氏名

カマタ セイイチロウ

鎌田 清一郎

職名

教授 (https://researchmap.jp/read0113140/)

所属

(大学院情報生産システム研究科)

連絡先

メールアドレス

メールアドレス
kam@waseda.jp

URL等

WebページURL

http://www.waseda.jp/sem-kamlabo011/(教育研究内容紹介)

研究者番号
00204602

本属以外の学内所属

学内研究所等

理工学総合研究センター

兼任研究員 2003年-2006年

理工学術院総合研究所(理工学研究所)

兼任研究員 2006年-2018年

理工学術院総合研究所(理工学研究所)

兼任研究員 2018年-

学歴・学位

学位

博士(工学) 論文 九州工業大学

所属学協会

電子情報通信学会

情報処理学会

映像情報メディア学会

Institute of Electrical and Electronics Engineers

委員歴・役員歴(学外)

2006年-2006年第11回知能メカトロニクスワークショップ 論文委員
2000年-2002年映像情報メディア学会編集委員会OB論文委員 論文委員
2000年-2001年電気学会非整備環境におけるパターン認識応用調査専門委員会委員 専門委員会委員
2000年-2001年電子情報通信学会画像工学専門委員会委員 専門委員会委員
2000年-2001年映像情報メディア学会九州支部評議員 評議員
2004年-Technical Program Committee for IEEE Asia-Pacific Conference on Circuits and Systems (APCCAS 2004), IEEE CAS Society- Taipei Chapter Technical Program Committee Member − 2004
2001年-2001年Program Committee for Sixth International Conference on Pattern Recognition and Information Processing, International Association on Pattern Recognition Program Committee Member
2000年-2000年Program Committee for Second International Conference on Digital Information Processing and Control in Extreme Situations, Academy of Sciences of Belarus Program Committee Member

受賞

最優秀論文賞(ICIEV2018)

2018年06月授与機関:ICIEV2018

最優秀論文賞(映像メディア処理シンポジウム,IMPS)

2017年09月授与機関:映像メディア処理シンポジウム

受賞者(グループ):杉本憲治郎、鎌田清一郎

最優秀論文賞(ICARCV2010)

2010年12月授与機関:ICARCV2010

研究分野

キーワード

画像情報処理、パターン認識、マルチメディア、信号処理

科研費分類

情報学 / 人間情報学 / 知能ロボティクス

工学 / 電気電子工学 / 通信・ネットワーク工学

共同研究希望テーマ

パターンマッチング応用に関する研究開発

希望連携機関:産学連携、民間を含む他機関等との共同研究等

目的:技術相談、受託研究

研究テーマ履歴

1988年-2014年画像処理

研究テーマのキーワード:画像処理

個人研究

1990年-2014年パターン認識

研究テーマのキーワード:パターン認識

個人研究

論文

Infrared Image Colorization Using a S-Shape Network

DONG, Ziyue DONG, 鎌田 清一郎, BRECKON, Toby

2018 25th IEEE International Conference on Image Processing (ICIP)査読有りp.2242 - 22462018年10月-2018年10月 

DOI

Sparse Graph based Deep Learning Networks for Face Recognition

WU,Renjie, 鎌田 清一郎

IEICE Transactions on Information and Systems査読有りE101-D(9)p.2209 - 22192018年09月-2018年09月 

DOI

Nuclei Segmentation of Cervical Cell Images based on Intermediate Segment Qualifier

WANG, Rui, 鎌田 清一郎

IEEE Proceedings of International Conference on Pattern Recognition (ICPR2018)査読有り2018年08月-2018年08月 

Frontal Gait Recognition from Incomplete RGB-D Streams Using Gait Cycle Analysis

ZOU,Wenyun, 鎌田 清一郎

Proc. of Joint 7th International Conference on Informatics, Electronics & Vision (ICIEV)査読有りp.453 - 4582018年06月-2018年06月 

Character Recognition in Japanese Historical Documents via Adaptive Multi-Region Model

WANG, Yueyu, 鎌田 清一郎

Proc. Joint 7th International Conference on Informatics, Electronics & Vision (ICIEV)査読有りp.404 - 4092018年06月-2018年06月 

Deep Neural Networks with Mixture of Experts Layers for Complex Event Recognition from Images

LI, Mingyao, 鎌田 清一郎

Proc. Joint 7th International Conference on Informatics, Electronics & Vision (ICIEV)査読有りp.410 - 4152018年06月-2018年06月 

Deep Metric Learning with Online Hard and Soft Selection for Person Re-identification

Yu Mingyang, 鎌田 清一郎

Proc. Joint 7th International Conference on Informatics, Electronics & Vision (ICIEV)査読有りp.426 - 4312018年06月-2018年06月 

Copy move image forgery detection based on polar Fourier representation

WANG Yitian, 鎌田 清一郎

International Journal of Machine Learning and Computing査読有り8(2)p.158 - 1632018年04月-2018年04月 

DOI

Guided Image Filtering with Arbitrary Window Function

福嶋 慶繁, 杉本 憲治郎, 鎌田 清一郎

Proc. IEEE International Conference, Acoustic, Signal Processing査読有りp.1523 - 15272018年04月-2018年04月 

DOI

Universal Approach for DCT-Based Constant-Time Gaussian Filter with Moment Preservation

杉本 憲治郎, 京地 清介, 鎌田 清一郎

Proc. IEEE International Conference, Acoustic, Signal Processing査読有りp.1498 - 15022018年04月-2018年04月 

DOI

Nuclei detection based on secant normal voting with skipping ranges in stained histopathological images

Lim, Xueting; Sugimoto, Kenjiro; Kamata, Sei Ichiro

IEICE Transactions on Information and SystemsE101D(2)p.523 - 5302018年02月-2018年02月 

DOIScopus

詳細

ISSN:09168532

概要:© Copyright 2018 The Institute of Electronics, Information and Communication Engineers. Seed detection or sometimes known as nuclei detection is a prerequisite step of nuclei segmentation which plays a critical role in quantitative cell analysis. The detection result is considered as accurate if each detected seed lies only in one nucleus and is close to the nucleus center. In previous works, voting methods are employed to detect nucleus center by extracting the nucleus saliency features. However, these methods still encounter the risk of false seeding, especially for the heterogeneous intensity images. To overcome the drawbacks of previous works, a novel detection method is proposed, which is called secant normal voting. Secant normal voting achieves good performance with the proposed skipping range. Skipping range avoids over-segmentation by preventing false seeding on the occlusion regions. Nucleus centers are obtained by mean-shift clustering from clouds of voting points. In the experiments, we show that our proposed method outperforms the comparison methods by achieving high detection accuracy without sacrificing the computational efficiency.

Deep Face Recognition under Eyeglass and Scale Variation Using Extended Siamese Network

Fan Qiu, Sei-ichiro Kamata and Lizhuang Ma

Proc. Asian Conference on Pattern Recognition査読有りp.471 - 4762017年11月-2017年11月 

Radical Region based CNN for Offline Handwritten Chinese Character Recognition

Weike Luo and Sei-ichiro Kamata

Proc. Asian Conference on Pattern Recognition査読有りp.542 - 5472017年11月-2017年11月 

Supervised Two-Step Hash Learning for Efficient Image Retrieval

Xinhui Wu, Sei-ichiro Kamata and Lizhuang Ma

Proc. Asian Conference on Pattern Recognition査読有りp.184 - 1892017年11月-2017年11月 

Face Recognition via Deep Sparse Graph Neural Networks

Renjie WU, Sei-ichiro KAMATA, Toby BRECKON

Proc. Workshop on Deep Learning on Irregular Domains, in British Machine Vision Conference 2017p.1 - 102017年09月-2017年09月 

Complex Coefficient Representation of IIR Bilateral Filter

Norishige Fukushima, Kenjiro Sugimoto and Sei-ichiro Kamata

Proc. IEEE international Conference on Image Processing査読有りp.2458 - 24622017年09月-2017年09月 

DOI

Robust registration of serial cell microscopic images using 3D Hilbert scan search

Lai, Yongwen; Kamata, Sei Ichiro; Fu, Zhizhong

Proceedings of the 15th IAPR International Conference on Machine Vision Applications, MVA 2017p.530 - 5332017年07月-2017年07月 

DOIScopus

詳細

概要:© 2017 MVA Organization All Rights Reserved. Microscopic images are quite helpful for us to observe the details of cells because of its high resolution. Furthermore it can benefit biologists and doctors to view the cell structure from any aspect by using a serial images to generate 3D cell structure. However each cell slice is placed at the microscopy respectively, which will bring in the arbitrary rotation and translation among the serial slices. What's more, the sectioning process will destroy the cell structure such as tearing or warping. Therefore we must register the serial slices before rendering the volume data in 3D. In this paper we propose a robust registration algorithm based on an improved 3D Hilbert scam search. Besides we put forward a simple but effective method to remove false matching in consecutive images. Finally we correct the local deformation based on optical-flow theory and adopt multi-resolution method. Our algorithm is tested, on a serial microscopy kidney cell images, and the experimental results show how accurate and robust of our method is.

Two-stage cross-based stereo disparity refinement

Xu, Zonglin; Kamata, Sei Ichiro; Zhang, Qieshi

Proceedings of the 15th IAPR International Conference on Machine Vision Applications, MVA 2017p.420 - 4232017年07月-2017年07月 

DOIScopus

詳細

概要:© 2017 MVA Organization All Rights Reserved. This paper proposed a disparity refinement method based on two-stage cross. First stage is anti-texture cross-based support region construction to build proper support regions for error pixels without being influenced by texture. Based on the support regions, second stage of the method is proposed, which is called weighted cross-based updating method. The experiments show that the proposed method could build the support region accurately and improve the accuracy of the disparity map in final results with fast speed, compared to other tree-based algorithms. It also outperforms the existing disparity refinement methods in preserving the boundaries of objects in the final disparity map.

Discriminative Histogram Intersection Metric Learning and Its Applications

Hao, Peng Yi; Xia, Yang; Li, Xiao Xin; Kamata, Sei ichiro; Chen, Sheng Yong

Journal of Computer Science and Technology32(3)p.507 - 5192017年05月-2017年05月 

DOIScopus

詳細

ISSN:10009000

概要:© 2017, Springer Science+Business Media New York. In this paper, a novel method called discriminative histogram intersection metric learning (DHIML) is proposed for pair matching and classification. Specifically, we introduce a discrimination term for learning a metric from binary information such as same/not-same or similar/dissimilar, and then combine it with the classification error for the discrimination in classifier construction. Compared with conventional approaches, the proposed method has several advantages. 1) The histogram intersection strategy is adopted into metric learning to deal with the widely used histogram features effectively. 2) By introducing discriminative term and classification error term into metric learning, a more discriminative distance metric and a classifier can be learned together. 3) The objective function is robust to outliers and noises for both features and labels in the training. The performance of the proposed method is tested on four applications: face verification, face-track identification, face-track clustering, and image classification. Evaluations on the challenging restricted protocol of Labeled Faces in the Wild (LFW) benchmark, a dataset with more than 7 000 face-tracks, and Caltech-101 dataset validate the robustness and discriminability of the proposed metric learning, compared with the recent state-of-the-art approaches.

Fast bilateral filter for multichannel images via soft-assignment coding

Sugimoto, Kenjiro; Fukushima, Norishige; Kamata, Sei Ichiro

2016 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference, APSIPA 20162017年01月-2017年01月 

DOIScopus

詳細

概要:© 2016 Asia Pacific Signal and Information Processing Association. This paper presents an acceleration method of the bilateral filter (BF) for multi-channel images. In most existing acceleration methods, the BF is approximated by an appropriate combination of convolutions. A major purpose under this framework is to achieve sufficient approximate accuracy by as few convolutions as possible. However, state-of-the-art methods for multi-channel images still requires hundreds of (e.g., 256) convolutions to achieve sufficient accuracy. The proposed method reduces the number of convolutions without a loss in accuracy via soft-assignment coding. This approach enables us to take two major advantages that two state-of-the-art methods (scalar quantization with linear interpolation and vector quantization) have individually provided. Experiments show that the proposed method can produce sufficiently-accurate resulting images by using 64-80 convolutions only.

Eigen-aging reference coding for cross-age face verification and retrieval

Tang, Kaihua; Tang, Kaihua; Kamata, Sei Ichiro; Hou, Xiaonan; Ding, Shouhong; Ma, Lizhuang

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)10113 LNCSp.389 - 4032017年01月-2017年01月 

DOIScopus

詳細

ISSN:03029743

概要:© Springer International Publishing AG 2017Recent works have achieved near or over human performance in traditional face recognition under PIE (pose, illumination and expression) variation. However, few works focus on the cross-age face recognition task, which means identifying the faces from same person at different ages. Taking human-aging into consideration broadens the application area of face recognition. It comes at the cost of making existing algorithms hard to maintain effectiveness. This paper presents a new reference based approach to address cross-age problem, called Eigen-Aging Reference Coding (EARC). Different from other existing reference based methods, our reference traces eigen faces instead of specific individuals. The proposed reference has smaller size and contains more useful information. To the best of our knowledge, we achieve state-of-the-art performance and speed on CACD dataset, the largest public face dataset containing significant aging information.

Efficient Keypoint Detection and Description Using Filter Kernel Decomposition in Scale Space

Ryo Okutani, Kenjiro Sugimoto and Sei-ichiro Kamata

Proc. IEEE International Conference on Image Processing査読有りp.31 - 352016年09月-2016年09月 

DOI

A Jointly Local Structured Sparse Deep Learning Network for Face Recognition

Renjie Wu and Sei-ichiro Kamata

Proc. IEEE international Conference on Image Processing査読有りp.3026 - 30302016年09月-2016年09月 

DOI

Efficient Constant-Time Bilateral Filter via Eigenvalue Decomposition

Kenjiro Sugimoto, Toby Breckon and Sei-ichiro Kamata

Proc. IEEE international Conference on Image Processing査読有りp.3319 - 33232016年09月-2016年09月 

DOI

Adaptive Sampling and Wavelet Tree based Compressive Sensing for MRI Reconstruction

Qieshi Zhang and Sei-ichiro Kamata

Proc. IEEE international Conference on Image Processing査読有りp.2524 - 25282016年09月-2016年09月 

DOI

Efficient keypoint detection and description via polynomial regression of scale space

Okutani, Ryo; Sugimoto, Kenjiro; Kamata, Sei Ichiro

ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings2016-Mayp.1357 - 13612016年05月-2016年05月 

DOIScopus

詳細

ISSN:15206149

概要:© 2016 IEEE.Keypoint detection and description using approximate continuous scale space are more efficient techniques than typical discretized scale space for achieving more robust feature matching. However, this state-of-the-art method requires high computational complexity to approximately reconstruct, or decompress, the value at an arbitrary point in scale space. Specifically, it has O(M2) computational complexity where M is an approximation order. This paper presents an efficient scale space approach that provides decompression operation with O(M) complexity without a loss of accuracy. As a result of the fact that the proposed method has much fewer variables to be solved, the least-square solution can be obtained through normal equation. This is easier to solve than the existing method which employs Karhunen-Loeve expansion and generalized eigenvalue problem. Experiments revealed that the proposed method performs as expected from the theoretical analysis.

Learning discriminative and shareable patches for scene classification

Ni, Shoucheng; Zhangg, Qieshi; Kamata, Sei Ichiro; Zhang, Chongyang

ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings2016-Mayp.1317 - 13212016年05月-2016年05月 

DOIScopus

詳細

ISSN:15206149

概要:© 2016 IEEE.This paper addresses the problem of scene classification and proposes learning discriminative and shareable patches (LDSP) method. The main idea of learning discriminative and shareable patches is to discover patches that exhibit both large between-class dissimilarity (discriminative) and large within-class similarity (shareable). A novel and efficient re-clustering, based on co-occurrence relationship of first-step clustering, is proposed and conducted to further enhance the visual similarity of patches within each cluster. In order to establish appropriate criteria for selecting desired patches, a condensed representation of image features called feature epitome is introduced. In the classification, a patch feature involving pre-trained convolutional neural network model is investigated. The experimental result outperforms existing single-feature methods on MIT 67 scene benchmark in term of mean Accuracy Precision.

A novel color space based on RGB color barycenter

Zhang, Qieshi; Kamata, Sei Ichiro

ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings2016-Mayp.1601 - 16052016年05月-2016年05月 

DOIScopus

詳細

ISSN:15206149

概要:© 2016 IEEE.Color space is one of the bases in the image processing area. Suitable color space can give the suitable description of colors for variant processing. However, in the image processing area, the existing color space cannot show the suitable distribution in color and lightness. In this paper, a novel color space based on RGB color barycenter (RGB-CB) is proposed to describe the color and lightness more intuitively. To prove the effectiveness of the proposed color space, YUV, HSV, L, and IPT color spaces are discussed and compared. Experimental results show the proposed color space can perform better effect than other color space in image processing.

Integrating multiple global and local features by product sparse coding for image retrieval

Tian, Li; Jia, Qi; Kamata, Sei Ichiro

IEICE Transactions on Information and SystemsE99D(3)p.731 - 7382016年03月-2016年03月 

DOIScopus

詳細

ISSN:09168532

概要:Copyright © 2016 The Institute of Electronics, Information and Communication Engineers.In this study, we propose a simple, yet general and powerful framework of integrating multiple global and local features by Product Sparse Coding (PSC) for image retrieval. In our framework, multiple global and local features are extracted from images and then are transformed to Trimmed-Root (TR)-features. After that, the features are encoded into compact codes by PSC. Finally, a two-stage ranking strategy is proposed for indexing in retrieval. We make three major contributions in this study. First, we propose TR representation of multiple image features and show that the TR representation offers better performance than the original features. Second, the integrated features by PSC is very compact and effective with lower complexity than by the standard sparse coding. Finally, the two-stage ranking strategy can balance the efficiency and memory usage in storage. Experiments demonstrate that our compact image representation is superior to the state-of-the-art alternatives for large-scale image retrieval.

Fisheye Image Correction Based on Straight-Line Detection and Preservation

Zhang, Qieshi; Kamata, Sei Ichiro

Proceedings - 2015 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2015p.1793 - 17972016年01月-2016年01月 

DOIScopus

詳細

概要:© 2015 IEEE.Fisheye lenses are widely used when the users want to capture the image/video with wide field of view (FoV) which is particularly suited to surveillance monitoring and vehicle camera. However, no projection from the actual scene in wide FoV image can avoid the distortion. If this problem cannot be solved, the fisheye image will difficult be used for object detection or analysis due to the distorted shapes of the scene objects. To correct this problem and obtain the natural-looking image, a two-step correction approach is proposed. Firstly, adaptive latitude and longitude correction are presented and the Hough transform is used to detect and estimate the straight-line. Secondly, the straight-line preserving and orientation, consistency based optimization is examined to obtain the final correction result. To compare the effectiveness of the proposed method, some fisheye correction methods are discussed. The experimental results demonstrate that the proposed method can obtain the coherent natural-looking.

Accurate system for automatic pill recognition using imprint information

Yu, Jiye; Chen, Zhiyuan; Kamata, Sei ichiro; Yang, Jie

IET Image Processing9(12)p.1039 - 10472015年12月-2015年12月 

DOIScopus

詳細

ISSN:17519659

概要:© The Institution of Engineering and Technology 2015.With rapidly advancing of contemporary medicine, it is necessary to help people identify various kinds of pills to prevent the adverse pill events. In this study, a high-accuracy automatic pill recognition system is proposed for accurate and automatic pill recognition. As pill imprint is main distinction between different pills, this system proposes algorithms on both imprint extraction and description parts to make use of imprint information. First, proposed modified stroke width transform is adopted to extract the imprint by detecting coherent strokes of imprint on the pill. Moreover, image segmentation by Loopy belief propagation is also added on printed imprint pills to solve the incoherent and coarse stroke problem. Second, a new descriptor named two-step sampling distance sets is proposed for accurate imprint description and successfully cut down the noise on extracted imprint. This strategy is based on the imprint partition - partitions the imprint on the basis of separated strokes, fragments and noise points. Recognition experiments are applied on extensive databases and result shows 90.46% rank-1 matching accuracy and 97.16% on top five ranks when classifying 12 500 query pill images into 2500 categories.

Robust road lane detection using extremal-region enhancement

Gu, Jingchen; Gu, Jingchen; Zhang, Qieshi; Kamata, Sei Ichiro

Proceedings - 3rd IAPR Asian Conference on Pattern Recognition, ACPR 2015査読有りp.519 - 5232015年11月-2015年11月 

DOIScopus

詳細

概要:© 2015 IEEE.Road lane detection is a key problem in advanced driver-assistance systems (ADAS). For solving this problem, vision-based detection methods are widely used and are generally focused on edge information. However, only using edge information leads to miss detection and error detection in various road conditions. In this paper, we propose a neighbor-based image conversion method, called extremal-region enhancement. The proposed method enhances the white lines in intensity, hence it is robust to shadows and illuminance changes. Both edge and shape information of white lines are extracted as lane features in the method. In addition, we implement a robust road lane detection algorithm using the extracted features and improve the correctness through probability tracking. The experimental result shows an average detection rate increase of 13.2% over existing works.

Compressive bilateral filtering

Sugimoto, Kenjiro; Kamata, Sei Ichiro

IEEE Transactions on Image Processing24(11)p.3357 - 33692015年11月-2015年11月 

DOIScopus

詳細

ISSN:10577149

概要:© 2015 IEEE.this paper presents an efficient constant-time bilateral filter that produces a near-optimal performance tradeoff between approximate accuracy and computational complexity without any complicated parameter adjustment, called a compressive bilateral filter (CBLF). The constant-time means that the computational complexity is independent of its filter window size. Although many existing constant-time bilateral filters have been proposed step-by-step to pursue a more efficient performance tradeoff, they have less focused on the optimal tradeoff for their own frameworks. It is important to discuss this question, because it can reveal whether or not a constant-time algorithm still has plenty room for improvements of performance tradeoff. This paper tackles the question from a viewpoint of compressibility and highlights the fact that state-of-the-art algorithms have not yet touched the optimal tradeoff. The CBLF achieves a near-optimal performance tradeoff by two key ideas: 1) an approximate Gaussian range kernel through Fourier analysis and 2) a period length optimization. Experiments demonstrate that the CBLF significantly outperforms state-of-the-art algorithms in terms of approximate accuracy, computational complexity, and usability.

Autonomous driving experiments by small electric vehicle in simulated road

Wu, Nan; Zhang, Qieshi; Pan, Xun; Beier, Hu; Ogai, Harutoshi; Kamata, Sei Ichiro; Inujima, Hiroshi; Tateno, Shigeyuki

2015 54th Annual Conference of the Society of Instrument and Control Engineers of Japan, SICE 2015p.1451 - 14522015年09月-2015年09月 

DOIScopus

詳細

概要:© 2015 The Society of Instrument and Control Engineers-SICE. In the city of Kitakyushu, a quarter more of people are older than 65 years old. Especially a large proportion of elderly people are living on their own. So how to let old people easily access the communal facilities like hospitals need be attended. Therefore, the bad connection between public transportation and home is need to be solved. Based on the survey result and road situation in Kitakyushu city, a new type small single-seat electrical vehicle (sEV) is studied to provide a solution for elder to easily access the public transportation. In this research, low cost and safe automatic driving electrical vehicle based on limited number of sensors is focused.

Disparity refinement with stability-based tree for stereo matching

Ji, Yuhang; Zhang, Qieshi; Sugimoto, Kenjiro; Kamata, Sei Ichiro

IEEE Intelligent Vehicles Symposium, Proceedings2015-Augustp.469 - 4742015年08月-2015年08月 

DOIScopus

詳細

概要:© 2015 IEEE. This paper proposes a disparity refinement method with stability-based tree. By developing stability-based tree to evaluate and reconstruct support regions for error parts, the proposed method achieves effective performance in removing outliers. This approach further improves the quality of raw disparity map in stereo matching, which makes the local methods results comparable to the global ones. Experiments exhibit that the proposed method reduces more than 70% aggregation time compared with traditional tree method without loss of accuracy. It also outperforms existing disparity refinement methods in removing large error parts.

Disparity Refinement with Stability-based Tree for Stereo Matching

Yuhang JI, Qieshi ZHANG, Kenjiro SUGIMOTO, Sei-ichiro KAMATA

Proc. of 2015 IEEE Intelligent Vehicles Symposium査読有りp.469 - 4742015年06月-2015年06月 

Sparse decomposition learning based dynamic MRI reconstruction

Zhu, Peifei; Zhang, Qieshi; Kamata, Sei Ichiro

Proceedings of SPIE - The International Society for Optical Engineering94452015年01月-2015年01月 

DOIScopus

詳細

ISSN:0277786X

概要:© 2015 SPIE. Dynamic MRI is widely used for many clinical exams but slow data acquisition becomes a serious problem. The application of Compressed Sensing (CS) demonstrated great potential to increase imaging speed. However, the performance of CS is largely depending on the sparsity of image sequence in the transform domain, where there are still a lot to be improved. In this work, the sparsity is exploited by proposed Sparse Decomposition Learning (SDL) algorithm, which is a combination of low-rank plus sparsity and Blind Compressed Sensing (BCS). With this decomposition, only sparsity component is modeled as a sparse linear combination of temporal basis functions. This enables coefficients to be sparser and remain more details of dynamic components comparing learning the whole images. A reconstruction is performed on the undersampled data where joint multicoil data consistency is enforced by combing Parallel Imaging (PI). The experimental results show the proposed methods decrease about 15∼20% of Mean Square Error (MSE) compared to other existing methods.

Disparity estimation from monocular image sequence

Zhang, Qieshi; Kamata, Sei Ichiro

Proceedings of SPIE - The International Society for Optical Engineering94452015年01月-2015年01月 

DOIScopus

詳細

ISSN:0277786X

概要:© 2015 SPIE. This paper proposes a novel method for estimating disparity accurately. To achieve the ideal result, an optimal adjusting framework is proposed to address the noise, occlusions, and outliners. Different from the typical multi-view stereo (MVS) methods, the proposed approach not only use the color constraint, but also use the geometric constraint associating multiple frame from the image sequence. The result shows the disparity with a good visual quality that most of the noise is eliminated, the errors in occlusion area are suppressed and the details of scene objects are preserved.

Optimized curvelet-based empirical mode decomposition

Wu, Renjie; Zhang, Qieshi; Kamata, Sei Ichiro

Proceedings of SPIE - The International Society for Optical Engineering94452015年01月-2015年01月 

DOIScopus

詳細

ISSN:0277786X

概要:© 2015 SPIE. The recent years has seen immense improvement in the development of signal processing based on Curvelet transform. The Curvelet transform provide a new multi-resolution representation. The frame elements of Curvelets exhibit higher direction sensitivity and anisotropic than the Wavelets, multi-Wavelets, steerable pyramids, and so on. These features are based on the anisotropic notion of scaling. In practical instances, time series signals processing problem is often encountered. To solve this problem, the time-frequency analysis based methods are studied. However, the time-frequency analysis cannot always be trusted. Many of the new methods were proposed. The Empirical Mode Decomposition (EMD) is one of them, and widely used. The EMD aims to decompose into their building blocks functions that are the superposition of a reasonably small number of components, well separated in the time-frequency plane. And each component can be viewed as locally approximately harmonic. However, it cannot solve the problem of directionality of high-dimensional. A reallocated method of Curvelet transform (optimized Curvelet-based EMD) is proposed in this paper. We introduce a definition for a class of functions that can be viewed as a superposition of a reasonably small number of approximately harmonic components by optimized Curvelet family. We analyze this algorithm and demonstrate its results on data. The experimental results prove the effectiveness of our method.

Efficient Constant-time Gaussian Filtering with Sliding DCT/DST-5and Dual-domain Error Minimization

Sugimoto, Kenjiro; Kamata, Sei Ichiro

ITE Transactions on Media Technology and Applications3(1)p.12 - 212015年01月-2015年01月 

Scopus

詳細

概要:© 2015 by ITE Transactions on Media Technology and Applications (MTA). This paper presents an efficient constant-time algorithm for Gaussian filtering and also Gaussian derivative filtering that provides a high approximate accuracy in a low computational complexity regardless of its filter window size. The proposed algorithm consists of two key techniques: second-order shift properties of the Discrete Cosine/Sine Transforms type-5 and dual-domain error minimization for finding optimal parameters. The former enables us to perform filtering in fewer number of arithmetic operations as compared than some state-of-the-art algorithms without integral images. The latter enables us to find the optimal filter size that provides the most accurate filter kernel approximation. Experiments show that the proposed algorithm clearly outperforms state-of-the-art ones in computational complexity, approximate accuracy, and accuracy stability.

Multi-histogram mapping and fusion based image contrast enhancement

Zhang, Qieshi; Kamata, Sei Ichiro

ITE Transactions on Media Technology and Applications3(1)p.2 - 112015年01月-2015年01月 

Scopus

詳細

概要:© 2015 by ITE Transactions on Media Technology and Applications (MTA). In this paper, a contrast enhancement method called Multi-Histogram Mapping and Fusion (MHMF) is proposed for color images. Histogram analysis based method has been successfully applied to contrast enhancement in some applications, but they are hard to enhance the dark and bright regions simultaneously for back-light images. To solve this problem, the color barycenter model (CBM) is extended to separate the color image into lightness and chroma components. Then multi-histogram mapping (MHM) is used to map the lightness component of one single color image into several new lightness components with different contrast. These new components are divided into several patches and the best patches are selected by calculating image entropy. Finally, the selected patches are fused to create the enhanced image, and mix-Gaussian filter is applied to remove the sharp transition. The experimental results show the effectiveness of proposed method comparing with other state-of-the-art methods.

Multi-Histogram Mapping and Fusion based Image Contrast Enhancement

Qieshi ZHANG and Sei-ichiro KAMATA

ITE Transactions on Media Technology and Applications3(1)p.2 - 112015年01月-2015年01月 

Efficient Constant-time Gaussian Filtering with Sliding DCT/DST-5 and Dual-domain Error Minimization

Kenjiro SUGIMOTO and Sei-ichiro KAMATA

ITE Transactions on Media Technology and Applications査読有り3(1)p.12 - 212015年01月-2015年01月 

O(1) transposed bilateral filtering for optimization

Sugimoto, Kenjiro; Shirai, Keiichiro; Kamata, Sei Ichiro

2014 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference, APSIPA 20142014年12月-2014年12月 

DOIScopus

詳細

概要:© 2014 Asia-Pacific Signal and Information Processing Ass. This paper presents an essential algorithm for optimization-based image processing using the bilateral filter (BLF), called constant-time transposed BLF (O(1) TBLF). Some iterative solvers for optimization problems require a pair of filters defined as multiplying a filter matrix or its transpose to vectorized images. Since the BLF can be described as a matrix form, its paired filter also exists, called a TBLF in this paper. BLF-based optimization achieves high smoothing performance; whereas, it requires much high computational complexity due to iterating both BLF and TBLF many times. Hence, this paper designs an O(1) TBLF algorithm to accelerate the iterative process. Experiments show that our O(1) TBLF runs in low complexity regardless of its filter window size and works effectively for flash/no-flash image integration via BLF-based optimization.

Pill Recognition Using Imprint Information by Two-step Sampling Distance Sets

Jiya YU, Zhiyuan CHEN and Sei-ichiro KAMATA

IEEE Proc. of Int. Conf. on Pattern Recognition2014年08月-

Development of autonomous small EV in Japan aging society

Wu, Nan; Ogai, Harutoshi; Ohshiro, Masakuni; Kamata, Seiichirou; Tateno, Shigeyuki; Uchida, Akira; Kai, Masahiko; Iida, Makio; Sano, Yuji

IFAC Proceedings Volumes (IFAC-PapersOnline)3(PART 1)p.966 - 9722014年01月-2014年01月 

DOIScopus

詳細

ISSN:14746670

概要:In Kitakyushu city, more than 25 % people are older than 65 years old. The roads in residential area for the people are very narrow, very steep slope and vulnerable. Some needs assessment for small electric vehicle at some event and community activity was done. Based on the survey result and topographical features, the small electric vehicle for elderly person was selected and was modified and the automatic driving system was built. Automatic driving and platooning using Zigbee or Digimesh to exchange driving data between vehicles were built and tested. © 2014 IFAC.

Pill recognition using imprint information by two-step sampling distance sets

Yu, Jiye; Chen, Zhiyuan; Kamata, Sei Ichiro

Proceedings - International Conference on Pattern Recognitionp.3156 - 31612014年01月-2014年01月 

DOIScopus

詳細

ISSN:10514651

概要:© 2014 IEEE. Huge variety of medicine cures diseases. But unlabeled pills sometimes confuse people, even causing adverse drug events. This paper introduces a high accuracy automatic pill recognition method based on pill imprint which is a main discriminative factor between different pills. To describe the imprint information clearly, we propose a Two-step Sampling Distance Sets (TSDS) descriptor based on Distance Sets (DS) using a two-step sampling strategy. The two-step sampling strategy applies a resampling according to imprint segmentation, which divides an imprint into separated strokes, fragments and noise points. The TSDS is able to take control over the selection of feature points, aiming to cut down the noise points and unwished fragments generated by imprint extraction which will cause disturbance on recognition. In the aspect of the imprint extraction, we preprocess the pill image by dynamic contrast adjustment to cope with the exposure problem. Modified Stroke Width Transform (MSWT) is used to extract the imprint by detecting the coherent strokes on the pill. Finally, several experimental results have shown 86.01%, rank-1 matching accuracy, and 93.64%, within top 5 ranks, when classifying pills into 2500 categories.

Improved Color Barycenter Model and Its Separation for Road Sign Detection

Qieshi ZHANG and Sei-ichiro KAMATA

IEICE Trans. on Information and SystemsE96-D(12)p.2839 - 28492013年12月-

Improved color barycenter model and its separation for road sign detection

Zhang, Qieshi; Kamata, Sei Ichiro

IEICE Transactions on Information and SystemsE96-D(12)p.2839 - 28492013年12月-2013年12月 

DOIScopus

詳細

ISSN:09168532

概要:This paper proposes an improved color barycenter model (CBM) and its separation for automatic road sign (RS) detection. The previous version of CBM can find out the colors of RS, but the accuracy is not high enough for separating the magenta and blue regions and the influence of number with the same color are not considered. In this paper, the improved CBM expands the barycenter distribution to cylindrical coordinate system (CCS) and takes the number of colors at each position into account for clustering. Under this distribution, the color information can be represented more clearly for analyzing. Then aim to the characteristic of barycenter distribution in CBM (CBM-BD), a constrained clustering method is presented to cluster the CBM-BD in CCS. Although the proposed clustering method looks like conventional K-means in some part, it can solve some limitations of K-means in our research. The experimental results show that the proposed method is able to detect RS with high robustness.© 2013 The Institute of Electronics, Information and Communication Engineers.

Fast Gaussian filter with second-order shift property of DCT-5

Sugimoto, Kenjiro; Kamata, Sei Ichiro

2013 IEEE International Conference on Image Processing, ICIP 2013 - Proceedingsp.514 - 5182013年12月-2013年12月 

DOIScopus

詳細

概要:This paper presents an efficient constant-time Gaussian filter which provides a high accuracy at a low cost over a wide range of scale σ. It requires only 14 multiplications per pixel in image filtering regardless of σ, which is fewer than state-of-the-art constant-time Gaussian filters. Main ideas of the paper are as follows: 1) introducing a second-order shift property of the discrete cosine transform type-5 (DCT-5) to convolve cosines faster, and 2) suppressing error propagation caused by the shift property. Experiments in image processing show that the proposed algorithm is 3.7× faster than a state-of-the-art recursive Gaussian filter and comparable to that of ±3σ-supported Gaussian convolution with σ = 2.33. The output accuracy is stable at around 80 [dB] all over σ [1, 128]. © 2013 IEEE.

A foreground object based quantitative assessment of dense stereo approaches for use in automotive environments

Hamilton, Oliver K.; Breckon, Toby P.; Bai, Xuejiao; Kamata, Sei Ichiro

2013 IEEE International Conference on Image Processing, ICIP 2013 - Proceedingsp.418 - 4222013年12月-2013年12月 

DOIScopus

詳細

概要:There has been significant recent interest in stereo correspondence algorithms for use in the urban automotive environment [1, 2, 3]. In this paper we evaluate a range of dense stereo algorithms, using a unique evaluation criterion which provides quantitative analysis of accuracy against range, based on ground truth 3D annotated object information. The results show that while some algorithms provide greater scene coverage, we see little differentiation in accuracy over short ranges, while the converse is shown over longer ranges. Within our long range accuracy analysis we see a distinct separation of relative algorithm performance. This study extends prior work on dense stereo evaluation of Block Matching (BM)[4], Semi-Global Block Matching (SGBM)[5], No Maximal Disparity (NoMD)[6], Cross[7], Adaptive Dynamic Programming (AdptDP)[8], Efficient Large Scale (ELAS)[9], Minimum Spanning Forest (MSF)[10] and Non-Local Aggregation (NLA)[11] using a novel quantitative metric relative to object range. © 2013 IEEE.

A fast multi-view based specular removal approach for pill extraction

Wang, Chengjie; Kamata, Sei Ichiro; Ma, Lizhuang

2013 IEEE International Conference on Image Processing, ICIP 2013 - Proceedingsp.4126 - 41302013年12月-2013年12月 

DOIScopus

詳細

概要:This paper presents a novel approach to remove the specular reflections on the transparent plastic medicine package and automatically extract the randomly distributed pills inside. In this approach, three cameras are employed to take images of the package from different viewpoints. And these three images are used as input image set while the output is a series of small images of a single pill. And these images can be directly applied to the traditional single pill recognition algorithms. The experimental results show the reliability of our approach by measuring correct detection rate (100%), false detection rate (0%) and pill separation accuracy (98.4%). And the proposed method processes a set of three 725×725 sized images at 0.15s averagely on a Core i5-2400 3.1GHz PC. © 2013 IEEE.

Maximum correntropy criterion for discriminative dictionary learning

Hao, Pengyi; Kamata, Sei Ichiro

2013 IEEE International Conference on Image Processing, ICIP 2013 - Proceedingsp.4325 - 43292013年12月-2013年12月 

DOIScopus

詳細

概要:In this paper, a novel discriminative dictionary learning with pairwise constraints by maximum correntropy criterion is proposed for pair matching problem. Comparing with the conventional dictionary learning approaches, the proposed method has several advantages: (i) It can deal with the outliers and noises problem more efficiently during the reconstruction step. (ii) It can be effectively solved by half-quadratic optimization algorithm, and in each iteration step, the complex optimization problem can be reduced to a general problem that can be efficiently solved by feature-sign search optimization. (iii) The proposed method is capable of analyzing non-Gaussian noise to reduce the influence of large outliers substantially, resulting in a robust and discriminative dictionary. We test the performance of the proposed method on two applications: face verification on the challenging restricted protocol of Labeled Faces in the Wild (LFW) benchmark and face-track identification on a dataset with more than 7,000 face-tracks. Compared with the recent state-of-the-art approaches, the outstanding performance of the proposed method validates its robustness and discriminability. © 2013 IEEE.

An efficient video retrieval scheme based on facial signatures

Hao, Pengyi; Kamata, Sei Ichiro

2013 IEEE International Conference on Image Processing, ICIP 2013 - Proceedingsp.2699 - 27032013年12月-2013年12月 

DOIScopus

詳細

概要:The topic of retrieving videos containing a desired person by just using facial content has many applications like video surveillance, social network, etc. In this paper, we propose a compact, discriminative and low-dimensional signature to describe an person with a set of high-dimensional features. The signature is generated by linear discriminant analysis with maximum correntropy criterion that is robust to outliers and noises. Based on the proposed signatures, a new video retrieval scheme is given for fast finding the desired videos by measuring the similarities between the signature of a query and the ones in the dataset. Evaluations on a large dataset of videos show that the proposed video retrieval scheme has the potential to substantially reduce the response time and slightly increase the mean average precision of retrieval. © 2013 IEEE.

Facial signatures for fast individual retrieval from video dataset

Hao, Pengyi; Kamata, Sei Ichiro

Proceedings - IEEE International Conference on Multimedia and Expo2013年10月-2013年10月 

DOIScopus

詳細

ISSN:19457871

概要:The topic of retrieving videos containing a desired person by using the content of faces without any help of textual information has many interesting applications like video surveillance, social network, video mining, etc. However, face-by-face matching leads to an unacceptable response time for a video dataset with a large number of detected faces and may also reduce the accuracy of searching. Therefore, in this paper we propose a scheme to generate facial signatures for fast retrieving videos containing the same person with a query. First, we summarize each video as a set of person-oriented individuals based on detected faces, which are represented as high dimensional vectors in a feature space. Then, each person with a collection of high dimensional vectors is projected to a compact and reduced dimensionality representation that is called facial signature for this person. The projection is realized by constructing a matcher using linear discriminant analysis with maximum correntropy criterion optimization. In this research, two kinds of signatures are provided, which are called 1D facial signature and 2D facial signature. The proposed searching scheme can support two types of queries: face image and video clip. Evaluations on a large dataset of videos show reliable measurement of similarities by using facial signature to represent each person generated from videos and also demonstrate that the proposed searching scheme has the potential to substantially reduce the response time and slightly increase the mean average precision of retrieval. © 2013 IEEE.

Superfast Gaussian Filter with Second-Order Shift Property of DCT-V

Kenjiro SUGIMOTO and Sei-ichiro KAMATA

IEEE Proc. of Int. Conf. on Image Processingp.514 - 5182013年09月-

A Foreground Object Based Quntitative Assessment of Dense Stereo Approaches for Use in Automotive Environments

Oliver HAMILTON, Toby BRECKON, Xuejiao BAI and Sei-ichiro KAMATA

IEEE Proc. of Int. Conf. on Image Processingp.418 - 4222013年09月-

A Fast Multi-View Based Specular Removal Approach for Pill Extraction

Chengjie WANG, Sei-ichiro KAMATA and Lizhuang MA

IEEE Proc. of Int. Conf. on Image Processingp.4126 - 41302013年09月-

An Efficient Video Retrieval Scheme Based on Facial Signatures

Pengyi HAO and Sei-ichiro KAMATA

IEEE Proc. of Int. Conf. on Image Processingp.2699 - 27032013年09月-

Maximum Correntropy Criterion for Discriminative Dictionary Learning

Pengyi HAO and Sei-ichiro KAMATA

IEEE Proc. of Int. Conf. on Image Processingp.4325 - 43292013年09月-

Efficient Large-scale Video Retrieval Via Discriminative Signatures

Pengyi HAO and Sei-ichiro KAMATA

IEICE Trans. on Information and SystemsE96-D(8)p.1800 - 18102013年08月-

Efficient large-scale video retrieval via discriminative signatures

Hao, Pengyi; Kamata, Sei Ichiro

IEICE Transactions on Information and SystemsE96-D(8)p.1800 - 18102013年08月-2013年08月 

DOIScopus

詳細

ISSN:09168532

概要:The topic of retrieving videos containing a desired person from a dataset just using the content of faces without any help of textual information has many interesting applications like video surveillance, social network, video mining, etc. However, traditional face matching against a huge number of detected faces leads to an unacceptable response time and may also reduce the accuracy due to the large variations in facial expressions, poses, lighting, etc. Therefore, in this paper we propose a novel method to generate discriminative "signatures" for efficiently retrieving the videos containing the same person with a query. In this research, the signature is defined as a compact, discriminative and reduced dimensionality representation, which is generated from a set of high-dimensional feature vectors of an individual. The desired videos are retrieved based on the similarities between the signature of the query and those of individuals in the database. In particular, we make the following contributions. Firstly, we give an algorithm of two directional linear discriminant analysis with maximum correntropy criterion (2DLDA-MCC) as an extension to our recently proposed maximum correntropy criterion based linear discriminant analysis (LDA-MCC). Both algorithms are robust to outliers and noise. Secondly, we present an approach for transferring a set of exemplars to a fixed-length signature using LDA-MCC and 2DLDA-MCC, resulting in two kinds of signatures that are called 1D signature and 2D signature. Finally, a novel video retrieval scheme is given based on the signatures, which has low storage requirement and can achieve a fast search. Evaluations on a large dataset of videos show reliable measurement of similarities by using the proposed signatures to represent the identities generated from videos. Experimental results also demonstrate that the proposed video retrieval scheme has the potential to substantially reduce the response time and slightly increase the mean average precision of retrieval. Copyright © 2013 The Institute of Electronics, Information and Communication Engineers.

Facial Signatures for Fast Individual Retrieval from Video Dataset

Pengyi HAO and Sei-ichiro KAMATA

IEEE Proc. of Int. Conf. on Multimedia and Expo (ICME)p.1 - 62013年07月-

A Novel Color Descriptor for Road-Sign Detection

Qieshi ZHANG and Sei-ichiro KAMATA

IEICE Trans. on Fundamentals of Electronics, Communications and Computer SciencesE96-A(5)p.971 - 9792013年05月-

A novel color descriptor for road-sign detection

Zhang, Qieshi; Kamata, Sei Ichiro

IEICE Transactions on Fundamentals of Electronics, Communications and Computer SciencesE96-A(5)p.971 - 9792013年05月-2013年05月 

DOIScopus

詳細

ISSN:09168508

概要:This paper presents a novel color descriptor based on the proposed Color Barycenter Hexagon (CBH) model for automatic Road- Sign (RS) detection. In the visual Driver Assistance System (DAS), RS detection is one of the most important factors. The system provides drivers with important information on driving safety. Different color combinations of RS indicate different functionalities; hence a robust color detector should be designed to address color changes in natural surroundings. The CBH model is constructed with barycenter distribution in the created color triangle, which represents RS colors in a more compact way. For detecting RS, the CBH model is used to segment color information at the initial step. Furthermore, a judgment process is applied to verify each RS candidate through the size, aspect ratio, and color ratio. Experimental results show that the proposed method is able to detect RS with robust, accurate performance and is invariant to light and scale in more complex surroundings. Copyright © 2013 The Institute of Electronics.

Linear discriminant analysis with maximum correntropy criterion

Zhou, Wei; Kamata, Sei Ichiro

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)7724 LNCS(PART 1)p.500 - 5112013年04月-2013年04月 

DOIScopus

詳細

ISSN:03029743

概要:Linear Discriminant Analysis (LDA) is a famous supervised feature extraction method for subspace learning in computer vision and pattern recognition. In this paper, a novel method of LDA based on a new Maximum Correntropy Criterion optimization technique is proposed. The conventional LDA, which is based on L2-norm, is sensitivity to the presence of outliers. The proposed method has several advantages: first, it is robust to large outliers. Second, it is invariant to rotations. Third, it can be effectively solved by half-quadratic optimization algorithm. And in each iteration step, the complex optimization problem can be reduced to a quadratic problem that can be efficiently solved by a weighted eigenvalue optimization method. The proposed method is capable of analyzing non-Gaussian noise to reduce the influence of large outliers substantially, resulting in a robust classification. Performance assessment in several datasets shows that the proposed approach is more effectiveness to address outlier issue than traditional ones. © 2013 Springer-Verlag.

Face recognition with learned local curvelet patterns and 2-directional L1-norm based 2DPCA

Zhou, Wei; Kamata, Sei Ichiro

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)7728 LNCS(PART 1)p.109 - 1202013年04月-2013年04月 

DOIScopus

詳細

ISSN:03029743

概要:In this paper, we propose Learned Local Curvelet Patterns (LLCP) for presenting the local features of facial images. The proposed method is based on curvelet transform which can overcome the weakness of traditional Gabor wavelets in higher dimension, and better capture the curve singularities and hyperplane singularities of facial images. Different from wavelet transform, curvelet transform can effectively and efficiently approximate the curved edges with very few coefficients as well as taking space-frequency information into consideration. First, LLCP designs several learned codebooks from Curvelet filtered facial images. Then each facial image can be encoded into multiple pattern maps and finally block-based histograms of these patterns are concatenated into an histogram sequence to be used as a face descriptor. In order to reduce the face feature descriptor, 2-Directional L1-Norm Based 2DPCA ((2D)2PCA-L1) is proposed which is simultaneously considering the row and column directions for efficient face representation and recognition. Performance assessment in several face recognition problem shows that the proposed approach is superior to traditional ones. © 2013 Springer-Verlag.

L1-Norm Based Linear Discriminant Analysis: An Application to Face Recognition

Wei ZHOU and Sei-ichiro KAMATA

IEICE Trans. on Information and SystemsE96-D(3)p.550 - 5582013年03月-

L1-norm based linear discriminant analysis: An application to face recognition

Zhou, Wei; Kamata, Sei ichiro

IEICE Transactions on Information and SystemsE96-D(3)p.550 - 5582013年03月-2013年03月 

DOIScopus

詳細

ISSN:09168532

概要:Linear Discriminant Analysis (LDA) is a well-known feature extraction method for supervised subspace learning in statistical pattern recognition. In this paper, a novel method of LDA based on a new L1-norm optimization technique and its variances are proposed. The conventional LDA, which is based on L2-norm, is sensitivity to the presence of outliers, since it used the L2-norm to measure the between-class and within-class distances. In addition, the conventional LDA often suffers from the so-called small sample size (3S) problem since the number of samples is always smaller than the dimension of the feature space in many applications, such as face recognition. Based on L1-norm, the proposed methods have several advantages, first they are robust to outliers because they utilize the L1-norm, which is less sensitive to outliers. Second, they have no 3S problem. Third, they are invariant to rotations as well. The proposed methods are capable of reducing the influence of outliers substantially, resulting in a robust classification. Performance assessment in face application shows that the proposed approaches are more effectiveness to address outliers issue than traditional ones. Copyright © 2013 The Institute of Electronics, Information and Communication Engineers.

A new accurate pill recognition system using imprint information

Chen, Zhiyuan; Kamata, Sei Ichiro

Proceedings of SPIE - The International Society for Optical Engineering90672013年01月-2013年01月 

DOIScopus

詳細

ISSN:0277786X

概要:Great achievements in modern medicine benefit human beings. Also, it has brought about an explosive growth of pharmaceuticals that current in the market. In daily life, pharmaceuticals sometimes confuse people when they are found unlabeled. In this paper, we propose an automatic pill recognition technique to solve this problem. It functions mainly based on the imprint feature of the pills, which is extracted by proposed MSWT (modified stroke width transform) and described by WSC (weighted shape context). Experiments show that our proposed pill recognition method can reach an accurate rate up to 92.03% within top 5 ranks when trying to classify more than 10 thousand query pill images into around 2000 categories. © 2013 SPIE.

Development of safety assist system for ultra-small EV - Efforts for safe mobility in Kitakyushu aging society

Ogai, Harutoshi; Kamata, Seiichirou; Wu, Nan; Ishi, Taro; Uchida, Akira; Kai, Masahiko; Iida, Makio

20th ITS World Congress Tokyo 20132013年01月-2013年01月 

Scopus

詳細

概要:In Kitakyushu city there are more than a quarter of people are older than 65-year-old. The road in residential area of this area has specific characters including very narrow road, very steep slope and vulnerable roadbed. Some needs assessment of small electric vehicle at some event activity and community was done. Based on the survey result and topographical features, the small electric vehicle for elder was designed and modified and automatic driving system was built. Automatic driving and platooning using Zigbee to exchange driving data between vehicles were built and tested.

Face Representation and Recognition with Local Curvelet Patterns

Wei ZHOU, Alireza AHRARY, Sei-ichiro KAMATA

IEICE Trans. on Information and SystemsE95-D(12)p.3078 - 30872012年12月-

Face representation and recognition with local curvelet patterns

Zhou, Wei; Ahrary, Alireza; Kamata, Sei Ichiro

IEICE Transactions on Information and SystemsE95-D(12)p.3078 - 30872012年12月-2012年12月 

DOIScopus

詳細

ISSN:09168532

概要:In this paper, we propose Local Curvelet Binary Patterns (LCBP) and Learned Local Curvelet Patterns (LLCP) for presenting the local features of facial images. The proposed methods are based on Curvelet transform which can overcome the weakness of traditional Gabor wavelets in higher dimensions, and better capture the curve singularities and hyperplane singularities of facial images. LCBP can be regarded as a combination of Curvelet features and LBP operator while LLCP designs several learned codebooks from patch sets, which are constructed by sampling patches from Curvelet filtered facial images. Each facial image can be encoded into multiple pattern maps and block-based histograms of these patterns are concatenated into an histogram sequence to be used as a face descriptor. During the face representation phase, one input patch is encoded by one pattern in LCBP while multi-patterns in LLCP. Finally, an effective classifier called Weighted Histogram Spatially constrained Earth Mover's Distance (WHSEMD) which utilizes the discriminative powers of different facial parts, the different patterns and the spatial information of face is proposed. Performance assessment in face recognition and gender estimation under different challenges shows that the proposed approaches are superior than traditional ones. Copyright © 2012 The Institute of Electronics, Information and Communication Engineers.

Unsupervised people organization and its application on individual retrieval from videos

Hao, Pengyi; Kamata, Sei Ichiro

Proceedings - International Conference on Pattern Recognitionp.2001 - 20042012年12月-2012年12月 

Scopus

詳細

ISSN:10514651

概要:In this paper, a method named histogram intersection metric learning from scene tracks is proposed for automatic organizing people in videos. We make the following contributions: (i) learning histogram intersection distance instead of Mahalanobis distance for widely used face features; (ii) learning the metric from scene tracks without manually labeling any examples, which enables learning across large variations in pose, expression, occlusion and illumination with small number of face pairs and can distinguish different people powerfully. We firstly test face identification, track clustering, and people organization on a long film, then individual retrieval based on people organization from a large video dataset is evaluated, demonstrating significantly increased search quality with respect to previous approaches on this area. © 2012 ICPR Org Committee.

Fast image filtering by DCT-based kernel decomposition and sequential sum update

Sugimoto, Kenjiro; Kamata, Sei Ichiro

Proceedings - International Conference on Image Processing, ICIPp.125 - 1282012年12月-2012年12月 

DOIScopus

詳細

ISSN:15224880

概要:This paper presents an approximate Gaussian filter which can run in one-pass with high accuracy based on spectrum sparsity. This method is a modification of the cosine integral image (CII), which decomposes a filter kernel into few cosine terms and convolves each cosine term with an input image in constant time per pixel by using integral images and look-up tables. However, they require much workspace and high access cost. The proposed method solves the problem with no decline in quality by sequentially updating sums instead of integral images and by improving look-up tables, which accomplishes a one-pass approximation with much less workspace. A specialization for tiny kernels are also discussed for faster calculation. Experiments on image filtering show that the proposed method can run nearly two times faster than CII and also than convolution even with small kernel. © 2012 IEEE.

An efficient window-based stereo matching algorithm using foreground disparity concentration

Bai, Xuejiao; Kamata, Sei Ichiro

2012 12th International Conference on Control, Automation, Robotics and Vision, ICARCV 2012p.1352 - 13572012年12月-2012年12月 

DOIScopus

詳細

概要:In this paper, we present an efficient window-based stereo matching algorithm that especially focuses on foreground objects. For decades, there are a lot of researches about the stereo matching algorithms. However, most of methods concentrate on the entire pixels, which are time consuming and meaningless in the real applications. To strength the accuracy of stereo correspondence in foreground objects, a simple locally support-weight method based on the selected prime key is proposed in our algorithm. Moreover, a background pre-detection method is also employed to get a primary background checking map, which is used to reduce the number of computed pixels in the disparity selection. After the refinement of both foreground disparity map and background checking map, our algorithm obtains accurate disparity results on the foreground and separate it with the background by the correspondence search simultaneously. The experimental results based on the Middlebury stereo datasets demonstrate that our method can achieve a better performance on foreground disparity computing than many other support-weight methods in terms of both accuracy and computational efficiency. In addition, our proposals can make foreground objects detection easier at the same time. © 2012 IEEE.

A fast and accurate interest points detection algorithm on 3D meshes using extension of harris operator combined with Hilbert scanning distance

Tibyani, Tibyani; Kamata, Sei Ichiro

ISIEA 2012 - 2012 IEEE Symposium on Industrial Electronics and Applicationsp.367 - 3712012年12月-2012年12月 

DOIScopus

詳細

概要:The main purpose of interest points detection (IPD) for 3D objects based on Harris operator is to find the fast computation of weighted average of the derivative data for different points. In this paper, we analyze the extension of Harris operator using Hausdorff distance (EHOHD) and propose the extension of Harris operator using Hilbert scanning distance (EHOHSD) as a new proposed method to IPD on 3D manifold triangular meshes data. Proposed EHOHSD method is 6-16 times faster than EHOHD method. The quality of this IPD with EHOHSD was measured using the repeatability criterion. © 2012 IEEE.

Registering 3D objects triangular meshes using an interest point detection on barycentric coordinates

Tibyani, Tibyani; Kamata, Sei Ichiro

2012 International Conference on Informatics, Electronics and Vision, ICIEV 2012p.122 - 1272012年11月-2012年11月 

DOIScopus

詳細

概要:In this paper,we put forward an interest point detection framework and combine with a spin image algorithm to register them. A framework is presented in this study. This method make use of the Harris Operator Extension method of interest point detection on 3D manifold triangular meshes in barycentric coordinate. Using this approach, we can extract the object correctly and effectively in noise situation. The unique advantage of this framework is its applicability to triangular meshes models. Experimental results on a different number of models are shown to demonstrate more accurate and effectively results for global registering 3D Objects triangular meshes for three pairs of corresponding interest point features. © 2012 IEEE.

Individual authentication through hand posture recognition using Multi-Hilbert Scanning Distance

Ryu, Jegoon; Kamata, Sei Ichiro

European Signal Processing Conferencep.1787 - 17902012年11月-2012年11月 

Scopus

詳細

ISSN:22195491

概要:In this paper, we propose a novel Hand Posture Recognition (HPR) for biometrics. This study uses the three dimensional point clouds for robust hand posture recognition at the rotation and scale. Multi-Hilbert Scanning Distance (MHSD) are also introduced for mathematical approaches of shape matching. HPR framework is divided into five parts: detecting hand region, removing the wrist, aligning the hand pose, extracting feature descriptor, and matching. Based on the experimental results, this framework showed superior results for hand posture recognition rate. © 2012 EURASIP.

Removal of transparent plastic film specular reflection based on multi-light sources

Wang, Chengjie; Kamata, Sei Ichiro

2012 Symposium on Photonics and Optoelectronics, SOPO 20122012年10月-2012年10月 

DOIScopus

詳細

概要:We present a novel method to remove the specular reflections on the surface of transparent plastic film. Our approach uses four light sources with strategic positions to get four images. Based on the information that both reflection and shadow move a lot from image to image, we reconstruct a high quality image free from reflection and shadow by using a image set which is consist of four images. © 2012 IEEE.

Fast Image Filtering by DCT-Based Kernel Decomposition and Sequential Sum Update

Kenjiro SUGIMOTO and Sei-ichiro KAMATA

Proc. of IEEE Int. Conf. on Image Processingp.125 - 1282012年09月-

A Histogram Separation and Mapping Framework for Image Contrast Enhancement

Qieshi Zhang, Sei-ichiro Kamata

IPSJ Transactions on Computer Vision and Applications4p.100 - 1072012年09月-

SSM-HPC: Front View Gait Recognition Using Spherical Space Model with Human Point Clouds

Jegoon RYU, Sei-ichiro KAMATA, Alireza AHRARY

IEICE Transactions on Information and SystemsE95-D(7)p.1969 - 19782012年07月-

A simple and effective clustering algorithm for multispectral images using space-filling curves

Zhang, Jian; Kamata, Sei Ichiro

IEICE Transactions on Information and SystemsE95-D(7)p.1749 - 17572012年07月-2012年07月 

DOIScopus

詳細

ISSN:09168532

概要:With the wide usage of multispectral images, a fast efficient multidimensional clustering method becomes not only meaningful but also necessary. In general, to speed up the multidimensional images' analysis, a multidimensional feature vector should be transformed into a lower dimensional space. The Hilbert curve is a continuous one-to-one mapping from N-dimensional space to one-dimensional space, and can preserves neighborhood as much as possible. However, because the Hilbert curve is generated by a recurve division process, 'Boundary Effects' will happen, which means data that are close in N-dimensional space may not be close in one-dimensional Hilbert order. In this paper, a new efficient approach based on the space-filling curves is proposed for classifying multispectral satellite images. In order to remove 'Boundary Effects' of the Hilbert curve, multiple Hilbert curves, z curves, and the Pseudo-Hilbert curve are used jointly. The proposed method extracts category clusters from one-dimensional data without computing any distance in N-dimensional space. Furthermore, multispectral images can be analyzed hierarchically from coarse data distribution to fine data distribution in accordance with different application. The experimental results performed on LANDSAT data have demonstrated that the proposed method is efficient to manage the multispectral images and can be applied easily. © 2012 The Institute of Electronics, Information and Communication Engineers.

SSM-HPC: Front view gait recognition using spherical space model with human point clouds

Ryu, Jegoon; Kamata, Sei Ichiro; Ahrary, Alireza

IEICE Transactions on Information and SystemsE95-D(7)p.1969 - 19782012年07月-2012年07月 

DOIScopus

詳細

ISSN:09168532

概要:In this paper, we propose a novel gait recognition framework - Spherical Space Model with Human Point Clouds (SSM-HPC) to recognize front view of human gait. A new gait representation - Marching in Place (MIP) gait is also introduced which preserves the spatiotemporal characteristics of individual gait manner. In comparison with the previous studies on gait recognition which usually use human silhouette images from image sequences, this research applies three dimensional (3D) point clouds data of human body obtained from stereo camera. The proposed framework exhibits gait recognition rates superior to those of other gait recognition methods. © 2012 The Institute of Electronics, Information and Communication Engineers.

Visibility restoration from single image based optical model

Zhang, Qieshi; Zhang, Qieshi; Kamata, Sei Ichiro

VISAPP 2012 - Proceedings of the International Conference on Computer Vision Theory and Applications1p.213 - 2162012年06月-2012年06月 

Scopus

詳細

概要:In this paper, we propose a segmentation based method to estimate the haze-free image by the optical model. In this work, we estimate the atmospheric light by color barycenter hexagon (CBH) model and use the watershed to segment the image to calculate transmission map by dark pixels with single image. Firstly, non-color region is segmented by CBH model and calculate the atmospheric light. Then, use the watershed with rang component of CBH model to segment the color image into several sub-regions, and estimate the transmission map. Finally, use the optical model with the parameters to restore the haze-free image. The experimental results show that our method is more effective and able to get better results than other compared single image based methods.

Efficiently Finding Individuals from Video Dataset

Pengyi HAO, Sei-ichiro KAMATA

IEICE Transactions on Information and SystemsE95-D(5)p.1280 - 12872012年05月-

A Linear Manifold Color Descriptor for Medicine Package Recognition

Kenjiro SUGIMOTO, Koji INOUE, Yoshimitsu KUROKI, Sei-ichiro KAMATA

IEICE Transactions on Information and SystemsE95-D(5)p.1264 - 12712012年05月-

Image Description with Local Patterns: An Application to Face Recognition

Wei ZHOU, Alireza AHRARY, Sei-ichiro KAMATA

IEICE Transactions on Information and SystemsE95-D(5)p.1494 - 15052012年05月-

Novel Algorithm for Polar and Spherical Fourier Analysis on Two and Three Dimensional Images

Zhuo YANG, Sei-ichiro KAMATA

IEICE Transactions on Information and SystemsE95-D(5)p.1248 - 12552012年05月-

Image description with local patterns: An application to face recognition

Zhou, Wei; Ahrary, Alireza; Kamata, Sei Ichiro

IEICE Transactions on Information and SystemsE95-D(5)p.1494 - 15052012年05月-2012年05月 

DOIScopus

詳細

ISSN:09168532

概要:In this paper, we propose a novel approach for presenting the local features of digital image using 1D Local Patterns by Multi-Scans (1DLPMS). We also consider the extentions and simplifications of the proposed approach into facial images analysis. The proposed approach consists of three steps. At the first step, the gray values of pixels in image are represented as a vector giving the local neighborhood intensity distrubutions of the pixels. Then, multi-scans are applied to capture different spatial information on the image with advantage of less computation than other traditional ways, such as Local Binary Patterns (LBP). The second step is encoding the local features based on different encoding rules using 1D local patterns. This transformation is expected to be less sensitive to illumination variations besides preserving the appearance of images embedded in the original gray scale. At the final step, Grouped 1D Local Patterns by Multi-Scans (G1DLPMS) is applied to make the proposed approach computationally simpler and easy to extend. Next, we further formulate boosted algorithm to extract the most discriminant local features. The evaluated results demonstrate that the proposed approach outperforms the conventional approaches in terms of accuracy in applications of face recognition, gender estimation and facial expression. Copyright © 2012 The Institute of Electronics, Information and Communication Engineers.

Novel algorithm for polar and spherical fourier analysis on two and three dimensional images

Yang, Zhuo; Kamata, Sei Ichiro

IEICE Transactions on Information and SystemsE95-D(5)p.1248 - 12552012年05月-2012年05月 

DOIScopus

詳細

ISSN:09168532

概要:Polar and Spherical Fourier analysis can be used to extract rotation invariant features for image retrieval and pattern recognition tasks. They are demonstrated to show superiorities comparing with other methods on describing rotation invariant features of two and three dimensional images. Based on mathematical properties of trigonometric functions and associated Legendre polynomials, fast algorithms are proposed for multimedia applications like real time systems and large multimedia databases in order to increase the computation speed. The symmetric points are computed simultaneously. Inspired by relative prime number theory, systematic analysis are given in this paper. Novel algorithm is deduced that provide even faster speed. Proposed method are 9-15% faster than previous work. The experimental results on two and three dimensional images are given to illustrate the effectiveness of the proposed method. Multimedia signal processing applications that need real time polar and spherical Fourier analysis can be benefit from this work. Copyright © 2012 The Institute of Electronics, Information and Communication Engineers.

A linear manifold color descriptor for medicine package recognition

Sugimoto, Kenjiro; Sugimoto, Kenjiro; Inoue, Koji; Kuroki, Yoshimitsu; Kamata, Sei Ichiro

IEICE Transactions on Information and SystemsE95-D(5)p.1264 - 12712012年05月-2012年05月 

DOIScopus

詳細

ISSN:09168532

概要:This paper presents a color-based method for medicine package recognition, called a linear manifold color descriptor (LMCD). It describes a color distribution (a set of color pixels) of a color package image as a linear manifold (an affine subspace) in the color space, and recognizes an anonymous package by linear manifold matching. Mainly due to low dimensionality of color spaces, LMCD can provide more compact description and faster computation than description styles based on histogram and dominant-color. This paper also proposes distance-based dissimilarities for linear manifold matching. Specially designed for color distribution matching, the proposed dissimilarities are theoretically appropriate more than J-divergence and canonical angles. Experiments on medicine package recognition validates that LMCD outperforms competitors including MPEG-7 color descriptors in terms of description size, computational cost and recognition rate. Copyright © 2012 The Institute of Electronics, Information and Communication Engineers.

Efficiently finding individuals from video dataset

Hao, Pengyi; Kamata, Sei Ichiro

IEICE Transactions on Information and SystemsE95-D(5)p.1280 - 12872012年05月-2012年05月 

DOIScopus

詳細

ISSN:09168532

概要:We are interested in retrieving video shots or videos containing particular people from a video dataset. Owing to the large variations in pose, illumination conditions, occlusions, hairstyles and facial expressions, face tracks have recently been researched in the fields of face recognition, face retrieval and name labeling from videos. However, when the number of face tracks is very large, conventional methods, which match all or some pairs of faces in face tracks, will not be effective. Therefore, in this paper, an efficient method for finding a given person from a video dataset is presented. In our study, in according to performing research on face tracks in a single video, we also consider how to organize all the faces in videos in a dataset and how to improve the search quality in the query process. Different videos may include the same person; thus, the management of individuals in different videos will be useful for their retrieval. The proposed method includes the following three points. (i) Face tracks of the same person appearing for a period in each video are first connected on the basis of scene information with a time constriction, then all the people in one video are organized by a proposed hierarchical clustering method. (ii) After obtaining the organizational structure of all the people in one video, the people are organized into an upper layer by affinity propagation. (iii) Finally, in the process of querying, a remeasuring method based on the index structure of videos is performed to improve the retrieval accuracy. We also build a video dataset that contains six types of videos: films, TV shows, educational videos, interviews, press conferences and domestic activities. The formation of face tracks in the six types of videos is first researched, then experiments are performed on this video dataset containing more than 1 million faces and 218,786 face tracks. The results show that the proposed approach has high search quality and a short search time. Copyright © 2012 The Institute of Electronics, Information and Communication Engineers.

Fast Hypercomplex Polar Fourier Analysis

Zhuo YANG, Sei-ichiro KAMATA

IEICE Transactions on Information and SystemsE95-D(4)p.1166 - 11692012年04月-

Fast hypercomplex polar fourier analysis

Yang, Zhuo; Kamata, Sci Ichiro

IEICE Transactions on Information and SystemsE95-D(4)p.1166 - 11692012年04月-2012年04月 

DOIScopus

詳細

ISSN:09168532

概要:Hypercomplex polar Fourier analysis treats a signal as a vector field and generalizes the conventional polar Fourier analysis. It can handle signals represented by hypercomplex numbers such as color images. Hypercomplex polar Fourier analysis is reversible that means it can reconstruct image. Its coefficient has rotation invariance property that can be used for feature extraction. However in order to increase the computation speed, fast algorithm is needed especially for image processing applications like realtime systems and limited resource platforms. This paper presents fast hypercomplex polar Fourier analysis based on symmetric properties and mathematical proper ies of trigonometric functions. Proposed fast hy percomplex polar Fourier analysis computes symmetric points simultane ously, which significantly reduce the computation time. Copyright © 2012 The Institute of Electronics, Information and Communication Engineers.

A generalized 3-D Hilbert scan using look-up tables

Zhang, Jian; Kamata, Sei Ichiro

Journal of Visual Communication and Image Representation23(3)p.418 - 4252012年04月-2012年04月 

DOIScopus

詳細

ISSN:10473203

概要:The Hilbert curve is a one-to-one mapping between multidimensional space and one-dimensional (1-D) space. Due to the advantage of preserving high correlation of multidimensional points, it receives much attention in many areas. Especially in image processing, Hilbert curve is studied actively as a scan technique (Hilbert scan). Currently there have been several Hilbert scan algorithms, but they usually have strict implementation conditions. For example, they use recursive functions to generate scans, which makes the algorithms complex and difficult to implement in real-time systems. Moreover the length of each side in a scanned region should be same and equal to the power of two, which limits the application of Hilbert scan greatly. In this paper, to remove the constraints and improve the Hilbert scan for a general application, an effective generalized three-dimensional (3-D) Hilbert scan algorithm is proposed. The proposed algorithm uses two simple look-up tables instead of recursive functions to generate a scan, which greatly reduces the computational complexity and saves storage memory. Furthermore, the experimental results show that the proposed generalized Hilbert scan can also take advantage of the high correlation between neighboring lattice points in an arbitrarily-sized cuboid region, and give competitive performance in comparison with some common scan techniques. © 2011 Elsevier Inc. All rights reserved.

An improved method for illumination invariant face recognition based on adaptive rescaling DCT coefficient in logarithm domain

Yu, Chao; Zhao, Xiaoqun; Kamata, Sei Ichiro

Lecture Notes in Electrical Engineering121 LNEEp.297 - 3042011年12月-2011年12月 

DOIScopus

詳細

ISSN:18761100

概要:This paper presents an improved method for robust face recognition using illumination normalization based on Discrete Cosine Transform (DCT) in logarithm domain. Two novel coefficients are designed to identify the lighting condition (LC), based on which the low-frequency DCT coefficients are adaptively rescaled except the first one (DC). As a result variations under different illumination conditions are minimized meanwhile original information contained in low-frequency is comparatively well preserved. Results of experiments on Yale B database and Extended Yale B database show that proposed method has better performance under variational input illumination conditions. The proposed method is fast in computation and could be easily implemented into real time face recognition systems. © 2011 Springer-Verlag.

Color distribution matching using a weighted subspace descriptor

Sugimoto, Kenjiro; Kamata, Sei Ichiro

Proceedings - International Conference on Image Processing, ICIPp.1697 - 17002011年12月-2011年12月 

DOIScopus

詳細

ISSN:15224880

概要:This paper presents a low-level color descriptor which describes the color distribution of a color image as a weighted subspace in the color space, namely eigenvectors and eigenvalues of the distribution. Thanks to low-dimensionality of color space, the proposed descriptor can provide compact description and fast computation. Furthermore, specialized for color distribution matching, it is more efficient than mutual subspace method (MSM). Experiments on medicine package recognition validate that the proposed descriptor outperforms MSM and MPEG-7 low-level color descriptors in terms of description size, computational cost and recognition rate. © 2011 IEEE.

Front view gait recognition using spherical space model with human point clouds

Ryu, Jegoon; Kamata, Sei Ichiro

Proceedings - International Conference on Image Processing, ICIPp.3209 - 32122011年12月-2011年12月 

DOIScopus

詳細

ISSN:15224880

概要:In this paper, we propose a novel gait recognition framework which is Spherical Space Model with Human Point Clouds (SSM-HPC). A new gait representation is also introduced, which is called Marching in Place (MIP) gait and preserves the spatiotemporal characteristics of individual gait manner. Various researches for gait recognition have used human silhouette images from moving picture. This research uses Three Dimensional (3D) point clouds data of human body obtained from stereo camera, which has the scale-invariant property. The framework is applied for frontal view gait recognition. This framework showed superior results for gait recognition rate than other gait recognition methods. © 2011 IEEE.

A common key encryption algorithm using N-dimensional Hilbert curves

Kamata, Sei Ichiro

Proceedings of the 2011 7th International Conference on Information Assurance and Security, IAS 2011p.275 - 2792011年12月-2011年12月 

DOIScopus

詳細

概要:There are a lot of previous works on common key encryptions such as DES, AES, etc, In this paper, a new common key encryption algorithm is proposed using Hilbert curves which are a one-to-one mapping between N-dimensional (N-D) spaces and 1-D space (a line). This is based on a property having a sharp rise in the number of Hilbert curve patterns in N-D spaces. In the case of N = 2, there are only four patterns, while if N is 5, the number of the patterns is more than 1 billions. Operations of addition and multiplication are denned on a curve, based on a mapping of a point in N-D spaces to a point on a line. In order to realize a cryptosystem, the algorithm utilizes Hilbert ordered point addresses, which is expressed as the coordinates of the points in N-dimensional space. © 2011 IEEE.

Multi balanced trees for face retrieval from image database

Hao, Pengyi; Kamata, Sei Ichiro

2011 IEEE International Conference on Signal and Image Processing Applications, ICSIPA 2011p.484 - 4892011年12月-2011年12月 

DOIScopus

詳細

概要:We are interested here in retrieving images containing a specific person in image database. Due to large variations in illumination conditions, hairstyles, facial expressions, etc. and the factors like occlusion, sunglasses, profile, etc., robust face matching has been a challenging problem. On the other hand, the speed of search is also a considerable issue, especially for the dataset with millions of face images. Inspired by face tracks in video retrieval which take advantages from the abundance of frames to get multiple exemplars, we present an approach named multi balanced trees for face retrieval from image dataset in this paper. Face images in the dataset are efficiently organized by the trees produced for persons. Multi sampling on the facial components employs the rich local information, which can help to differentiate different persons. Given a query face, a sorted face set with similarities is obtained by inserting the query into a tree. It is easy and fast to get the search results in respect that it avoids calculating the distances between query and elements in the cluster. In addition, a rectification strategy is given in the query process to rectify the error occurred in the generation of trees, resulting in a significant improvement of retrieval quality. Experimental results show the better face grouping ability in comparison with traditional methods. The speed of searching is improved as well. © 2011 IEEE.

Anisotropic diffusion with edge projection

Qiao, Gang; Qiao, Gang; Zhang, Wei; Kamata, Sei Ichiro

Proceedings - 4th International Congress on Image and Signal Processing, CISP 20112p.665 - 6672011年12月-2011年12月 

DOIScopus

詳細

概要:In image processing, anisotropic diffusion provides a forward method to remove noise while preserving edges accurate and sharp. However, due to the inappropriate edge estimation by gradient, some isolated noise points still exist and edge location is inaccurate. In this representation, isolated noise points and edges are distinguished by the significant difference of their "lengths", which are computed by orthogonally projecting their pixels to the corresponding normalized gradient directions and recording the number of the same projections. Combining gradient and "length" to estimate edges, isolated noise points are further suppressed while edges are re-located and enhanced. © 2011 IEEE.

Hypercomplex polar Fourier analysis for color image

Yang, Zhuo; Kamata, Sei Ichiro

Proceedings - International Conference on Image Processing, ICIPp.2117 - 21202011年12月-2011年12月 

DOIScopus

詳細

ISSN:15224880

概要:Fourier transform is a significant tool in image processing and pattern recognition. By introducing hypercomplex number, hypercomplex Fourier transform [1] treats signal as vector field and generalizes conventional Fourier transform. Inspired from that, hypercomplex polar Fourier analysis is proposed in this paper. This work extends conventional polar Fourier analysis [5]. The proposed method can handle hypercomplex number represented signals like color image. The hypercom-plex polar Fourier analysis is reversible that means it can be used to reconstruct image. The hypercomplex polar Fourier descriptor has rotation invariance property that can be used for feature extraction. Due to the noncommutative property of quaternion multiplication, both left-side and right-side hypercomplex polar Fourier analysis are discussed and their relationships are also established in this paper. The experimental results on image reconstruction, rotation invariance and color plate test are given to illustrate the usefulness of the proposed method as an image analysis tool. © 2011 IEEE.

Fast color matching using weighted subspace on medicine package recognition

Sugimoto, Kenjiro; Kamata, Sei Ichiro

Proceedings of the 12th IAPR Conference on Machine Vision Applications, MVA 2011p.287 - 2902011年12月-2011年12月 

Scopus

詳細

概要:This paper presents a color matching technique using weighted subspace on medicine package recognition. The proposed method is more compact and lowercomplex than scalable color descriptor and dominant color descriptor, which are employed by MPEG-7. Our method is based on subspace matching: A color object is treated as a subspace derived from its color distribution. Unlike mutual subspace method, it is specially designed for color matching. Specifically, weighted subspace and a distance-based dissimilarity are employed instead of normalized subspace and similarity based on canonical angles of MSM. Experiments show that the proposed method outperforms the conventional methods in terms of description size, building/matching speed, and recognition rate.

Fast Polar Cosine Transform for image description

Yang, Zhuo; Kamata, Sei Ichiro

Proceedings of the 12th IAPR Conference on Machine Vision Applications, MVA 2011p.320 - 3232011年12月-2011年12月 

Scopus

詳細

概要:Polar Cosine Transform (PCT) is one of the Polar Harmonic Transforms that those kernels are basic waves and harmonic in nature. They are proposed to represent invariant patterns for two dimensional image description and are demonstrated to show superiorities comparing with other methods on extracting rotation invariant patterns for images. However in order to increase the computation speed, fast algorithm for PCT is proposed for real world applications like limited computing environments, large image databases and realtime systems. Based on our previous work, this paper novelly employs relative prime number theory to develop Fast Polar Cosine Transform (FPCT). The proposed FPCT is averagely over 11 ∼ 12.5 times faster than PCT that significantly boost computation process. The experimental results are given to illustrate the effectiveness of the proposed method.

Fast Hypercomplex Polar Fourier Analysis for Image Processing

Zhuo YANG and Seiichiro KAMATA

Proc. of The Fifth Pacific-Rim Symposium on Image and Video TechnologyLNCS 7088p.141 - 1482011年11月-

An Improved Method for Illumination Invariant Face Recognition Base on Adaptive Rescaling DCT Coefficient in Logarithm Domain

Chao YU, Xiaoqqun ZHAO and Sei-ichiro KAMATA

Proc. of the 2011 International Symposium on Computer, Communication, Control and Automation121p.297 - 3042011年11月-

Multi Balanced Trees for Face Retrieval From Image Database

Pengyi HAO, Seiichiro KAMATA

Proc. of Int. Conf. on Signal and Image Processing Applicationsp.478 - 4832011年11月-

Manifold learning based on multi-feature for road-sign recognition

Zhang, Qieshi; Zhang, Qieshi; Kamata, Sei Ichiro

Proceedings of the SICE Annual Conferencep.1143 - 11462011年11月-2011年11月 

Scopus

詳細

概要:In this paper, a multi-feature selection and application based manifold learning metric method is proposed for Road-Sign Recognition (RSR). Firstly, the manifold metric between manifold from subspace is discussed in detail. After that, the multi-feature analyzing, selection, classification and application are introduced for rough recognition and create the manifold. Then the proposed method is used to evaluate the distance between the manifolds. Finally, the RSR results suggest that the proposed method is robust than other methods. © 2011 SICE.

Fast hypercomplex polar Fourier analysis for image processing

Yang, Zhuo; Kamata, Sei Ichiro

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)7088 LNCS(PART 2)p.141 - 1482011年11月-2011年11月 

DOIScopus

詳細

ISSN:03029743

概要:Hypercomplex polar Fourier analysis treats a signal as a vector field and generalizes the conventional polar Fourier analysis. It can handle signals represented by hypercomplex numbers such as color images. It is reversible that can reconstruct image. Its coefficient has rotation invariance property that can be used for feature extraction. With these properties, it can be used for image processing applications like image representation and image understanding. However in order to increase the computation speed, fast algorithm is needed especially for image processing applications like realtime systems and limited resource platforms. This paper presents fast hypercomplex polar Fourier analysis that based on symmetric properties and mathematical properties of trigonometric functions. Proposed fast hypercomplex polar Fourier analysis computes symmetric eight points simultaneously that significantly reduce the computation time. © 2011 Springer-Verlag.

Anisotropic Diffusion with Edge Projection

Gang QIAO, Wei ZHANG and Sei-ichiro KAMATA

Proc. of 2011 International Congress on Image and Signal Processingp.680 - 6822011年10月-

Single Image Based Haze Removal Method

Qieshi ZHANG and Sei-ichiro KAMATA

Proc. of the 4th International Conference on Computer and Electrical Engineeringp.365 - 3692011年10月-

Color Distribution Matching Using A Weighted Subspace Descriptor

Kenjiro SUGIMOTO and Seiichiro KAMATA

Proc. of IEEE International Conference on Image Processingp.1733 - 17362011年09月-

Front View Gait Recognition Using Spherical Space Model With Human Point Clouds

Jegoon RYU and Seiichiro KAMATA

Proc. of IEEE International Conference on Image Processingp.3270 - 32732011年09月-

Hypercomplex Polar Fourier Analysis for Color Image

Zhuo YANG and Seiichiro KAMATA

Proc. of IEEE International Conference on Image Processingp.2161 - 21642011年09月-

Hypercomplex Polar Fourier Analysis for Image Representation

Zhuo YANG and Seiichiro KAMATA

IEICE Transactions on Information and SystemsE94-D(8)p.1663 - 16702011年08月-

Hypercomplex polar fourier analysis for image representation

Yang, Zhuo; Kamata, Sei Ichiro

IEICE Transactions on Information and SystemsE94-D(8)p.1663 - 16702011年08月-2011年08月 

DOIScopus

詳細

ISSN:09168532

概要:Fourier transform is a significant tool in image processing and pattern recognition. By introducing a hypercomplex number, hypercomplex Fourier transform treats a signal as a vector field and generalizes the conventional Fourier transform. Inspired from that, hypercomplex polar Fourier analysis that extends conventional polar Fourier analysis is proposed in this paper. The proposed method can handle signals represented by hypercomplex numbers as color images. The hypercomplex polar Fourier analysis is reversible that means it can be used to reconstruct image. The hypercomplex polar Fourier descriptor has rotation invariance property that can be used for feature extraction. Due to the noncommutative property of quaternion multiplication, both left-side and right-side hypercomplex polar Fourier analysis are discussed and their relationships are also established in this paper. The experimental results on image reconstruction, rotation invariance, color plate test and image retrieval are given to illustrate the usefulness of the proposed method as an image analysis tool. Copyright © 2011 The Institute of Electronics, Information and Communication Engineers.

Hilbert Scan Based Bag-of-features for Image Retrieval

Pengyi HAO and Seiichiro KAMATA

IEICE Transactions on Information and SystemsE94-D(6)p.1260 - 12682011年06月-

Fast Polar Cosine Transform for Image Description

Zhuo YANG and Sei-ichiro KAMATA

Proc. of 12th. IAPR Conference on Machine Vision Applications (MVA2011)p.320 - 3232011年06月-

Color Matching Using Weighted Subspace

Kenjiro SUGIMOTO and Sei-ichiro KAMATA

Proc. of 12th. IAPR Conference on Machine Vision Applications (MVA2011)p.287 - 2902011年06月-

Hilbert scan based bag-of-features for image retrieval

Hao, Pengyi; Kamata, Sei Ichiro

IEICE Transactions on Information and SystemsE94-D(6)p.1260 - 12682011年06月-2011年06月 

DOIScopus

詳細

ISSN:09168532

概要:Generally, two problems of bag-of-features in image retrieval are still considered unsolved: one is that spatial information about descriptors is not employed well, which affects the accuracy of retrieval; the other is that the trade-off between vocabulary size and good precision, which decides the storage and retrieval performance. In this paper, we propose a novel approach called Hilbert scan based bag-of-features (HS-BoF) for image retrieval. Firstly, Hilbert scan based tree representation (HSBT) is studied, which is built based on the local descriptors while spatial relationships are added into the nodes by a novel grouping rule, resulting of a tree structure for each image. Further, we give two ways of codebook production based on HSBT: multi-layer codebook and multi-size codebook. Owing to the properties of Hilbert scanning and the merits of our grouping method, sub-regions of the tree are not only flexible to the distribution of local patches but also have hierarchical relations. Extensive experiments on caltech-256, 13-scene and 1 million ImageNet images show that HS-BoF obtains higher accuracy with less memory usage. Copyright © 2011 The Institute of Electronics, Information and Communication Engineers.

A Fast Homology Search Algorithm Using Dynamic Seeding

Haijiang TANG, Sei-ichiro KAMATA, Toshimasa YAMAZAKI

Proc. of The 21th International Conference on Genome Informatics2010年12月-

Fast Polar Harmonic Transforms

Zhuo YANG, Sei-ichiro KAMATA

Proc. of 11th. Int. Conf. Control, Automation, Robotics and Vision (ICARCV2010)p.673 - 6772010年12月-

Image description with 1D local patterns by multi-scans: An application to face recognition

Zhou, Wei; Ahrary, Alireza; Kamata, Sei Ichiro

Proceedings - International Conference on Image Processing, ICIPp.4553 - 45562010年12月-2010年12月 

DOIScopus

詳細

ISSN:15224880

概要:In this work, a novel approach which called 1D local patterns by multi-scans(1DLPMS) for presenting the local features is proposed and its simplifications and extensions to facial image analysis are also considered. First, multi-scans are applied to capture different spatial information on the image with less computation than some traditional ways, such as Local Binary Patterns(LBP). Then, some 1D local patterns are given to encode the local features based on different coding rules. To make the proposed approach computationally simpler and easy to extend, grouped 1D local patterns by multiscans( G1DLPMS) is studied, which divides 1DLPMS into several groups and uses the co-occurrences of these groups. Performance assessment in face recognition under different challenges shows that the proposed approach is superior than traditional ones. © 2010 IEEE.

3D object matching based on spherical Hilbert scanning

Tong, Can; Kamata, Sei Ichiro

Proceedings - International Conference on Image Processing, ICIPp.2941 - 29442010年12月-2010年12月 

DOIScopus

詳細

ISSN:15224880

概要:This paper describes a novel method to match objects in cluttered scenes. This method makes use of Hilbert scanning of feature points in Hough space. We use a 3D Hough transform to obtain a spectrum on which 3D features are concentrated on the sphere. Then, based on the obtained Hough Spectrum, we apply Hilbert scanning on the sphere to match the objects. Using this approach, we can match the object correctly and robustly in both overlapping and noise situation. The characteristic of this method is that it is a global matching method without an estimate of the rotation first and suffering from computational complexity brought by voting/correlation procedure. The experiment results show that the method is more effective compared to existing methods in both matching rate and robustness. © 2010 IEEE.

Adaptive histogram analysis for image enhancement

Zhang, Qieshi; Inaba, Hiroshi; Kamata, Sei Ichiro

Proceedings - 4th Pacific-Rim Symposium on Image and Video Technology, PSIVT 2010p.408 - 4132010年12月-2010年12月 

DOIScopus

詳細

概要:One image processing application is to reconstruct the original scene from the low quality images. Considering the idea histogram distribution can reflect good vision effect. So many histogram analyzing based methods have been studied recently. However, some methods require users to set some parameters or condition, and cannot get the optimal results automatically. To overcome those short come, this paper presents an Adaptive Histogram Separation and Mapping (AHSM) method for Backlight image enhancement. First, we separate the histogram by binary tree structure with the proposed Adaptive Histogram Separation Unit (AHSU). And then mapping the Low Dynamic Range (LDR) histogram partition into High Dynamic Range (HDR). By doing this, the excessive or scarcity enhancement can be avoid. The experimental results show that the proposed method can gives better enhancement results, also compared with some histogram analyzing based methods and get better results. © 2010 IEEE.

Face detection in color images based on skin color models

Zou, Li; Kamata, Sei Ichiro

IEEE Region 10 Annual International Conference, Proceedings/TENCONp.681 - 6862010年12月-2010年12月 

DOIScopus

詳細

概要:Face finding is a very important initial step towards building up a fully automated face recognition system. Face detection by detecting skin like colors can achieve a high detection rate. In this paper, we presented a novel algorithm for face detection in color images with complex backgrounds. First a parallel structure for skin color detection is proposed to improve the accuracy of detections. The concept of the probability image has been introduced to utilize the color information in the traditional face detection methods specific for gray-scale images. After that, a classifier obtained from Adaboost training is applied to the result of skin detection to reduce the false positives. An experiment has been implemented to verify the improvement of this proposed research. And the proposed approach achieved a better result in this experiment. © 2010 IEEE.

Hilbert scan based tree representation for image search

Hao, Pengyi; Kamata, Sei Ichiro

IEEE Region 10 Annual International Conference, Proceedings/TENCONp.499 - 5042010年12月-2010年12月 

DOIScopus

詳細

概要:In this paper, Hilbert scan based tree representation (HSBT) is presented for image search. Unlike common ways decreasing the number of interest points or reducing the dimensions of features or using searching methods to match interest points, the proposed method builds a tree for each image and gives a new distance measure to calculate the similarity between the query and images in database. In the proposed approach, Hilbert scan for arbitrarily-sized arrays is used to map the interest points from two-dimensional space to one-dimensional space at first. Then, interest points set is divided into several parts by a separation way, and a grouping strategy is given to build a tree for each image. Experimental results show that the proposed approach is space saving. That is because it only stores clustering center and relevant information of each node in the tree. It is also time saving since the similarity calculation is up to the nodes of tree rather than all the descriptors of image. At the same time, the retrieval precision is good, because Hilbert scanning preserves the correlation in two-dimensional image, so nodes of tree are shaped according to the compactness of interest points which can employ the local information as much as possible. © 2010 IEEE.

Document layout analysis and reading order determination for a reading robot

Pan, Yucun; Zhao, Qunfei; Kamata, Seiichiro

IEEE Region 10 Annual International Conference, Proceedings/TENCONp.1607 - 16122010年12月-2010年12月 

DOIScopus

詳細

概要:In this paper an efficient approach of document layout analysis and reading order determination is proposed for a reading robot. Firstly the input document images are preprocessed to remove noises, connect lines and domains, and to reduce the computation time. Secondly a bottom-up, parameter-independent, two-step layout analysis algorithm based on morphology is used, which outlines the geometry of the maximum homogeneous regions and classifies them into texts, tables, and pictures. Finally the reading order is determined, by a top-down recursive hierarchy algorithm derived from XY-cut, using a set of rules depending on layout information. Important parameters are acquired using statistic information of the given images to adapt to different types of documents. The proposed algorithm is applied to a large number of document images and the experimental results show that it makes the reading robot be able to read paper documents of different languages, even with complex layout structure. ©2010 IEEE.

3D reconstruction from a single image for a Chinese talking face

Liu, Ning; Fang, Ning; Kamata, Seiichiro

IEEE Region 10 Annual International Conference, Proceedings/TENCONp.1613 - 16162010年12月-2010年12月 

DOIScopus

詳細

概要:This paper proposes an automatic 3D reconstruction approach for a Chinese talking face by a generic model and a single image. Firstly, an improved color-based ASM method is used to detect the face area and get the 2D face feature points automatically from the given image, which is not restricted to full frontal one. Then, color information is used to correct the location of face feature points. Finally, after text mapping, a particular and realistic 3D face model is deformed from a generic model. Using ASM face feature points extraction and correction based on skin color model, the problem of side face information missing is successfully resolved. Depending on only one image and one generic model, the computing cost of memory and time is largely reduced. The 3D face reconstructed can be easily deformed to form different expressions and mouth shapes. Experiments show that this approach is fast and efficient and has an output of a lifelike Chinese talking face. ©2010 IEEE.

Face recognition with local gradient derivative patterns

Zheng, Xianchun; Kamata, Sei Ichiro; Yu, Liang

IEEE Region 10 Annual International Conference, Proceedings/TENCONp.667 - 6702010年12月-2010年12月 

DOIScopus

詳細

概要:In this work, we present a novel local pattern descriptor, Local Gradient Derivative Pattern (LGDP) to face recognition which considers more detailed information than the Local Binary Pattern (LBP). The face image is first divided into several small regions from which Local Gradient Derivative Pattern (LGDP) histograms are extracted and concatenated into a single, spatially enhanced feature vector to be used as a face descriptor. Three well-known and challenge-ORL, Yale and FERET face databases are used in the performances to evaluate the method. The experiments result clearly show that the proposed method give us a better performance than some other methods. © 2010 IEEE.

A novel face representation toward pose invariant face recognition

Yu, Liang; Yu, Liang; Kamata, Sei Ichiro; Fang, Yong

IEEE Region 10 Annual International Conference, Proceedings/TENCONp.179 - 1832010年12月-2010年12月 

DOIScopus

詳細

概要:Though elastic bunch graph matching (EBGM) has a good performance on face recognition in the distortion of facial expression, it is still not robust enough to in-depth rotation. To solve this problem, a novel face representation approach based on the space-filling tree is proposed in this paper. This kind of representation shows a better performance than Elastic bunch graph matching (EBGM) in in-depth rotation of pose especially when there are only frontal images in the training set. With the proposed face representation approach, the face recognition system is built. Experimental results on the FERET standard database show that the proposed face representation approach is more effective and robust to the in-depth rotation of pose when there are only frontal images in the training set. ©2010 IEEE.

Fast polar harmonic transforms

Yang, Zhuo; Kamata, Sei Ichiro

11th International Conference on Control, Automation, Robotics and Vision, ICARCV 2010p.673 - 6772010年12月-2010年12月 

DOIScopus

詳細

概要:Polar Harmonic Transform (PHT) is termed to represent a set of transforms those kernels are basic waves and harmonic in nature. PHTs consist of Polar Complex Exponential Transform (PCET), Polar Cosine Transform (PCT) and Polar Sine Transform (PST). They are proposed to represent invariant image patterns for two dimensional image retrieval and pattern recognition tasks. They are demonstrated to show superiorities comparing with other methods on describing rotation invariant patterns for images. Kernel computation of PHTs is also simple and has no numerical stability issue. However in order to increase the computation speed, fast computation method is needed especially for real world applications like limited computing environments, large image databases and realtime systems. This paper presents Fast Polar Harmonic Transforms (FPHTs) including Fast Polar Complex Exponential Transform (FPCET), Fast Polar Cosine Transform (FPCT) and Fast Polar Sine Transform (FPST) that are deduced based on mathematical properties of trigonometric functions. The proposed FPHTs are averagely over 6 ∼ 8 times faster than PHTs that significantly boost computation process. The experimental results on both synthetic and real data are given to illustrate the effectiveness of the proposed fast transforms. © 2010 IEEE.

Hilbert Scan Based Tree Representation for Image Search

Pengyi HAO, Sei-ichiro KAMATA

Proc. of IEEE TENCON2010p.499 - 5042010年11月-

A Novel Face Representation Toward Pose Invariant Face Recognition

Liang YU, Sei-ichiro KAMATA, Yong FANG

Proc. of IEEE TENCON2010p.179 - 1832010年11月-

Face Detection in Color Images Based on Skin Color Models

Li ZOU, Sei-ichiro KAMATA

Proc. of IEEE TENCON2010p.681 - 6862010年11月-

3D Reconstruction from a Single Image for a Chinese Talking Face

Ning Liu, Ning Fang, Seiichiro KAMATA

Proc. of IEEE TENCON2010p.1613 - 16162010年11月-

Document Layout Analysis and Reading Order Determination for a Reading Robot

Yucun PAN, Qunfei ZHAO, Seiichiro KAMATA

Proc. of IEEE TENCON2010p.1607 - 16122010年11月-

A Color Distribution Descriptor for Medicine Package Recognition

Kenjiro Sugimoto, Koji Inoue, Kuroki Yoshimitsu and Sei-ichiro Kamata

Proc. of 2nd China-Japan-Korea Joint Workshop of Pattern Recognition2010年11月-

Fast polar and spherical fourier descriptors for feature extraction

Yang, Zhuo; Kamata, Sei Ichiro

Proceedings - International Conference on Pattern Recognitionp.975 - 9782010年11月-2010年11月 

DOIScopus

詳細

ISSN:10514651

概要:Polar Fourier Descriptor(PFD) and Spherical Fourier Descriptor(SFD) are rotation invariant feature descriptors for two dimensional(2D) and three dimensional(3D) image retrieval and pattern recognition tasks. They are demonstrated to show superiorities compared with other methods on describing rotation invariant features of 2D and 3D images. However in order to increase the computation speed, fast computation method is needed especially for applications like realtime systems and large image databases. This paper presents fast computation method for PFD and SFD that based on mathematical properties of trigonometric functions and associated Legendre polynomials. Proposed fast PFD and SFD are 8 and 16 times faster than traditional ones that significantly boost computation process. © 2010 IEEE.

Image Description With 1D Local Pattern By Multi-Scans: An Application To Face Recognition

Wei Zhou, Alireza AHRARY and Sei-ichiro KAMATA

Proc. of 2010 IEEE Int. Conf. on Image Processingp.4553 - 45562010年09月-

3D Object Matching Based on Spherical HILBERT Scanning

Can TONG and Sei-ichiro KAMATA

Proc. of 2010 IEEE Int. Conf. on Image Processingp.2941 - 29442010年09月-

A Study on Fast Random Access Decompression Using Start-step-stop Coding and Rank/Select Dictionary

Kenjiro SUGIMOTO and Sei-ichiro KAMATA

Proc. of 1st Int. Conf. on Advanced Computing and Communications (ACC-2010)p.132 - 1352010年09月-

Fast Polar and Spherical Fourier Descriptors for Feature Extraction

Zhou YANG and Sei-ichiro KAMATA

Proc. of 20th Int. Conf. on Pattern Recognitionp.975 - 9782010年08月-

Interscale Stein's Unbiased Risk Estimate and Intrascale Feature Patches Distance Constraint for Image Denoising

Qieshi ZHANG, Sei-ichiro KAMATA and Alireza AHRARY

IEICE Transactions on Fundamentals of Electronics, Communications and Computer SciencesE93-A(8)p.1434 - 14412010年08月-

Interscale Stein's unbiased risk estimate and intrascale feature patches distance constraint for image denoising

Zhang, Qieshi; Kamata, Sei Ichiro; Ahrary, Alireza

IEICE Transactions on Fundamentals of Electronics, Communications and Computer SciencesE93-A(8)p.1434 - 14412010年08月-2010年08月 

DOIScopus

詳細

ISSN:09168508

概要:The influence of noise is an important problem on image acquisition and transmission stages. The traditional image denoising approaches only analyzing the pixels of local region with a moving window, which calculated by neighbor pixels to denoise. Recently, this research has been focused on the transform domain and feature space. Compare with the traditional approaches, the global multi-scale analyzing and unchangeable noise distribution is the advantage. Apparently, the estimation based methods can be used in transform domain and get better effect. This paper proposed a new approach to image denoising in orthonormal wavelet domain. In this paper, we adopt Stein's unbiased risk estimate (SURE) based method to denoise the low-frequency bands and the feature patches distance constraint (FPDC) method also be proposed to estimate the noise free bands in Wavelet domain. The key point is that how to divide the lower frequency sub-bands and the higher frequency sub-bands, and do interscale SURE and intrascale FPDC, respectively. We compared our denoising method with some well-known and new denoising algorithms, the experimental results show that the proposed method can give better performance and keep more detail information in most objective and subjective criteria than other methods. Copyright © 2010 The Institute of Electronics,.

Fast Polar and Spherical Fourier Descriptors for Feature Extraction

Zhou YANG and Sei-ichiro KAMATA

IEICE Transactions on Information and SystemsE93-D(7)p.1708 - 17152010年07月-

Fast Polar Harmonic Transforms

Zhou YANG, Alireza AHRARY and Sei-ichiro KAMATA

The Journal of the IIEEJ39(4)p.399 - 4082010年07月-

Stein's unbiased risk estimate (SURE) and distance constraint combined image denoising in Wavelet domain

Zhang, Qieshi; Kamata, Sei Ichiro

Proceedings of the 7th IASTED International Conference on Signal Processing, Pattern Recognition and Applications, SPPRA 2010p.196 - 2012010年07月-2010年07月 

Scopus

詳細

概要:Image denoising is a lively research field now. For solving this problem, non-linear filters based methods are the classical approach. These methods are based on local analysis of pixels with a moving window in spatial domain, but also have some shortcoming. Recently, because of the properties of Wavelet transform, this research has been focused on the wavelet domain. Compared to the classical nonlinear filters, the global multi-scale analysis characteristic of Wavelet is better for image denoising. So this paper proposed a new approach to use orthonormal Wavelet transform and distance constraint to solve this. Here, by minimizing the Stein's unbiased risk estimate (SURE) method to calculate the low frequency sub-band images for estimating. And convert the high frequency sub-band images to feature space, then use distance constraint to denoise by trained samples set. The experimental results show that the proposed method is efficiency and keep the detail ideally.

Fast polar and spherical fourier descriptors for feature extraction

Yang, Zhuo; Kamata, Sei Ichiro

IEICE Transactions on Information and SystemsE93-D(7)p.1708 - 17152010年07月-2010年07月 

DOIScopus

詳細

ISSN:09168532

概要:Polar Fourier Descriptor(PFD) and Spherical Fourier Descriptor( SFD) are rotation invariant feature descriptors for two dimensional( 2D) and three dimensional(3D) image retrieval and pattern recognition tasks. They are demonstrated to show superiorities compared with other methods on describing rotation invariant features of 2D and 3D images. However in order to increase the computation speed, fast computation method is needed especially for machine vision applications like realtime systems, limited computing environments and large image databases. This paper presents fast computation method for PFD and SFD that are deduced based on mathematical properties of trigonometric functions and associated Legendre polynomials. Proposed fast PFD and SFD are 8 and 16 times faster than direct calculation that significantly boost computation process. Furthermore, the proposed methods are also compact for memory requirements for storing PFD and SFD basis in lookup tables. The experimental results on both synthetic and real data are given to illustrate the efficiency of the proposed method. Copyright © 2010 The Institute of Electronics, Information and Communication Engineers.

An Adaptive Tone Mapping Algorithm for High Dynamic Range Images

Jian Zhang and Sei-ichiro Kamata

The Journal of The Institute of Image Information and Television Engineers64(6)p.850 - 8602010年06月-

An adaptive tone mapping algorithm for high dynamic range images

Zhang, Jian; Kamata, Sei Ichiro

Kyokai Joho Imeji Zasshi/Journal of the Institute of Image Information and Television Engineers64(6)p.850 - 8602010年06月-2010年06月 

DOIScopus

詳細

ISSN:13426907

概要:A common task of tone mapping algorithms is to reproduce high dynamic range images (HDR) on low dynamic range (LDR) display devices such as printers and monitors. We present a new tone mapping algorithm for the display of HDR images that was inspired by the adaptive process of the human visual system. The proposed algorithm is based on center/surround Retinex processing. Our method has two novel aspects. The input luminance image is first compressed by a global tone mapping curve. The curvature of the compression curve is adapted locally based on the pseudo-Hilbert scan technique, so it can provide a better overall impression before the subsequent local processing. Second, the local details are enhanced according to a non-linear adaptive spatial filter (Gaussian filter), whose shape (filter variance) is adapted to the high-contrast edges of the image. The proposed method takes advantage of the properties of both global and local processing while overcoming their respective disadvantages. Therefore, the algorithm can preserve visibility and contrast impression of high dynamic range scenes in standard display devices. We tested the proposed method on a variety of HDR images and also compared it to previous research. The results indicated that our method was effective for displaying images with high visual quality.

On-line Signature Matching Based on Hilbert Scanning Patterns

Alireza Ahrary, Jian Zhang and Sei-ichiro Kamata

The Journal of the IIEEJ39(2)p.175 - 1842010年03月-

Pixel color feature enhancement for road signs detection

Zhang, Qieshi; Kamata, Sei Ichiro

Proceedings of SPIE - The International Society for Optical Engineering75462010年03月-2010年03月 

DOIScopus

詳細

ISSN:0277786X

概要:Road signs play an important role in our daily life which used to guide drivers to notice variety of road conditions and cautions. They provide important visual information that can help drivers operating their vehicles in a manner for enhancing traffic safety. The occurrence of some accidents can be reduced by using automatic road signs recognition system which can alert the drivers. This research attempts to develop a warning system to alert the drivers to notice the important road signs early enough to refrain road accidents from happening. For solving this, a non-linear weighted color enhancement method by pixels is presented. Due to the advantage of proposed method, different road signs can be detected from videos effectively. With suitably coefficients and operations, the experimental results have proved that the proposed method is robust, accurate and powerful in road signs detection. © 2010 Copyright SPIE - The International Society for Optical Engineering.

ポイントクラウドと球面空間モデルを利用した前方視点からの歩容認識

柳 済群;鎌田 清一郎

電気関係学会九州支部連合大会講演論文集2010(0)p.311 - 3122010年-2010年

CiNii

詳細

概要:In this paper, we propose a novel gait recognition framework which is spherical space model using human point clouds (SSM-HPC). The framework is applied for frontal view gait recognition.Various researches for gait recognition have been used human silhouette images from moving picture. This research used three dimensional point cloud data from stereo camera. This framework can get good result from gait recognition rate than silhouette image.

高ダイナミックレンジ画像トーンマッピング

張 兼;鎌田 清一郎

電気関係学会九州支部連合大会講演論文集2010(0)p.312 - 3132010年-2010年

CiNii

詳細

概要:A common task of tone mapping is to reproduce high dynamic range images (HDR) on low dynamic range (LDR) display devices such as printers and monitors. In this paper, a new tone mapping algorithm is proposed. Compared to the previous algorithms, our approach uses an adaptive surround instead of the traditional pre-defined circular. So the shape of a surround can be changed according to the high-contrast edges, which can effectively avoid halo artifacts but preserve visibility of local details. The experimental results show that this algorithm is effective and easy to use.

Fingerprint image enhancement by super resolution with early stopping

Yuan, Zhi; Wu, Jiong; Kamata, Sei Ichiro; Ahrary, Alireza; Yan, Peimin

Proceedings - 2009 IEEE International Conference on Intelligent Computing and Intelligent Systems, ICIS 20094p.527 - 5312009年12月-2009年12月 

DOIScopus

詳細

概要:This paper addressed the problem of multi-frame image super resolution and its implementation to fingerprint image. The use of computers in fingerprints recognition is highly desirable in many applications where security is an important concern. However, as the performance of fingerprint recognition algorithm relies heavily on the quality of the input fingerprint images, it is essential to incorporate a fingerprint enhancement algorithm in the minutiae extraction module. Typical approaches use Gabor filter to raise the contrast between fingerprint ridges and valleys. But few of them try to increase the very basic resolution of fingerprint image. Therefore, we proposed a computer vision solution to this problem which involves the technique of super resolution. This approach can be incorporated into any traditional fingerprint enhancement algorithm as a pre-processing step, rendering a better overall enhancement result. ©2009 IEEE.

3D face recognition based on fast feature detection and non-rigid iterative closest point

Tong, Can; Kamata, Sei Ichiro; Ahrary, Alireza

Proceedings - 2009 IEEE International Conference on Intelligent Computing and Intelligent Systems, ICIS 20094p.509 - 5122009年12月-2009年12月 

DOIScopus

詳細

概要:This paper presents a 3D face recognition algorithm using fast landmark detection and non-rigid iterative Closest Point (ICP) algorithm. The proposed approach can estimate the facial feature region using the anthropometric face model after pose correction, and accurately detect 9 facial landmarks (nose tip, sellion, inner and outer eye corners, nostrils and mouth center). An extension of ICP algorithm has also been proposed to matching the non-rigid 3D face shapes. Experimental results demonstrate that compared to the existing methods, the proposed approach can efficiently detect human facial landmarks and satisfactorily deal with the 3D face matching problem. ©2009 IEEE.

NIR: Content based image retrieval on cloud computing

Yang, Zhuo; Kamata, Sei Ichiro; Ahrary, Alireza

Proceedings - 2009 IEEE International Conference on Intelligent Computing and Intelligent Systems, ICIS 20093p.556 - 5592009年12月-2009年12月 

DOIScopus

詳細

概要:NIR is an open source cloud computing enabled content based image retrieval system. With the development and popularization of cloud computing, more and more researchers from different research areas do research with the help of cloud computing. Nowadays content based image retrieval as one of the challenging and emerging technologies is high computation task because of the algorithm computation complexity and big amount of data. As based on cloud computing infrastructure, NIR is easy to extent and flexible for deployment. As an open source project, NIR can be improved on demand and integrated to other existing systems. This paper presents our ideas, findings, design and the system from our work of NIR. ©2009 IEEE.

Linear predictor using 3-D projection for video lossless compression

Bang, Daejung; Tang, Haijiang; Kamata, Sei Ichiro

IEEE International Symposium on Industrial Electronicsp.1914 - 19182009年12月-2009年12月 

DOIScopus

詳細

概要:Recently, video lossless compression has been developed for applying it to digital cinema, video archiving of contents, etc. Video lossless compression is important in image processing problem since a large image requires a large amount of storage space. The purpose of this paper is to enhance the predictor used for the lossless compression of video. In this paper, we propose the 3-dimensional predictor for the effective prediction. In addition, the three-dimensional spatio-temporal gradient is adopted to improve the conventional image compression methods such as GAP, MED which are two-dimensional predictions based on horizontal and vertical gradients. The spatio-temporal gradient is a spatial data resulted from the projection of triangular prism composed of the neighborhood pixels. From the experimental results compared with the previous prediction methods, we confirmed that the prediction using proposed method is more efficient. ©2009 IEEE.

Face recognition with local feature patterns and histogram spatially bonstrained earth mover's distance

Zhou, Wei; Ahrary, Alireza; Kamata, Sei Ichiro

ICSIPA09 - 2009 IEEE International Conference on Signal and Image Processing Applications, Conference Proceedingsp.374 - 3792009年12月-2009年12月 

DOIScopus

詳細

概要:In this work, two novel local feature patterns-Modified Local Binary patterns (MLBP) and local Ternary patterns (LIP), are proposed for extract features in the facial image, which use some distinct rule to code the values in a label, respectively. These patterns are more invariant to illuminance and face expression compared to traditional one. After getting the local feature patterns, in order to take alignment of face into account, a novel matching method called Histogram Spatially constrained Earth Mover's Distance(HSEMD) is proposed. In this step, the source image is partitioned into non-overlapping local regions while the destination image is represented as a set of overlapping local regions at different positions. Meanwhile, multi-scale cascade mechanism is studied for extracting more feature patterns and obtaining global information of the face.The performance of the proposed method is assessed in the face recognition problem under different challenges. The experimental results show that the proposed method has higher accuracy than some other classic methods.

Face Recognition with Local Feature Patterns and Histogram Spatially constrained Earth Mover's Distance

Wei Zhou, Alireza Ahrary, Sei-ichiro Kamata

Proc. of the IEEE International Conference on Signal and Image Processing Applications (ICSIPA)p.374 - 3792009年11月-

5-3 図書に記された文字の認識と読み上げ(第5部門 ヒューマンインフォメーション2)

小松原 幸弘;山内 幸治;鎌田 清一郎

映像情報メディア学会冬季大会講演予稿集(2009)p."5 - 3-1"2009年11月-2009年11月 

CiNii

詳細

ISSN:13434357

概要:In this study, we try to read printed characters on the book. This paper describes the method to segment a sentence to characters and search a pronunciation of the character by the dictionary.

荒天時の車載カメラ映像におけるフロントガラスへの付着物による前方遮蔽部の修繕に関する一考察(高精細画像の処理・表示,及び一般)

稲葉 洋;鎌田 清一郎

電子情報通信学会技術研究報告. IE, 画像工学109(292)p.33 - 382009年11月-2009年11月 

CiNii

詳細

ISSN:09135685

概要:車載カメラは車両周辺の環境を入力するためのセンサとして広く用いられる.しかしながら,センサとしてのカメラは晴天時では有効に働くものの,荒天時ではレンズへの付着物により画像の有効部分が減り性能低下が懸念される.同様の状況は,ドライブレコーダのようなカメラを有する車内設置型の装置において,車両のガラスを通して車外を撮影する場合にも起こりうる.本研究では,荒天時の車載カメラ映像に対する視認性改善に向けた基礎的検討として後者の状況に着目し,フロントガラスへの付着物により前方が遮蔽された領域の修繕を試みる.手法は,付着物として雨滴を想定し,一台の一般的なカメラを用いて雨滴の検出,および,その部分の修繕を行い,雨滴が除去された映像を生成する.各時刻における画像の修繕は,直前の数フレーム分の画像を用い,それらから雨滴の存在しない部分を統合して行う.本文では一台のカメラの映像を用いて修繕を行う際の問題点について考察し,実験においていくつかの修繕例を示す.

局所的指標による予測器選択を用いた可逆画像圧縮

杉本憲治郎,黒木祥光,鎌田清一郎

電子情報通信学会論文誌(D)J92-D(10)p.1698 - 17012009年10月-

A new on-line signature verification algorithm using hilbert scanning patterns

Ahrary, Alireza; Kamata, Sei Ichro

Digest of Technical Papers - IEEE International Conference on Consumer Electronicsp.276 - 2792009年10月-2009年10月 

DOIScopus

詳細

ISSN:0747668X

概要:Signature verification is a challenging task, because only a small set of genuine samples can be acquired and usually no forgeries are available in real application. In this paper, we propose a new function-based method with Hilbert Scanning patterns for automatic on-line signature verification. The proposed method is compared to other state-of-the-art methods based on the results of the First International Signature Verification Competition (SVC 2004). Experiments are conducted to verify the effectiveness of our method. ©2009 IEEE.

局所的指標による予測器選択を用いた可逆画像圧縮(研究速報,画像符号化,<特集>画像符号化・映像メディア処理レター)

杉本 憲治郎;黒木 祥光;鎌田 清一郎

電子情報通信学会論文誌. D, 情報・システム92(10)p.1698 - 17012009年10月-2009年10月 

CiNii

詳細

ISSN:18804535

概要:予測符号化において,隣接画素間の差分(こう配)に基づき予測器を選択する手法が提案されている.しかしながらこう配と予測器間の関係についての議論は定性的な場合が多い.本論文では,こう配の代わりに平均予測値との差分を局所的指標として用い,従来手法における予測器との関係を定量的に議論する.また局所的指標を用いた予測手法及びスキームを提案し,その予測効率と符号量を測定する.

Image registration based on genetic algorithm and weighted feature correspondences

Yuan, Zhi; Ahrary, Alireza; Yan, Peimin; Yan, Peimin; Yan, Peimin; Kamata, Sei Ichiro

Digest of Technical Papers - IEEE International Conference on Consumer Electronicsp.42 - 462009年10月-2009年10月 

DOIScopus

詳細

ISSN:0747668X

概要:Super resolution is a technique of enhancing image resolution by combining information from multiple images. It is widely applied in fields like camera surveillance, satellite imaging, pattern recognition, etc. One challenging problem of super resolution is its high demand on image registration accuracy. This paper introduces a high accuracy registration approach for the purpose of super resolution. It is invariant to translation, scaling, rotation, and noise, and can be used to automatically obtain the Maximize a Likelihood Estimation (MLE) of image homography (registration result) using information only contained within the images themselves. An effective Genetic Algorithm based approach is used to filter out all the mismatches. Comparison with RANSAC and Keren's method will be given to prove the effectiveness of the proposed method. ©2009 IEEE.

Face recognition using local quaternion patters and weighted spatially constrained earth mover's distance

Zhou, Wei; Ahrary, Alireza; Kamata, Sei Ichiro

Digest of Technical Papers - IEEE International Conference on Consumer Electronicsp.285 - 2892009年10月-2009年10月 

DOIScopus

詳細

ISSN:0747668X

概要:This paper presents a novel algorithm for face recognition. Local Quaternion Patters (LQP) is proposed for presenting the feature parts in the face. To keep the spatial feature of the face, an asymmetric similarity measure Weighted Spatially constrained Earth Mover's Distance (WSEMD) is studied for classification. In this step, the source image is partitioned into non overlapping local patches while the destination image is represented as a set of overlapping local patches at different positions and Gaussian Kernel is used. Finally, local and global weighting is applied to get a more accurate classifier. To evaluate the proposed method and its performance, three well-known and challenge face databases - ORL, Yale and FERET are used in our study. The experimental results show that the proposed method has higher accuracy than some other classic methods. ©2009 IEEE.

An adaptive tone mapping algorithm for high dynamic range images

Zhang, Jian; Kamata, Sei Ichro

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)5646 LNCSp.207 - 2152009年09月-2009年09月 

DOIScopus

詳細

ISSN:03029743

概要:Real world scenes contain a large range of light intensities which range from dim starlight to bright sunlight. A common task of tone mapping algorithms is to reproduce high dynamic range(HDR) images on low dynamic range(LDR) display devices such as printers and monitors. In this paper, a new tone mapping algorithm is proposed for the display of HDR images. Inspired by the adaptive process of the human visual system, the proposed algorithm utilized the center-surround Retinex processing. The novelty of our method is that the local details are enhanced according to a non-linear adaptive spatial filter (Gaussian filter), whose shape is adapted to high-contrast edges of the image. The proposed method uses an adaptive surround instead of the traditional pre-defined circular. Therefore, the algorithm can preserve visibility and contrast impression of high dynamic range scenes in the common display devices. The proposed method is tested on a variety of HDR images, and we also compare it to previous work. The results show good performance of our method in terms of visual quality. © 2009 Springer Berlin Heidelberg.

On-Line signature matching based on hilbert scanning patterns

Ahrary, Alireza; Ahrary, Alireza; Chiang, Hui Ju; Kamata, Sei Ichiro

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)5558 LNCSp.1190 - 11992009年09月-2009年09月 

DOIScopus

詳細

ISSN:03029743

概要:Signature verification is a challenging task, because only a small set of genuine samples can be acquired and usually no forgeries are available in real application. In this paper, we propose a novel approach based on Hilbert scanning patterns and Gaussian mixture models for automatic on-line signature verification. Our system is composed of a similarity measure based on Hilbert scanning patterns and a simplified Gaussian mixture model for decision-level evaluation. To be practical, we introduce specific simplification strategies for model building and training. The system is compared to other state-of-the-art systems based on the results of the First International Signature Verification Competition (SVC 2004). Experiments are conducted to verify the effectiveness of our system. © Springer-Verlag Berlin Heidelberg 2009.

Face recognition using Local Quaternion Patters and Weighted Spatially constrained Earth Mover's Distance

Wei ZHOU; Alireza AHRARY and Seiichiro KAMATA

Proc. of IEEE 13th Int. Symp. on Consumer Electronics, ISCE2009p.285 - 2892009年05月-

Image registration based on Genetic Algorithm and weighted feature correspondences

Zhi YUAN; Alireza AHRARY, P. YAN, Seiichiro KAMATA

Proc. of IEEE 13th Int. Symp. on Consumer Electronics, ISCE2009p.42 - 462009年05月-

接近する人物に対しての歩容認証の一検討

萩尾 和也;アハラリ アリレザ;鎌田 清一郎

電気関係学会九州支部連合大会講演論文集2009(0)p.408 - 4082009年-2009年

CiNii

詳細

概要:人物の歩行特徴から個人を特定する歩容認証は,被験者にかかる負担が小さな生体認証の一つとして近年注目を集めている.しかし,多くの研究では,カメラから十分に離れた人物が側面から撮影されていることを想定しており,歩行方向やカメラの視線方向,被験者とカメラ間の距離の変化の及ぼす影響については十分に議論されていない.そこで本研究では,独自の撮影環境を設営し,カメラに対して被験者が接近してくる場合の歩容認証について検討する.また,撮影にはステレオカメラを用い,歩容認証における立体情報の有用性ついても検討する.この研究の成果は,建物での入退室管理への応用などに活かすことができると考えられる.

Face Recognition with Multi-Scan and Histogram Spatially constrained Earth Mover's Distance

周 蔚;アハラリ アリレザ;鎌田 清一郎

電気関係学会九州支部連合大会講演論文集2009(0)p.527 - 5272009年-2009年

CiNii

詳細

概要:In this work, a multi-scan is applied for each block in the facial image while entropy is used for selecting the best scan order in that block. Based on that scan order, local feature pattern is proposed to obtain feature histograms in the face. Then, a novel matching method called Histogram Spatially constrained Earth Mover's Distance is proposed to take alignment of face into account. The experimental results show that the proposed method has higher accuracy than some other classic methods.

Fast Facial Feature Point Detection for 3D Face Recognition

童 燦;鎌田 清一郎

電気関係学会九州支部連合大会講演論文集2009(0)p.528 - 5282009年-2009年

CiNii

詳細

概要:A face recognition system that utilizes three-dimensional shape information is more robust to arbitrary view, lighting, and facial appearance. The main problem in 3D face recognition is how to detect the feature points correctly and efficiently. In this paper, we present a novel method to detect feature points for 3D face recognition. The experiment result shows that the proposed method performs better in both accuracy and efficiency than other methods.

Single Medicine Recognition using Color Histogram

Cai Qi;アハラリ アリレザ;鎌田 清一郎

電気関係学会九州支部連合大会講演論文集2009(0)p.529 - 5292009年-2009年

CiNii

詳細

概要:In this paper, a new method using color histogram is proposed for recognizing printed character of single medicine. In different brightness condition, the three-dimensional color histogram also changes and a naive three-dimensional partition of color space often supports poor indexing. To circumvent this problem, image RGBs are mapped to brightness-independent chromaticity prior to indexing. Then, chi-square is used for measuring the similarity between the sample medicines and testing medicine. The experimental results show that the proposed method gives an acceptable result.

パラレル順次走査に基づく領域抽出のための新色記述

Yang Zhuo;アハラリ アリレザ;鎌田 清一郎

電気関係学会九州支部連合大会講演論文集2009(0)p.530 - 5302009年-2009年

CiNii

詳細

概要:A color descriptor is a numeric quantity that describes a color feature of an image. Color can be extracted from the image as a whole, a global characterization; or separately from different regions, producing a local characterization. One of the drawbacks of extracting color histograms globally is that it does not take into account the spatial distribution of the color across different areas of the image. In this paper a new color descriptor based on parallel progressive scan is proposed that solve real world problems in region detection for color images.

Adaptive Local Contrast Enhancement for the Visualization of High Dynamic Range Images

Jian ZHANG and Sei-ichiro KAMATA

Proc. of 19th Int. Conf. on Pattern Recognition2008年12月-

Automatic Road Sign Detection Method Based on Color Barycenters Hexagon Model

Qieshi ZHANG and Sei-ichiro KAMATA

Proc. of 19th Int. Conf. on Pattern Recognition2008年12月-

An iterative image enhancement algorithm and a new evaluation framework

Tian, Li; Kamata, Sei Ichiro

IEEE International Symposium on Industrial Electronicsp.992 - 9972008年12月-2008年12月 

DOIScopus

詳細

概要:Image enhancement is important for images captured in low contrast and low illumination conditions. In this study, we propose a new iterative algorithm for image enhancement based on analysis on embedded surfaces of images. In our method, scaled surface area and the surface volume are proposed and used to reconstruct the image iteratively for contrast enhancement, and the illumination of the reconstructed image can also be adjusted simultaneously. On the other hand, the most common methods for measuring the quality of enhanced images Mean Square Error (MSE) or Peak Signal-to-Noise-Ratio (PSNR) have been recognized as inadequate measures because they do not evaluate the result in the way that the human vision system does. This paper also presents a new framework for evaluating image enhancement using both objective and subjective measures. This framework can also be used for other image quality evaluations such as denoising evaluation. We compare our enhancement method with some well-known enhancement algorithms, including wavelet and curvelet methods, using the new evaluation framework. The results show that our method gives better performance in most objective and subjective criteria than the conventional methods. © 2008 IEEE.

Face detection and tracking in color images using color centroids segmentation

Zhang, Qieshi; Kamata, Sei Ichiro; Zhang, Jun

2008 IEEE International Conference on Robotics and Biomimetics, ROBIO 2008p.1008 - 10132008年12月-2008年12月 

DOIScopus

詳細

概要:Human face detection plays an important role in many application areas such as video surveillance, human computer interface, face recognition, face search and face image database management etc. In human face detection applications, face region usually form an inconsequential part of images. Consequently, preliminary segmentation of images into regions that contain "non-face" objects and regions that may contain "face" candidates can greatly accelerate the process of human face detection. Color information based methods take a great attention, because colors have obviously character and robust visual cue for detection. This paper proposed a new method based on RGB color centroids segmentation (CCS) for face detection. This paper include two parts, first part is color image thresholding based on CCS and the second part is face detection based on region growing and facial features structure character combined method. The experimental results show the ideal thresholding result and better than the result of other color space analysis based thresholding methods. Proposed method can conquer the influence of different background conditions, position, scale instance and orientation in images from several photo collections and database; the effect is also better than existing skin color segmentation based methods. © 2008 IEEE.

Adaptive local contrast enhancement for the visualization of high dynamic range images

Zhang, Jian; Kamata, Sei Ichiro

Proceedings - International Conference on Pattern Recognition2008年12月-2008年12月 

Scopus

詳細

ISSN:10514651

概要:In this paper, we present a new tone mapping algorithm for the display of high dynamic range images, inspired by adaptive process of the human visual system. The proposed algorithm is based on the center-surround Retinex processing. In our method, the local details are enhanced according to a non-linear adaptive spatial filter (Gaussian filter), whose shape (filter variance) is adapted to high-contrast edges of the image. Thus our method does not generate halo artifacts meanwhile preserves visibility and contrast impression of high dynamic range scenes in the common display devices. The proposed method is tested on a variety of HDR images and the results show the good performance of our method in terms of visual quality. © 2008 IEEE.

Automatic road sign detection method based on Color Barycenters Hexagon model

Zhang, Qieshi; Kamata, Sei Ichiro

Proceedings - International Conference on Pattern Recognition2008年12月-2008年12月 

Scopus

詳細

ISSN:10514651

概要:Road sign detection is one of the major concerned topics in the field of driving safety and intelligent vehicle. In this paper, a novel model based on Color Barycenters Hexagon (CBH) is proposed and used to detect road sign usefully. In CBH model, full color images are calculated the color barycenters and get the barycenters region, then automatic select the idea threshold curves to separate the Region of Interest (ROI) of barycenters aiming to detect the road sign. Because of the practically images have many noise, and the existing color space cannot separate the ROI ideally. The proposed CBH model can thresholding the principal color of ROI and have high robust. With suitably thresholding and operations, road sign on various scene images can be detected. © 2008 IEEE.

An N-dimensional pseudo-hilbert scan for arbitrarily-sized hypercuboids

Zhang, Jian; Kamata, Sei Ichiro

IEICE Transactions on Fundamentals of Electronics, Communications and Computer SciencesE91-A(3)p.846 - 8582008年12月-2008年12月 

DOIScopus

詳細

ISSN:09168508

概要:The N-dimensional (N-D) Hilbert curve is a one-to-one mapping betweenN-D space and one-dimensional (1-D) space. It is studied actively in the area of digital image processing as a scan technique (Hilbert scan) because ofits property of preserving the spatial relationship of the N-D patterns.Currently there exist several Hilbert scan algorithms. However, these algorithms have two strict restrictions in implementation. First, recursive functions are used to generate a Hilbert curve, which makes the algorithms complexand computationally expensive. Second, all the sides of the scanned region must have the same size and the length must be a power of two, which limitsthe application of the Hilbert scan greatly. Thus in order to remove these constraints and improve the Hilbert scan for general application, a nonrecursive N-D Pseudo-Hilbert scan algorithm based on two look-up tables is proposed in this paper. The merit of the proposed algorithm is that implementation is much easier than the original one while preserving the original characteristics. The experimental results indicate that the Pseudo-Hilbert scancan preserve point neighborhoods as much as possible and take advantage of the high correlation between neighboring lattice points, and it also shows the competitive performance of the Pseudo-Hilbert scan in comparison with other common scan techniques. We believe that this novel scan technique undoubtedly leads to many new applications in those areas can benefit from reducing the dimensionality of the problem. © 2008 The Institute of Electronics, Information and Communication Engineers.

An automatic image-map alignment algorithm based on Mutual Information and Hilbert scan

Tian, Li; Kamata, Sei Ichiro

European Signal Processing Conference2008年12月-2008年12月 

Scopus

詳細

ISSN:22195491

概要:An algorithm for automatic image-map alignment problem using a new similarity measure named Edge-Based Code Mutual Information (EBCMI) and Hilbert scan is presented in this study. Because image and map are very different in their representations, the normal Mutual Information (MI) using the intensity in traditional alignment method may result in misalignment. To solve the problem, codes which are robust to the differences between the image-map pairs are constructed and Mutual Information of the codes is computed as the similarity measure for the alignment. We convert the 3-D transformation search space in alignment to a 1-D search space sequence by using 3-D Hilbert Scan. A new search strategy is also proposed on the 1-D search space sequence. The experimental results show that the proposed EBCMI outperformed the normal MI and some other similarity measures and the proposed search strategy gives flexibility between efficiency and accuracy for automatic imagemap alignment task.

平均予測値との差分による予測器選択を用いた可逆画像圧縮(高精細度画像処理・表示及び一般)

杉本 憲治郎;黒木 祥光;鎌田 清一郎

電子情報通信学会技術研究報告. IE, 画像工学108(324)p.65 - 692008年11月-2008年11月 

CiNii

詳細

ISSN:09135685

概要:予測符号化において,局所的な指標に基づき予測器を切り替えることで高効率化を狙う手法が数多く提案されている.局所的な指標として隣接画素間の差分(勾配)を用いた手法がMEDやGAPをはじめ多くあるが,勾配と予測器間の関係についての定量的な議論は少ない.本稿では局所的な指標として平均予測値との差分を用い,その指標と予測器間の関係の定量的な議論を試みる.またその議論を基に設計された予測手法およびスキームを提案する.性能比較実験では,提案予測手法はGAPに比べ予測誤差エントロピを0.070[bits/pixel]減少でき,提案スキームはCALICに比べて平均符号長を0.016[bits/pixel]削減できた.

動きベクトルを用いた車載カメラ映像に含まれる雨滴の抽出に関する一考察(高精細度画像処理・表示及び一般)

稲葉 洋;鎌田 清一郎

電子情報通信学会技術研究報告. IE, 画像工学108(324)p.59 - 632008年11月-2008年11月 

CiNii

詳細

ISSN:09135685

概要:本研究では,降雨時の車載カメラ映像に含まれる雨滴により生じる視界不良部の抽出を試みる.提案手法は,映像の隣接フレームにおける画素の動きベクトルを用い,各画素における過去の動きベクトルの解析に基づき雨滴を抽出するものである.本手法を,降雨時の車載カメラ映像1例に適用した結果,4割程度の雨滴を抽出した.具体的には,短時間において,雨滴に背景全体が写り込み,かつ,雨滴の大きさが小さい場合抽出が行え,雨滴に近距離の前方が写り込む,また,雨滴がガラス上部にある場合抽出が困難であった.

A two-stage point pattern matching algorithm using ellipse fitting and dual Hilbert Scans

Tian, Li; Kamata, Sei Ichiro

IEICE Transactions on Information and SystemsE91-D(10)p.2477 - 24842008年10月-2008年10月 

DOIScopus

詳細

ISSN:09168532

概要:Point Pattern Matching (PPM) is an essential problem in many image analysis and computer vision tasks. This paper presents a two-stage algorithm for PPM problem using ellipse fitting and dual Hilbert scans. In the first matching stage, transformation parameters are coarsely estimated by using four node points of ellipses which are fitted by Weighted Least Square Fitting (WLSF). Then, Hilbert scans are used in two aspects of the second matching stage: it is applied to the similarity measure and it is also used for search space reduction. The similarity measure named Hilbert Scanning Distance (HSD) can be computed fast by converting the 2-D coordinates of 2-D points into 1-D space information using Hilbert scan. On the other hand, the N-D search space can be converted to a 1-D search space sequence by N-D Hilbert Scan and an efficient search strategy is proposed on the 1-D search space sequence. In the experiments, we use both simulated point set data and real fingerprint images to evaluate the performance of our algorithm, and our algorithm gives satisfying results both in accuracy and efficiency. Copyright © 2008 The Institute of Electronics, Information and Communication Engineers.

18-6 動画像可逆圧縮のための時空間勾配を用いた3次元予測(第18部門 符号化・セキュリティ)

房 大政;唐 海江;鎌田 清一郎

映像情報メディア学会年次大会講演予稿集(2008)p."18 - 6-1"-"18-6-2"2008年08月-2008年08月 

CiNii

詳細

ISSN:13431846

概要:In this paper, we propose video lossless compression based on 3-dimensional prediction using spatio-temporal gradients. The proposed method predicted the target pixel using the LGM(Local Gradient Magnitude), which is the gradient between the neighborhood pixel and the current pixel in the spatio-temporal. From the experimental results, we demonstrated that the spatio-temporal gradients prediction is more efficient.

18-11 Online signature matching based on Hilbert-Scanning patterns

CHIANG Huiju;ZHANG Jian;AHRARY Alireza;KAMATA Seiichiro

映像情報メディア学会年次大会講演予稿集(2008)p."18 - 11-1"-"18-11-2"2008年08月-2008年08月 

CiNii

詳細

ISSN:13431846

概要:Signature-based personal identification systems are used and accepted widely due to its distinctness and stability. In this paper, we propose a new function-based method with Hilbert-Scanning patterns for signature matching. We evaluate the performance by using the online signature database Signature Verification Competition (SVC) 2004.

Image enhancement by analysis on embedded surfaces of images and a new framework for enhancement evaluation

Tian, Li; Kamata, Sei Ichiro

IEICE Transactions on Information and SystemsE91-D(7)p.1946 - 19542008年07月-2008年07月 

DOIScopus

詳細

ISSN:09168532

概要:Image enhancement plays an important role in many machine vision applications on images captured in low contrast and low illumination conditions. In this study, we propose a new method for image enhancement based on analysis on embedded surfaces of images. The proposed method gives an insight into the relationship between the image intensity and image enhancement. In our method, scaled surface area and the surface volume are proposed and used to reconstruct the image iteratively for contrast enhancement, and the illumination of the reconstructed image can also be adjusted simultaneously. On the other hand, the most common methods for measuring the quality of enhanced images areMean Square Error (MSE) or Peak Signal-to-Noise-Ratio (PSNR) in conventional works. The two measures have been recognized as inadequate ones because they do not evaluate the result in the way that the human vision system does. This paper also presents a new framework for evaluating image enhancement using both objective and subjective measures. This framework can also be used for other image quality evaluations such as denoising evaluation. We compare our enhancement method with some well-known enhancement algorithms, including wavelet and curvelet methods, using the new evaluation framework. The results show that our method can give better performance in most objective and subjective criteria than the conventional methods. Copyright © 2008 The Institute of Electronics, Information and Communication Engineers.

Automatic Image-Map Alignment Using Edge-Based Code Mutual Information and 3-D Hilbert Scan

Li TIAN and Sei-ichiro KAMATA

The Journal of IIEEJ37(3)p.223 - 2302008年05月-

An N-Dimensional Pseudo-Hilbert Scan for Arbitrarily-sized Hypercuboids

Jian ZHANG and Sei-ichiro KAMATA

IEICE Transactions on Fundamentals of Electronics, Communications and Computer SciencesE91-A(3)p.846 - 8582008年03月-

TK-2-4 空間充填曲線による画像圧縮検索(TK-2. 北九州での知的クラスター創成事業(第1期)の概要と成果・課題と展望,大会委員会企画)

鎌田 清一郎

電子情報通信学会総合大会講演論文集2008(1)p."SSS - 6"-"SSS-7"2008年03月-2008年03月 

CiNii

D-11-55 可逆画像圧縮のためのメディアン適応予測の改善(D-11. 画像工学,一般セッション)

杉本 憲治郎;黒木 祥光;鎌田 清一郎

電子情報通信学会総合大会講演論文集2008(2)2008年03月-2008年03月 

CiNii

D-12-109 点パターンマッチングのための二段階マッチングアルゴリズム(D-12. パターン認識・メディア理解,一般セッション)

田 黎;鎌田 清一郎

電子情報通信学会総合大会講演論文集2008(2)2008年03月-2008年03月 

CiNii

分散と重み付きDPマッチングを用いた顔認識

周 蔚;鎌田 清一郎

電気関係学会九州支部連合大会講演論文集2008(0)p.616 - 6162008年-2008年

CiNii

詳細

概要:This paper presents a novel algorithm for face recognition. Variance is used for extracting the feature vector and then Dynamic Programming (DP) is applied for matching, since the length of each feature vector is different. At last, some weighted values are added for final recognition. These weighted values can improve the recognition rate greatly. To evaluate the proposed method and its performance, a well-known face database ORL is used in our study. The experimental results show that the proposed method is much better than other existing method, such as PCA, 2DPCA, LDA, LBP and so on.

色分析に基づく道路標識の自動検出方法

Zhang Qieshi;鎌田 清一郎

電気関係学会九州支部連合大会講演論文集2008(0)p.617 - 6172008年-2008年

CiNii

詳細

概要:Road sign detection is one of the major concerned topics in the field of driving safety and intelligent vehicle. In this paper, a novel model based on Color Barycenters Hexagon (CBH) is proposed and used to detect road sign usefully. In CBH model, full color images are calculated the color barycenters and get the barycenters region, then automatic select the idea threshold curves to separate the barycenters region of interest (ROI) aiming to detect the road sign. Because of the practically images have many noise, and the existing color space cannot separate the ROI ideally, the proposed CBH model can thresholding the principal color of ROI and have high robust. With suitably thresholding and operations, road sign on various scene images can be detected.

POLSAR画像を用いた橋高度の測定

李 虎栄;王 海鵬;鎌田 清一郎

電気関係学会九州支部連合大会講演論文集2008(0)p.63 - 632008年-2008年

CiNii

詳細

概要:POLSARデータを用いた橋高度の測定方法を提案する。de-orientation理論と分類パラメータを用い,Pi-SARのポライメトリックデータでの橋目標の1次、2次と3次散乱の画像生成原理を分析し、画像での位置を判断する。またフィルタリングとクラスタリング処理を行い,SAR画像からそれぞれの散乱画像を抽出し、2次散乱と3次散乱の画像距離によって橋の高度を測定する。最後に日本のナルト大橋のPi-SAR画像と中国の東海大橋のALOS-PALSAR画像を利用してその高度を測定し、実際高度と比較して本手法の有効性を示す。

Fingerprint matching using dual hilbert scans

Tian, Li; Chen, Liang; Kamata, Sei Ichiro

Proceedings - International Conference on Signal Image Technologies and Internet Based Systems, SITIS 20073(1)p.593 - 6002007年12月-2007年12月 

DOIScopus

詳細

概要:A new fingerprint matching algorithm using dual Hilbert scans is presented in this study. We treat the fingerprint matching as point pattern matching problem and Hilbert scans are used in two aspects of the matching problem: one is applied to the similarity measure and the other is used in search space reduction. The similarity measure named Hilbert Scanning Distance (HSD) can be computed fast by converting the 2-D coordinates of 2-D images into 1-D space information using Hilbert scan. On the other hand, the 3-D search space can be converted to a 1-D search space sequence. The proposed method has been tested on FVC2002 database. The experimental results show that our method can implement fingerprint matching robustly and efficiently. The performance evaluation EER (Equal-Error Rate) generally used is very low by our algorithm. © 2008 IEEE.

An N-dimensional pseudo-Hilbert scan algorithm for an arbitrarily-sized hypercuboid

Zhang, Jian; Kamata, Sei Ichiro

IECON Proceedings (Industrial Electronics Conference)33(1)p.2459 - 24642007年12月-2007年12月 

DOIScopus

詳細

概要:The N-dimensional (N-D) Hilbert curve is a one-to-one mapping between N-D space and one-dimensional (1-D) space. It is studied actively in the area of digital image processing as a scan technique (Hilbert scan) because of its property of preserving the spacial relationship of the N-D patterns. Currently there exist several Hilbert scan algorithms. However, these algorithms have two strict restrictions in implementation. First, recursive functions are used to generate a Hilbert curve, which makes the algorithms complex and computationally expensive. Second, all the sides of the scanned region must have same size and each size must be a power of two, which limits the application of the Hilbert scan greatly. In this paper, a nonrecursive N-D Pseudo-Hilbert scan algorithm based on two look-up tables is proposed. The merit of the algorithm is that the computation is fast and the implementation is much easier than the original one. The simulation indicates that the Pseudo-Hilbert scan can preserve point neighborhoods as much as possible and take advantage of the high correlation between neighboring lattice points. It also shows competitive performance of the Pseudo-Hilbert scan in comparison with other common scan techniques. ©2007 IEEE.

Image contrast enhancement by analysis on embedded surfaces of images

Tian, Li; Kamata, Sei Ichiro

Proceedings of IAPR Conference on Machine Vision Applications, MVA 200710(1)p.90 - 932007年12月-2007年12月 

Scopus

詳細

概要:Image contrast enhancement plays an important role in many machine vision applications. In this study, we propose a new method for edge enhancement based on analysis on embedded surfaces of images. The proposed method gives an insight into the relationship between the image intensities (also the gradients) and image contrast. In our method, scaled surface area and the surface volume are used to reconstruct the image for edge enhancement, and then the contrast of the reconstructed image is adjusted by a 'strengthen-weaken' process. Although, current method for edge enhancement such as curvelet transform can enhance the edge part, it does not provide good tonal rendition or color constancy sometimes. The experimental results show that our method can give good performance not only in edge enhancement, but also in tonal rendition and color constancy.

Near-duplicate detection using a new framework of constructing accurate affine invariant regions

Tian, Li; Kamata, Sei Ichiro

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)4781 LNCSp.61 - 722007年12月-2007年12月 

Scopus

詳細

ISSN:03029743

概要:In this study, we propose a simple, yet general and powerful framework for constructing accurate affine invariant regions and use it for near-duplicate detection problem. In our framework, a method for extracting reliable seed points is first proposed. Then, regions which are invariant to most common affine transformations are extracted from seed points by a new method named the Thresholding Seeded Growing Region (TSGR). After that, an improved ellipse fitting method based on the Direct Least Square Fitting (DLSF) is used to fit the irregularly-shaped contours of TSGRs to obtain ellipse regions as the final invariant regions. At last, SIFT-PCA descriptors are computed on the obtained regions. In the experiment, our framework is evaluated by retrieving near-duplicate in an image database containing 1000 images. It gives a satisfying result of 96.8% precision at 100% recall. © Springer-Verlag Berlin Heidelberg 2007.

A Pseudo-Hilbert scan for arbitrarily-sized cuboid region

Zhang, Jian; Kamata, Sei Ichiro

Sixth IEEE International Symposium on Signal Processing and Information Technology, ISSPITp.764 - 7692007年12月-2007年12月 

DOIScopus

詳細

概要:The 3-dimensional (3-D) Hilbert scan is a one-to-one mapping between 3-D data and 1-D data along the 3-D Hilbert curve. It has been applied widely in image processing, such as image compression, object recognition, and image clustering, etc. Now, although there exist some 3-D Hilbert scanning algorithms, they usually have strict limitation on the scanned region. This makes Hilbert scan difficult to be applied in practice. So an effective scanning algorithm for arbitrarily-sized cuboid region is significant to improve the correlative digital image processing technology. In this paper, we proposed a novel Pseudo-Hilbert scanning algorithm based on the look-up tables method for arbitrarily-sized cuboid region. Although the proposed algorithm is designed for 3-D space scanning, it can be also applied in an arbitrary-sized rectangle. The algorithm does not only remove the strict constrains but also reserve the good property of the Hilbert curve preserving point neighborhoods as much as possible. The good performance of the algorithm is demonstrated by the simulation results. © 2006 IEEE.

Diffusion geodesic path: A common feature for automatic image-map registration

Tian, Li; Kamata, Sei Ichiro

Sixth IEEE International Symposium on Signal Processing and Information Technology, ISSPITp.944 - 9492007年12月-2007年12月 

DOIScopus

詳細

概要:The purpose of image-map registration is to revise the digital map included in Geographic Information System (GIS) with aerial image. The traditional method for this task requires the manual selection of features such as tie points in both the image and the map. Determining how to automatically extract common features between image and map is a difficult problem. In this study, we propose a novel framework of extracting the common feature named Diffusion Geodesic Path (DGP) between image and map, and use it for image-map registration. In order to overcome the differences in presentations between image and map, we first construct two new edge images from the image and the map using the diffusion process in physics to diffuse the differences between them. Then, we extract several paths between feature points (corner points) on these new edge images using the geodesic. Finally, a part of the extracted paths are automatically selected and are used for registration. The experimental results implicate that our DGP is a common and robust feature between image and map, and it can provide sufficient information for automatic image-map registration. The whole framework works well on automatic image-map registration. © 2006 IEEE.

A simple tone mapping for high dynamic range image visualization using a Pseudo-Hilbert scan

Zhang, Jian; Kamata, Sei Ichiro; Tian, Li

Proceedings of IAPR Conference on Machine Vision Applications, MVA 2007p.363 - 3662007年12月-2007年12月 

Scopus

詳細

概要:The Hilbert curve is one of space-filling curves published by G. Peano. There are several applications using this curve, such as image processing, computer graphics, etc. In this paper, we concentrate on a tone mapping technique for high dynamic range images using the Pseudo-Hilbert curve. Based on the neighbourhood property of the Pseudo-Hilbert scan, a fast and flexible tone reproduction method is proposed. The proposed new technique preserves visibility and contrast impression of high dynamic range scenes in low dynamic range reproduction devices. From the experimental results, we have confirmed that the proposed method produces good results on a variety of high dynamic range images.

A new framework for constructing accurate affine invariant regions

Tian, Li; Tian, Li; Kamata, Sei Ichiro

IEICE Transactions on Information and SystemsE90-D(11)p.1831 - 18402007年11月-2007年11月 

DOIScopus

詳細

ISSN:09168532

概要:In this study, we propose a simple, yet general and powerful framework for constructing accurate affine invariant regions. In our framework, a method for extracting reliable seed points is first proposed. Then, regions which are invariant to most common affine transformations can be extracted from seed points by two new methods the Path Growing (PG) or the Thresholding Seeded Growing Region (TSGR). After that, an improved ellipse fitting method based on the Direct Least Square Fitting (DLSF) is used to fit the irregularly-shaped contours from the PG or the TSGR to obtain ellipse regions as the final invariant regions. In the experiments, our framework is first evaluated by the criterions of Mikolajczyk's evaluation framework [1], and then by near-duplicate detection problem [2]. Our framework shows its superiorities to the other detectors for different transformed images under Mikolajczyk's evaluation framework and the one with TSGR also gives satisfying results in the application to near-duplicate detection problem. Copyright © 2007 The Institute of Electronics, Information and Communication Engineers.

Adaptive Tone Reproduction for High Dynamic Range Image

Jian ZHANG and Sei-ichiro KAMATA

Proc. of IEVC2007: IIEEJ Image Electronics and Visual Computing Workshop35(1)p.1 - 42007年11月-

Automatic Image-Map Alignment by Maximization of Edge-Based Code Mutual Information

Li TIAN and Sei-ichiro KAMATA

Proc. of IEVC2007: IIEEJ Image Electronics and Visual Computing Workshop35(1)p.1 - 42007年11月-

A Study on PCA-based Fourier Descriptor in Complete and Incomplete Contour Representations

Li TIAN and Sei-ichiro KAMATA

Proc. of subspace2007: Subspace 2007 Workshop on ACCV20071(1)p.75 - 812007年11月-

時空間勾配を用いた3次元予測による動画像の可逆圧縮(高精細画像の処理・表示及び一般)

安 昭映;唐 海江;鎌田 清一郎

電子情報通信学会技術研究報告. IE, 画像工学107(358)p.109 - 1132007年11月-2007年11月 

CiNii

詳細

ISSN:09135685

概要:近年,放送映像などの素材蓄積,ディジタツシネマなどへの応用を目的として,動画像の可逆符号化が検討されている.本論文では,動画像を対象とし,時空間勾配を利用した3次元予測による可逆圧縮について述べる.これは,従来GAR,MEDなどの水平および垂直方向のエッジに着目した予測方式に対して,エッジの方向をより細かく捉えた,2次元空間勾配を利用した静止画像圧縮法を拡張し,3次元時空間勾配を利用した動画像圧縮へ適用したものである.実験の結果,従来手法と比較してより効率の良い時空間予測が実現できるものである.

Navigation of an Autonomous Sewer Inspection Robot Based on Stereo Camera Images and Laser Scanner Data

Alireza AHRARY, Li TIAN, Seiichiro KAMATA, Masumi ISHIKAWA

Int. Journal on Artificial Intelligence Tools16(4)p.611 - 6252007年08月-2007年08月 

DOIScopus

詳細

ISSN:02182130

概要:Sewer environment is composed of cylindrical pipes, in which only a few landmarks such as manholes, inlets and pipe joints are available for localization. This paper presents a method for navigation of an autonomous sewer inspection robot in a sewer pipe system based on detection of landmarks. In this method, location of an autonomous sewer inspection robot in the sewer pipe system is estimated from stereo camera images. The laser scanner data are also used to ensure accurate localization of the landmarks and reduce the error in distance estimation by image processing. The method is implemented and evaluated in a sewer pipe test field using a prototype robot, demonstrating its effectiveness. © World Scientific Publishing Company.

Near-Duplicate Detection Using A New Framework of Constructing Accurate Affine Invariant Regions

Li TIAN and Sei-ichiro KAMATA

Proc. of The 9th Int. Conf. on Visual Information Systems9(1)p.1 - 122007年06月-

A Simple Tone Mapping for High Dynamic Range Image Visualization Using a Pseudo-Hilbert Scan

Jian ZHANG, Sei-ichiro KAMATA, Li TIAN

Proc. of IAPR Conf. on Machine Vision Applications(MVA2007)10(1)p.363 - 3662007年05月-

Fast and accurate singular point extraction of fingerprint

Xu, Xiao; Kamata, Sei Ichiro

International Conference on Signal Processing Proceedings, ICSP22007年05月-2007年05月 

DOIScopus

詳細

概要:Nowadays, biometric verification occupies important position in people's daily life. Because fingerprints are easy to use and provide good performance, fingerprint identification is one of the most popular technologies. In order to maximize the robustness of fingerprint identification and to minimize the identification time, it is better to align the fingerprint first by referring the singular points of fingerprint. In this paper, we proposed a new algorithm to localize the singular points of fingerprint. The proposed approach is less computational complexity and higher precision. This paper shows how, from the x-gradient component of orientation filed and its sine component, very accurate extraction of singular points can be obtained. This localization can be used for accurate registration of two fingerprints. © 2006 IEEE.

A pseudo-hilbert scan for arbitrarily-sized arrays

Zhang, Jian; Kamata, Sei Ichiro; Ueshige, Yoshifumi

IEICE Transactions on Fundamentals of Electronics, Communications and Computer SciencesE90-A(3)p.682 - 6902007年03月-2007年03月 

DOIScopus

詳細

ISSN:09168508

概要:The 2-dimensional (2-D) Hilbert curve is a one-to-one mapping between 2-D space and one-dimensional (1-D) space. It is studied actively in the area of digital image processing as a scan technique (Hilbert scan) because of its property of preserving the spacial relationship of the 2-D patterns. Currently there exist several Hilbert scan algorithms. However, these algorithms have two strict restrictions in implementation. First, recursive functions are used to generate a Hilbert curve, which makes the algorithms complex and computationally expensive. Second, both sides of the scanned rectangle must have same size and each size must be a power of two, which limits the application of the Hilbert scan greatly. In this paper, a Pseudo-Hilbert scan algorithm based on two look-up tables is proposed. The proposed method improves the Hilbert scan to be suitable for real-time processing and general application. The simulation indicates that the Pseudo-Hilbert scan can preserve point neighborhoods as much as possible and take advantage of the high correlation between neighboring lattice points. It also shows competitive performance of the Pseudo-Hilbert scan in comparison with other scan techniques.

D-12-116 点パターンマッチングための非線形最小二乗フィッティングに基づく変換パラメーター推定(D-12.パターン認識・メディア理解,一般講演)

田 黎;鎌田 清一郎

電子情報通信学会総合大会講演論文集2007(2)2007年03月-2007年03月 

CiNii

An efficient algorithm for point matching using hilbert scanning distance

Tian, Li; Kamata, Sei Ichiro

Proceedings - International Conference on Pattern Recognition3(2)p.873 - 8762006年12月-2006年12月 

DOIScopus

詳細

ISSN:10514651

概要:A fast and accurate similarity named Hilbert Scanning Distance(HSD) [9] has recently been presented for point matching. In this study, we improved an efficient algorithm of search strategy for HSD in the large search space. This search strategy is associated with two ideas: a relaxation greedy search, and an accelerating process using Monte Carlo sampling. The experimental results implicate that this improved algorithm is robust and efficient for point matching using HSD. It also makes a tradeoff between accuracy and speed under different requirements.

A low-complexity deformation invariant descriptor

Tian, Li; Kamata, Sei Ichiro

Proceedings - International Conference on Pattern Recognition2(1)p.227 - 2302006年12月-2006年12月 

DOIScopus

詳細

ISSN:10514651

概要:In this paper, we propose a descriptor which is invariant to general deformations (only intensity locations change but not their value) by using Hilbert scanning. In our method, an image is converted to a 1-D sequence through Hilbert scanning at first. Then, we embed this sequence as a 1-D curve in the 2-D space. Because Hilbert scanning preserves the coherence in a 2-D image, it is easily to understand that the area under the curve is invariant to intensity location changes, naturally. Hence, we use some areas for an interest point as a deformation invariant descriptor. This descriptor can be computed in the 2-D space efficiently than other approaches where an image is embedded in the 3-D space or the dimensions of descriptors are very large. The experimental results show that our descriptor is low-complexity and superior to other approaches on interest point matching in deformation images. © 2006 IEEE.

A study of bias correction methods for enhancing median edge detector prediction

Tang, Haijiang; Kamata, Sei Ichiro; Tsuneyoshi, Kazuyuki

2005 IEEE 7th Workshop on Multimedia Signal Processing2006年12月-2006年12月 

DOIScopus

詳細

概要:In this paper, we present three novel lossless compression approaches for gray-scale continuous tone natural image. Our methods enhance the median edge detector (MED), which is the core part of JPED-LS algorithm, by reducing the entropy of the prediction error via adaptive regression. These modified predictors improve the prediction accuracy by reducing the negative effect due to MED's oversimplified edge orientation detection. The experimental results show that our approaches achieve evidently better performance than MED with only neglectable increasing of computational complexity and without introduce extra pixels into the causal template.

Fast and Accurate Singular Point Extraction of Fingerprint

Xiao Xu, Sei-ichiro Kamata

Proceedings of IEEE TENCON200627(1)p.1 - 42006年11月-

複数の直交基底を用いた最近傍コードワードの高速探索法

黒木 祥光;高橋 幸太郎;上繁 義史;鎌田 清一郎

映像情報メディア学会技術報告30(62)p.29 - 342006年11月-2006年11月 

CiNii

詳細

ISSN:13426893

3次元空間における一般ヒルベルト走査(高精細画像の処理・表示,及び一般)

張 兼;鎌田 清一郎

電子情報通信学会技術研究報告. IE, 画像工学106(397)p.35 - 392006年11月-2006年11月 

CiNii

詳細

ISSN:09135685

概要:The three-dimensional Hilbert curve is a one-to-one mapping between three-dimensional (3-D) space and one-dimensional (1-D) space. Due to the advantage of preserving high correlation of 3-D points, it receives much attention in many areas. Especially as a scan technique (Hilbert scan), the Hilbert curve is studied actively in image processing. Although there exist several Hilbert scan algorithms now, they have strict implementation conditions. First, they use recursive functions to generate scanning curves, which makes the algorithms complex and difficult to implement in real-time systems. Second, the scanned region must be a square and the length must be the power of two, which limits the application of the Hilbert scan greatly, In this paper, an effective 3-D Pseudo-Hilbert scan algorithm for an arbitrarily-sized cuboid is proposed. The proposed algorithm improves the Hilbert scan for general application. Moreover, it greatly reduces the computational complexity and saves storage memory by using two simple look-up tables instead of recursive functions. Therefore the algorithm is suitable for real-time processing. The experimental results show that the Pseudo-Hilbert scan preserves the most structures of the Hilbert scan. Although the proposed algorithm is 3-D case, it is also feasible in 2-D space. We believe this novel scan technique undoubtedly leads to many new applications in those areas which can benefit from reducing the dimensionality of the problem.

画像の一般変形に対する低次元の不変特徴記述子(高精細画像の処理・表示,及び一般)

田 黎;鎌田 清一郎

電子情報通信学会技術研究報告. IE, 画像工学106(397)p.41 - 442006年11月-2006年11月 

CiNii

詳細

ISSN:09135685

概要:本論文では、輝度の位置だけは変化し、輝度値は変化しない一般変形画像を対象とし、ヒルベルト走査に基づいた低次元の不変特徴記述子を提案する。本手法は、まずヒルベルト走査を利用して画像情報を1次元情報に変換し、その1次元情報を曲線と見なして面積を計算するものである。これは、ヒルベルト走査の性質により、曲線下の面積が一般変形に不変となる。本記述子は、従来の記述子と比較して次元が低く、一般変形に対して不変となることを確認した。

Diffusion Geodesic Path: A common feature for automatic image-map registration

Li Tian, Sei-ichiro Kamata

Proc. of the 6th IEEE Symposium on Signal Processing and Information Technology (ISSPIT2006)6(1)p.944 - 9492006年08月-

A Pseudo-Hilbert Scan for Arbitrarily-Sized Rectangle Region

Jian Zhang, Sei-ichiro Kamata and Yoshifumi Ueshige

Proc. of the Int. Workshop on Intelligent Computing in Pattern Analysis/Synthesis (IWICPAS2006)LNCS4153(1)p.290 - 2992006年08月-

Voting weighted modified Hausdorff distance through multiscale space for automatic image-map registration

Li Tian , Sei-ichiro Kamata

Proc. of the 18th Int. Conf. on Pattern Recognition (ICPR2006)18(2)p.837 - 8402006年08月-2006年12月 

DOIScopus

詳細

ISSN:10514651

概要:The purpose of image-map registration is to revise the digital map included in Geographic Information System (GIS) with aerial image. The traditional method for this task requires the manual selection of tie points in both image and map. In this study, we propose a distance measure named Voting Weighted Modified Hausdorff Distance (VWMHD) for this task. In order to overcome the differences in representations between image and map in urban area, after several times of edge extraction through multiscale space, we give weights to each edge point in the initial scale based on its voting times and then compute the VWMHD for registration. The experimental results implicate that our VWMHD can provide sufficient information for automatic image-map registration and is robust to noises. © 2006 IEEE.

I_029 拡散過程による自動画像地図照合ための共通特徴の抽出(I分野:画像認識・メディア理解)

田 黎;鎌田 清一郎;恒吉 和幸

情報科学技術フォーラム一般講演論文集5(3)p.67 - 702006年08月-2006年08月 

CiNii

A gradient based predictive coding for lossless image compression

Tang, Haijiang; Kamata, Sei Ichiro

IEICE Transactions on Information and SystemsE89-D(7)p.2250 - 22562006年07月-2006年07月 

DOIScopus

詳細

ISSN:09168532

概要:Natural, continuous tone images have a very important property of high correlation of adjacent pixels. Images which we wish to compress are usually non-stationary and can be reasonably modeled as smooth and textured areas separated by edges. This property has been successfully exploited in LOCO-I and CALIC by applying gradient based predictive coding as a major de-correlation tool. However, they only examine the horizontal and vertical gradients, and assume the local edge can only occur in these two directions. Their over-simplified assumptions hurt the robustness of the prediction in higher complex areas. In this paper, we propose an accurate gradient selective prediction (AGSP) algorithm which is designed to perform robustly around any type of image texture. Our method measures local texture information by comparison and selection of normalized scalar representation of the gradients in four directions. An adaptive predictor is formed based on the local gradient information and immediate causal pixels. Local texture properties are also exploited in the context modeling of the prediction error. The results we obtained on a test set of several standard images are encouraging. On the average, our method achieves a compression ratio significantly better than CALIC without noticeably increasing of computational complexity. Copyright © 2006 The Institute of Electronics, Information and Communication Engineers.

ヒルベルト曲線による点照合ための新しい類似度計算法(一般セッション(5),CVのためのパターン認識・学習理論の新展開)

田 黎;鎌田 清一郎;恒吉 和幸

電子情報通信学会技術研究報告. PRMU, パターン認識・メディア理解105(674)p.161 - 1662006年03月-2006年03月 

CiNii

詳細

ISSN:09135685

概要:本研究では、ヒルベルト曲線を利用した点照合ための新しい類似度計算法について述べる.これは、ヒルベルト曲線を利用して、二次元の点情報を一次元点情報に変換し、一次元上で高速に類似度を計算するものである.点照合ための従来手法と比較して、計算量が少なく、雑音の影響を受けにくいことを確認した.

A pseudo-hilbert scan algorithm for arbitrarily-sized rectangle region

Zhang, Jian; Kamata, Sei Ichiro; Ueshige, Yoshifumi

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)4153 LNCS(1)p.290 - 2992006年01月-2006年01月 

Scopus

詳細

ISSN:03029743

概要:The 2-dimensional Hilbert scan (HS) is a one-to-one mapping between 2-dimensional (2-D) space and one-dimensional (1-D) space along the 2-D Hilbert curve. Because Hilbert curve can preserve the spatial relationships of the patterns effectively, 2-D HS has been studied in digital image processing actively, such as compressing image data, pattern recognition, clustering an image, etc. However, the existing HS algorithms have some strict restrictions when they are implemented. For example, the most algorithms use recursive function to generate the Hilbert curve, which makes the algorithms complex and takes time to compute the one-to-one correspondence. And some even request the sides of the scanned rectangle region must be a power of two, that limits the application scope of HS greatly. Thus, in order to improve HS to be proper to real-time processing and general application, we proposed a Pseudo-Hilbert scan (PHS) based on the look-up table method for arbitrarily-sized arrays in this paper. Experimental results for both HS and PHS indicate that the proposed generalized Hilbert scan algorithm also reserves the good property of HS that the curve preserves point neighborhoods as much as possible, and gives competitive performance in comparison with Raster scan. © Springer-Verlag Berlin Heidelberg 2006.

A fast and accurate algorithm for matching images using hilbert scanning distance with threshold elimination function

Tian, Li; Tian, Li; Kamata, Sei Ichiro; Kamata, Sei Ichiro; Kamata, Sei Ichiro; Tsuneyoshi, Kazuyuki; Tsuneyoshi, Kazuyuki; Tang, Haijiang

IEICE Transactions on Information and SystemsE89-D(1)p.291 - 2972006年01月-2006年01月 

DOIScopus

詳細

ISSN:09168532

概要:To find the best transformation between a "model" point set and an "image" point set is the main purpose of point pattern matching. The similarity measure plays a pivotal role and is used to determine the degree of resemblance between two objects. Although some well-known Hausdorff distance measures work well for this task, they are very computationally expensive and suffer from the noise points. In this paper, we propose a novel similarity measure using the Hilbert curve named Hilbert scanning distance (HSD) to resolve the problems. This method computes the distance measure in the one-dimensional (1-D) sequence instead of in the two-dimensional (2-D) space, which greatly reduces the computational complexity. By applying a threshold elimination function, large distance values caused by noise and position errors (e.g. those that occur with feature or edge extraction) are removed. The proposed algorithm has been applied to the task of matching edge maps with noise. The experimental results show that HSD can provide sufficient information for image matching within low computational complexity. We believe this sets a new direction for the research of point pattern recognition. Copyright © 2006 The Institute of Electronics, Information and Communication Engineers.

2次元空間における擬似ヒルベルト走査

張 兼;鎌田 清一郎;上繁 義史

電気関係学会九州支部連合大会講演論文集2006(0)p.478 - 4782006年-2006年

CiNii

指紋画像の高精度特異点抽出

許 霄;鎌田 清一郎;黒木 祥光

電気関係学会九州支部連合大会講演論文集2006(0)p.488 - 4882006年-2006年

CiNii

An autonomous sewer robots navigation based on stereo camera information

Ahrary, Alireza; Tian, Li; Kamata, Sei Ichiro; Ishikawa, Masumi

Proceedings - International Conference on Tools with Artificial Intelligence, ICTAI2005(1)p.628 - 6332005年12月-2005年12月 

DOIScopus

詳細

ISSN:10823409

概要:In this paper, we propose a method for an autonomous sewer robots to navigate through a sewer pipe system based on stereo camera information. In this method, a local features such as manholes and pipe joints are extracting as a feature pixels in the Region of Interest (ROI) of left image. Then, an accurate and fast stereo matching measure named Linear Computation is implemented in this ROI image to compute the distance between the robots and local features. Finally, the distance data can be used for navigation map in sewer pipe system. The experimental results show that our method can provide sufficient information for an autonomous sewer robots navigation. © 2005 IEEE.

Matching vehicles using hubert scanning distance

Li, Tian; Kamata, Sei Ichiro; Tsuneyoshi, Kazuyuki

2005 IEEE International Conference on Vehicular Electronics and Safety Proceedings2005(1)p.149 - 1542005年12月-2005年12月 

DOIScopus

詳細

概要:Matching objects is a fundamental problem for any object detection system. Feature-based methods in matching objects such as vehicles often encounter the problem of correspondences between features of two related patterns. The features may be points, lines, curves and regions. Point pattern matching (PPM) is a primary and essential approach for establishing a correspondence within two related patterns. Although some well-known Hausdorff distance measures work well for this task, they are very computational expensive and suffer from the noise of images. In this paper, we propose a novel similarity measure using Hilbert curve named Hilbert scanning distance (HSD) to resolve the problems. This method computes the distance measure in one-dimensional (1-D) sequence in stead of in two-dimensional (2-D) image space, which greatly reduce the computational complexity. By applying a threshold elimination function, extreme distances caused by noise and position errors (e.g. those that occur with feature or edge extraction) are removed. The experimental results show that HSD can provide sufficient information for matching -vehicles within low computational complexity. We believe this point out a new direction for the research of PPM. © 2005 IEEE.

Correction Methods for Enhancing Median Edge Detector Prediction

Haijiang Tang, Sei-ichiro Kamata and Kazuyuki Tsuneyoshi

Proc. of 2005 IEEE Seventh Workshop on Multimedia Signal Processing7(1)p.545 - 5482005年11月-

A cooperative stereo matching algorithm for sewer inspection robots

Alireza Ahrary, Li Tian, Sei-ichiro Kamata and Masumi Ishikawa

Proc. of IASTED International Conference on Robotics and Applications11(1)p.294 - 2992005年10月-

Image compression and retrieval using Hilbert curve

Noritatsu Matsuo, Sei-ichiro Kamata and Kazuyuki Tsuneyoshi

Proc. of International Astronautical Congress56(B1.P.09)p.1 - 62005年10月-2005年12月 

Scopus

詳細

概要:Huge amount of high resolution satellite images for remote sensing are available in a couple of decades. Image compression method is required or highly desired in this task. Image coding standards such as JPEG and MPEG are widely used for compressing these images. However, JPEG (or MPEG) compressed data stream can not be processed without decoding. In this paper, we concentrate on image compression and retrieval without decoding by using Hilbert curve. Hubert curve is one of space filling curves which pass through all points in a space. The merits of our compression method is that (1) the compression ratio is almost equivalent to JPEG2000, (2) the computation is quite simple, (3) the required memory is rather small. This paper presents that our compression algorithm can be utilized for image retrieval from image database without decoding the compressed data stream. The block color histogram generated from the compressed data stream is used to compute the similarity measure between two images. The experimental results show that our approaches achieve better performance than conventional methods.

A new image matching algorithm for change detection using Hilbert curve

Li Tian, Sei-ichiro Kamata, Yoshifumi Ueshige and Yoshimitsu Kuroki

Proc. of International Astronautical Congress56(B1.P.04)p.1 - 62005年10月-2005年12月 

Scopus

詳細

概要:Finding significant change in high resolution sensed image is an important task in maintaining GIS database. A class of these algorithms detects changed regions by means of edge comparisons. After extraction of feature points from a sensed image and a reference image, the feature points matching is a pivotal key in change detection. In general, given two point sets, find the minimum or maximal value of some measuring distances under the (affine) transformation. Because of the measurement errors and some outlying points, it is important that the measuring distances should be robust. Recently, a well known robust measuring distance called (partial) Hausdorff distance is widely used in feature points matching. It is more efficient than other conventional methods and has been applied in many fields. Although it is a reliable similarity measure, it is also a computational task. In this paper, we present a new algorithm using Hubert curve in order to resolve the computational complexity problem. This distance can be computed in the 1-D space rather than in the 2-D space that can greatly reduce the computational complexity. Our algorithm shows good performances for this task.

H.264/AVCの直交変換におけるパディング法

黒木 祥光;廣重徹;上繁 義史;鎌田 清一郎

情報処理学会研究報告オーディオビジュアル複合情報処理(AVM)2005(98)p.17 - 222005年10月-2005年10月 

CiNii

詳細

ISSN:09196072

概要:本論文は,任意形状画像の符号化においてしばしば用いられるパディングについて考察するものである.パディングは,一般に,任意形状領域内の画素を用いて領域外部を埋めて矩形ブロックを作成することを意味し,動画像符号化MPEG4においても,low-pass extrapolation (LPE) と称する手法が採用されている.1次元DCTに対するパディング法として,既に,ShenとLiouは,領域内画素数と符号化すべきDCT係数の個数が等しいことを保証する手法を報告している.彼等は,同時に,画像信号への適用法も2種類提案しているが,彼等の手法は1次元DCTを基本としているため,領域内画素と符号化すべきDCT係数の個数が一致しない.本論文では,Shenらの手法をH.264/AVCで用いられる整数精度DCTを用いた2次元DCTに拡張する.提案する手法では,直交変換後の符号化処理を考慮し,領域内画素数と同数のDCT係数がジグザグ順序の初めに出現するため,Shenらの手法に比べ,更なる符号量の削減が期待できる.This paper presents a new padding technique for arbitrarily shaped coding. Padding background pixels using pixels in shapes, square blocks are generated; therefore, traditional orthogonal transforms work availably. A padding technique called low-pass extrapolation (LPE) is indeed employed in MPEG-4. Shen and Liou demonstrate a sophisticated padding for 1D-DCT, which guarantees that as many high frequency DCT coefficients as background pixels become zero. In order to apply their method to 2D-DCT, they also show two solutions. However, the solutions do not accomplish the equivalence between the numbers of the high frequency coefficients and the background pixels because the proposals are merely based on the 1D-DCT. In this paper, we enhance their padding methods to the orthogonal transform used in H.264/AVC, namely integer DCT. In the proposed method, the number of the DCT coefficients to be coded is equivalent to the number of the pixels lie in the shape and is appear in the zigzag order.

An automatic image-map registration using partial Hausdorff distance

Li Tian, Sei-ichiro Kamata, Yoshifumi Ueshige and Yoshimitsu Kuroki

Proc. of IEEE International Geoscience And RemoteSensing Symposium25(6)p.3534 - 35372005年07月-

画像照合のためのヒルベルト走査距離

田 黎;鎌田 清一郎;恒吉 和幸

電気関係学会九州支部連合大会講演論文集2005(0)p.544 - 5442005年-2005年

CiNii

時空間ヒルベルト走査を用いた動画像の可逆圧縮

塚野 真司;鎌田 清一郎;上繁 義史;黒木 祥光

電気関係学会九州支部連合大会講演論文集2005(0)p.373 - 3732005年-2005年

CiNii

走査パターンに着目したカラードキュメント画像の可逆圧縮

古海 頌悟;鎌田 清一郎

電気関係学会九州支部連合大会講演論文集2005(0)p.374 - 3742005年-2005年

CiNii

DCT係数に対するベッセル分布の適合性

黒木 祥光;上繁 義史;鎌田 清一郎

電気関係学会九州支部連合大会講演論文集2005(0)p.379 - 3792005年-2005年

CiNii

Lossless Image Compression via Multi-Scanning and Adaptive Linear Prediction

Haijiang Tang, Sei-ichiro Kamata, Kazuyuki Tsuneyoshi and Masa-aki Kobayashi

Proc. of 2004 IEEE Asia-Pacific Conference on Circuits and Systems7(1)p.81 - 842004年12月-2004年12月 

Scopus

詳細

概要:This paper presents an efficient lossless compression approach for gray scale images. The main contributions are 1) we divide the input image to blocks of pixels and use different scanning methods, we select the one that minimizes the prediction error for coding; 2) a new adaptive linear prediction is applied. Experiment results show that our method outperformed JPEG-LS 2∼10% in compression efficiency. © 2004 IEEE.

空間充填曲線と画像処理応用

鎌田 清一郎

情報処理学会研究報告グラフィクスとCAD(CG)2004(121)p.25 - 302004年11月-2004年11月 

CiNii

詳細

ISSN:09196072

概要:G.ペアノ(Peano)は 1890年『平面領域内の全ての点を通過するような曲線』を発見し その存在を明らかにした. 現在 線分を単位超立方体全体へ移すこのような連続曲線は 空間充填曲線 あるいはペアノ曲線と呼ばれている.空間充填曲線の中で 応用研究の最も多い曲線はヒルベルト曲線である.例えば ヒルベルト曲線の応用としては画像圧縮 スペクトル画像分類 データベース情報検索 計算機ホログラムなど 様々な分野に及ぶ.本論文では 空間充填曲線について定義と3つの例を紹介し 次にヒルベルト曲線を中心とした画像処理への応用研究を幾つか概観する.In 1890, G.Peano found a curve which passes through all points in a space and proved its existance. So far such a countinuous curve which maps a unit interval into a unit hypercube is called a space-filling curve (SFC) or a Peano curve. There are several applications in the area of image processing, computer graphics, database retrieval, etc. Among the SFC's, the most applicable curve is a Hilbert curve. This paper describes the definition and some examples of the SFC, and then overviews image processing applications of the SFC, especially the Hilbert curve.

A multi-scan adaptive linear prediction approach for Lossless Image Compression

Haijiang Tang, Sei-ichiro Kamata and Masa-aki Kobayashi

Proc. of IEEE Int. Conf. on Systems, Man and Cybernetics34(5)p.3004 - 30092004年10月-

静止画像の動的予測符号化における演算負荷の少ない予測器の選定法

黒木 祥光;上繁 義史;鎌田 清一郎

情報処理学会研究報告オーディオビジュアル複合情報処理(AVM)2004(99)p.13 - 182004年10月-2004年10月 

CiNii

詳細

ISSN:09196072

概要:ディジタル画像データを無歪みで圧縮する手法として,一般に予測符号化が用いられる.予測符号化は符号化済みの画素を用いて符号化すべき画素を予測し,予測誤差をエントロピ符号化する手法である.線形予測における最適な予測係数は,予測誤差電力を最小化するという観点から,正規方程式により算出される.正規方程式の解は,実数で与えられるが,演算負荷の少ない予測器を実現するには,数回のシフト演算と加減算で算出可能な値,例えば±1/2,±1/4,±3/4といった値が好ましい.本研究では,線形予測に用いる画素を,JPEG,JPEG-LSの可逆モードと同様に,着目画素の近傍3画素とし,予測係数の代数和は1であるとの条件の下,数回のシフト演算と加減算のみで算出可能な,演算負荷の少ない予測器に対する予測誤差電力の定量的評価を通じ,各予測器が最適となる条件を示す.また,11種および6種類から成る予測器の組を提案し,ブロック適応予測に使用した場合の性能評価も示す.Predictive coding is generally employed in lossless coding algorithms for digitized images. Pixel values to be coded in the predictive coding schemes are predicted from already-coded adjacent pixels, and then, an entropy coder encodes the resulting prediction residuals. This paper describes an estimate and a selection method of the simple predictors in a theoretical viewpoint, which constrained as follows: (1) The predicted values are computed by weighted summation of the neighboring three pixels identical to JPEG and JPEG-LS; (2) The predicted values are obtained with a small number of shift operations, additions, and/or subtractions; (3) The predictor coefficients sum up to 1. The comparison of the total squared error between the simple predictors and the optimum predictors obtained by solving the normal equations is discussed. With the results of the theoretical analysis, we propose a six-predictor and an eleven-predictor set predictors. The performance of the proposed predictors on actual images is also presented.

複数走査を用いた自然画像の可逆圧縮方法

小林 正明, 鎌田清一郎

電子情報通信学会論文誌(D-II)J87-D-II(8)p.1603 - 16122004年08月-

複数走査を用いた自然画像の可逆圧縮法(画像・映像処理)

小林 正明;鎌田 清一郎

電子情報通信学会論文誌. D-II, 情報・システム, II-パターン処理87(8)p.1603 - 16122004年08月-2004年08月 

CiNii

詳細

ISSN:09151923

概要:画像通信,蓄積においては膨大な情報量を処理するために画像圧縮が求められる.医療分野や歴史的文化財などのディジタルアーカイビング分野で使用される自然画像は,データの精度が重要であり,可逆圧縮が要求される.白熱画像に対する圧縮方法は,JPEG-LSの標準化作業においてもいくつかの方法が提案され,その多くが予測符号化に基づいている.自然画像は2次元方向での近傍画素間に高い相関があり,また,エッジ,テクスチャ,グラデーションなどによって局所的に類似した濃度変化をする.しかし,従来手法はラスタ走査順に符号化を行っているため,十分にこれらの冗長度が削減されていないと考えられる.そこで,本論文では,正方領域単位で画像を分割し,ラスタ走査順ではなく複数の走査パターンから最適な走査を選択し,選択された走査に沿って予測式の重み係数を適応的に更新する予測符号化について検討を行った.いくつかの自然画像を用いた評価実験の結果,JPEG-LSに比べ2〜10%程度高い符号化効率が実現できることを確認した.

Efficient Updating Digital Maps by Remote Sensing Images and Their Usage for Applied Tasks

Sergey Ablameyko, Aleksandr Kryuchkov, S.Borichev, Sei-ichiro Kamata

Proc. of 24th Int. Sympo. on Space Technology and Science24(1)p.1 - 52004年06月-

北九州学術研究都市 早稲田大学大学院情報生産システム研究科メディア情報部門

岡田 稔;鎌田 清一郎

画像電子学会誌 = Imaging & Visual Computing The Journal of the Institute of Image Electronics Engineers of Japan33(2)p.289 - 2932004年03月-2004年03月 

CiNii

詳細

ISSN:02859831

A Method of Computing a Space Filling Curve for Arbitrarily Shaped Region

Sei-ichiro Hiratsuka, Yoshifumi Ueshige, Sei-ichiro Kamata

Proc. of Int. Workshop on Advanced Image Technology7(1)p.181 - 1842004年01月-

A modified method of adaptive space-filling coding

Yoshifumi Ueshige, Sei-ichiro Hiratsuka, Sei-ichiro Kamata

Proc. of Int. Workshop on Advanced Image Technology7(1)p.177 - 1802004年01月-

複数走査による可逆画像圧縮法

唐 海生;鎌田 清一郎;小林 正明

映像情報メディア学会技術報告27(72)p.77 - 822003年12月-2003年12月 

CiNii

詳細

ISSN:13426893

A new probability density function of DCT coefficients

Yoshimitsu Kuroki, Yoshifumi Ueshige and Sei-ichiro Kamata

Proc. of 3rd Int. Workshop on Spectral Methods and Multirate Signal Processing3(1)p.129 - 1332003年09月-

カラードキュメント画像の可逆圧縮法

小林 正明, 鎌田清一郎

電子情報通信学会論文誌(D-II)J85-D-II(4)p.584 - 5932002年04月-

カラー静止画像の高速可逆圧縮方法

小林 正明;鎌田 清一郎

画像電子学会誌31(5)p.778 - 7862002年-2002年

CiNii

詳細

ISSN:0285-9831

概要:RGBカラー静止画像をR-,G-,B-の三つの色プレーンに分けた場合,各色プレーンごとの画像は高い相関を持つことが知られている.また,各画像は局所領域ごとに異なる性質(コンテクスト)を持つことも知られている.本論文では,これらの性質を利用することにより画像の持つ冗長度を除去しRGBカラー静止画像の可逆圧縮を行う方法について提案する.色プレーンごとに予測変換を行った予測誤差データに対して色プレーン間の相関を利用して予測誤差の色差成分を生成し,生成された予測誤差の色差成分を局所的な性質を利用してコンテクストごとに分離し,分離されたコンテクストごとに符号化を行う.従来手法との比較実験から,LOCO-I, CALICに比べ符号化効率が14%,13%程度改善され,CREWに対して同等以上の符号化効率を実現できることを確認した.また,処理時間はこれらの従来手法より高速に実現できることを確認した.

Lossless Compression for Compound Color Document Images

Masa-aki Kobayashi, Sei-ichiro Hiratsuka and Sei-ichiro Kamata

Proc. of the 21st Int. Display Research Conf.21(2)p.1525 - 15282001年10月-

Color thinning with applications to biomedical images

Alexandr Nedzved, Yurii Ilyich, Sergey Ablameyko and Sei-ichiro Kamata

Proc. of 9th Int. Conf. Computer Analysis of Images and Patterns9(1)p.256 - 2632001年09月-

An efficient correlation computation method for binary images based on matrix factorisation

R. Bogush, S. Maltsev, Sergey Ablameyko, Sei-ichi Uchida, and Sei-ichiro Kamata

Proc. of 6th International Conference on Document Analysis and Recognition6(1)p.312 - 3162001年09月-

Correlation binary image processing based on matrix representation

R.Bogush, S.Maltsev, Sergey Ablameyko, Sei-ichiro Kamata

Proc. of 6th Int. Conf. on Pattern Recognition and Information Processing6(1)p.87 - 942001年05月-

N-dimensional Hilbert scanned hierachical histogram representation for cluster analysis

Sei-ichiro Kamata and Sergey Ablameyko

Proc. of 6th Int. Conf. on Pattern Recognition and Information Processing6(1)p.113 - 1202001年05月-

Region-Based Image Coding with Multiple Algorithms

Maria Petrou, Peixin Hou, Sei-ichiro Kamata and Craig Ian Underwood

IEEE Trans. on Geoscience and Remote Sensing39(3)p.562 - 5702001年03月-

時空間ヒルベルト走査を用いた適応的線形予測符号化による動画像可逆圧縮

坂東 幸浩, 横山 貴政, 鎌田 清一郎

映像情報メディア学会誌55(3)p.447 - 4542001年03月-

N次元空間における一般化ヒルベルト走査の一計算法

坂東 幸浩, 鎌田清一郎

電子情報通信学会論文誌(A)J83-A(12)p.1368 - 13812000年12月-

Region-based scanning for image compression

Sei-ichiro Kamata and Yusuke Hayashi

Proc. of 2000 IEEE Int. Conf. on Image Processing7(1)p.895 - 8982000年09月-

An Address Generator for an N-Dimensional Pseudo-Hilbert Scan in a Hyper-rectangular Parallelepiped Region

Yukihiro Bandoh and Sei-ichiro Kamata

Proc. of 2000 IEEE Int. Conf. on Image Processing7(1)p.737 - 7402000年09月-

顔認識のためのK3-スパースグラフ・畳み込みネットワーク

呉仁傑、鎌田清一郎

査読有り1(1)p.174 - 1792018年11月-2018年11月 

DOI

詳細

掲載種別:研究論文(国際会議プロシーディングス)

3次元ボリュームデータに対するGPUフレンドリーな近似バイラテラルフィルタ

矢野光一、杉本憲治郎、鎌田清一郎

Proc.2018 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC)査読有り1(1)p.2054 - 20582018年11月-2018年11月 

DOI

詳細

掲載種別:研究論文(国際会議プロシーディングス)

DCT-1を使った短窓を持ったガウスフィルタの高速化

矢野貴大、杉本憲治郎、黒木祥光、鎌田清一郎

Proc.2018 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC)査読有り1(1)p.129 - 1322018年11月-2018年11月 

DOI

詳細

掲載種別:研究論文(国際会議プロシーディングス)

病状重症度による糖尿病網膜症のための眼底画像識別

阪口愛紀、鎌田清一郎

Proc.9th International Conference on Biomedical Engineering and Technology (ICBET 2019)査読有り1(1)p.1 - 62019年03月-2019年03月 

詳細

掲載種別:研究論文(国際会議プロシーディングス)

医用カラー画像に対するPCAによるガイドバイラテラルフィルタ

鹿毛俊喜、杉本憲治郎、鎌田清一郎

Proc.9th International Conference on Biomedical Engineering and Technology (ICBET 2019)査読有り1(1)p.1 - 62019年03月-2019年03月 

詳細

掲載種別:研究論文(国際会議プロシーディングス)

画像トラッキングのためのランキングによるアテンションアプローチ

ペン・シェンフイ、鎌田清一郎、ブレッコン・トビー

Proc.IEEE international Conference on Image Processing (ICIP2019)査読有り2019年09月-2019年09月 

詳細

掲載種別:研究論文(国際会議プロシーディングス)

3次元フラクタル次元複雑性マップによるADHDの構造MRI画像の分類

ワン・ティアンイー、鎌田清一郎

Proc.IEEE international Conference on Image Processing (ICIP2019)査読有りp.1 - 62019年09月-2019年09月 

詳細

掲載種別:研究論文(国際会議プロシーディングス)

SVDとタイリング法を用いた200FPSをもつ定数時間バイラテラルフィルタ

杉本憲治郎、福嶋慶繁、鎌田清一郎

Proc.IEEE international Conference on Image Processing (ICIP2019)査読有り2019年09月-2019年09月 

詳細

掲載種別:研究論文(国際会議プロシーディングス)

書籍等出版物

画像処理−画像表現・圧縮・フラクタル−

鎌田清一郎

サイエンス社2003年 03月-

詳細

ISBN:4-7819-1029-7

特許

整理番号:327

画像符号化装置及び画像符号化方法(日本)

鎌田 清一郎

特願2003-276898、特開2005- 39723

整理番号:455

画像検索装置及び画像検索方法(日本)

鎌田 清一郎

特願2004-329564、特開2006-139608

整理番号:456

ヒストグラム近似復元装置及びヒストグラム近似復元方法、並びに画像検索装置及び画像検索方法(日本, PCT, アメリカ)

鎌田 清一郎

特願2004-329231、特開2006-139579、特許第4575751号

整理番号:457

硬貨識別装置および硬貨識別方法(日本)

鎌田 清一郎, 唐 海江

特願2004-358839、特開2006-164192、特許第4444089号

整理番号:482

画像検索方法及び画像検索装置ならびにプログラム(日本)

鎌田 清一郎

特願2005- 90114、特開2006-268799

整理番号:483

マッチング方法およびマッチング装置ならびにプログラム(日本)

鎌田 清一郎

特願2005-078260、特開2006-260311、特許第4570995号

整理番号:598

予測器及び画像符号化器(日本)

鎌田 清一郎, 唐 海江

特願2006- 83873、特開2007-259306

整理番号:599

画像変換パラメータ演算装置及び画像変換パラメータ演算方法(日本)

鎌田 清一郎, 田 黎

特願2006- 83878、特開2007-257530

整理番号:857

指紋特異点抽出装置及び指紋特異点抽出方法(日本)

鎌田 清一郎, 許 霄

特願2008-090359、特開2008-276758

整理番号:917

認証装置及び撮影装置(日本)

鎌田 清一郎

特願2008-236370、特開2010- 72693

整理番号:994

粒状物品種検査装置(日本)

鎌田 清一郎

特願2008-292824、特開2010-117331、特許第5163985号

整理番号:1138

画像識別装置及びプログラム(日本)

鎌田 清一郎, 杉本 憲治郎

特願2010-246946、特開2012- 98191、特許第5582610号

整理番号:1139

相同性検索装置及びプログラム(日本)

唐 海江, 鎌田 清一郎

特願2010-279658、特開2012-128672

整理番号:1221

暗号処理システム、暗号化装置、復合装置、及びプログラム、並びに暗号処理方法(日本)

鎌田 清一郎

特願2011-268023、特開2012-177893

整理番号:1522

画像フィルタ演算装置及びガウシアン・カーネル演算装置並びにプログラム(日本)

鎌田 清一郎, 杉本 憲治郎

特願2014-263701、特開2016-122430

外部研究資金

科学研究費採択状況

研究種別:

空間充填曲線による画像認識・検索に関する研究

2012年-0月-2015年-0月

配分額:¥5330000

研究種別:

空間充填曲線による画像検索および暗号に関する研究

配分額:¥4550000

研究種別:

インターネット上でのプライバシ保護が可能なオンライン生体認証システムの構築

配分額:¥3630000

研究種別:

乳腺用3次元超音波画像取得装置に関する基礎研究

配分額:¥3200000

研究種別:

物理モデル駆動によるノンフォトリアリスティック画像創成と知的符号化

配分額:¥13550000

研究種別:

電子透かし技術を応用した画像情報管理システムの開発

配分額:¥8500000

研究種別:

空間充填走査を用いたディジタル画像システムの構築

配分額:¥3200000

研究種別:

2次元パターンのワープ法に関する研究

配分額:¥3200000

研究種別:

ヒルベルト走査を利用したディジタル画像システムの構築

配分額:¥2100000

研究種別:

仮想空間構築における実動画像情報を用いたモデル自動生成の高品位実時間処理の研究

配分額:¥12200000

研究種別:

ヒルベルト走査による画像表示装置の開発

配分額:¥1000000

研究種別:

ヒルベルト曲線による走査アドレス・ジェネレータの開発

配分額:¥900000

研究種別:

SD式モデルに基づく英日会話文データベースからの背景知識と意志や意図情報の抽出

配分額:¥1300000

研究種別:

SD式モデルに基づく英日会話文データベースからの背景知識と意志や意図情報の抽出

配分額:¥1900000

研究種別:

SD式モデルに基づく英日会話文デ-タベ-スからの背景知識と意志や意図情報の抽出

配分額:¥2500000

研究種別:

英語・日本語会話文のデ-タベ-ス化と意味情報による検索利用システムの研究

配分額:¥800000

研究種別:

リモ-トセンシング画像を対象とした画像理解に基づく知的インタ-フェイスの研究

配分額:¥1900000

研究種別:基盤研究(C)

ビジュアルビッグデータの高速画像検索・認識に関する研究

2015年-2017年

研究分野:知覚情報処理

配分額:¥4680000

研究種別:

スパースグラフ・ニューラルネットワークによる画像認識および応用

2018年-0月-2021年-0月

配分額:¥4420000

研究種別:

ビジュアルビッグデータの高速画像検索・認識に関する研究

2015年-0月-2018年-0月

配分額:¥4680000

研究種別:

多次元画像のスパースフーリエ変換と深層学習の高速化

2016年-0月-2018年-0月

配分額:¥3900000

研究資金の受入れ状況

実施形態:共同研究

高速パターンマッチング応用2007年-2011年

実施形態:共同研究

ハイブリッド画像圧縮システム2003年-2007年

実施形態:受託教育

カラー静止画像圧縮に関する研究開発2000年-2003年

学内研究制度

特定課題研究

幾何イメージ思考のための支援ツールの開発

2005年度

研究成果概要:学生の大半が空間次元拡張をなかなかイメージできないのが現状である。画像処理研究の道具として使っている空間充填曲線(space filling curve、日本ではペアノ曲線とよく呼ばれる)は、曲線による次元拡張の例を示す最もよい道...学生の大半が空間次元拡張をなかなかイメージできないのが現状である。画像処理研究の道具として使っている空間充填曲線(space filling curve、日本ではペアノ曲線とよく呼ばれる)は、曲線による次元拡張の例を示す最もよい道具である。これは、簡単なパターンの回転、縮小、平行移動などの写像を使って興味深い幾何模様が描けるという特長がある。このような現状から、約10年前から自らが翻訳した「空間充填曲線とフラクタル」のテキスト(シュプリンガー・フェアラークから1998年に出版)を使用して、学生に情報数学教育を行っている。図を使うことはイメージを膨らます最もよい方法だと言われるが、空間次元の拡張も同様に思考能力を向上させる重要なテーマである。本研究では、空間充填曲線の生成を基礎として幾何イメージの思考能力を高めるための支援ツールの開発を行った。ペアノ、ヒルベルト、シェルピンスキー、ムーアなどの大数学者が発見した空間充填曲線を使って、2次元平面での生成方法をグラフィックスでわかりやすい教材コンテンツを作成した。次に、これらの曲線を3次元に拡張し、3次元空間における様々な空間充填曲線の模型を作り、3次元立体視表示ソフトとコンピュータグラフィックスを使って2次元から3次元への幾何イメージを思考訓練させるコンテンツを作成した。さらに、n次元空間(n≧4)に拡張したときに幾何図形の認識理解を助ける表示方法を検討した。ただし、表示装置は2次元あるいは3次元の表示系であるため、表示方法は、幾何図形を2次元平面あるいは3次元空間への投影、あるいは、2次元あるいは3次元断面の生成を基本としたものである。

複数走査を用いた画像圧縮および画像検索技術の開発

2005年度

研究成果概要:遠隔監視,遠隔医療などにおいて,画像の圧縮伝送・蓄積に関する研究開発が社会的なニーズとともに益々重要となっている.例えば,遠隔監視では,セキュリティ対象画像を24時間圧縮伝送し,記憶装置に大量の圧縮データを蓄積させている.しかし,...遠隔監視,遠隔医療などにおいて,画像の圧縮伝送・蓄積に関する研究開発が社会的なニーズとともに益々重要となっている.例えば,遠隔監視では,セキュリティ対象画像を24時間圧縮伝送し,記憶装置に大量の圧縮データを蓄積させている.しかし,現状では,国際標準化方式である Motion JPEG,あるいは MPEG などに頼っており,(a)DCT,ウェーブレット変換等の画像変換を利用した低域周波数通過型の圧縮方式であるために画質がぼけてしまうという問題,(b)検索したい人物画像を蓄積データの中から高速に検索できない問題,(c)圧縮データフォーマットが公開されているための情報漏洩の問題,など大きな問題が指摘されている.本研究では,このような遠隔監視あるいは遠隔医療等への応用を目的として,複数の画像走査を利用した,画像変換を行わない高効率な画像圧縮技術と,蓄積した圧縮データから高速に類似画像等の検索を行う画像検索技術を開発した.具体的な成果は,次の通りである.(1) 画像上の各部分領域を圧縮に適した走査方法で走査し,エッジを保存するような画質明瞭性に優れた新たな圧縮技術を開発し,前述の国際標準化方式と比較した定量的評価によって,本開発手法の画質明瞭性を明らかにした. 本手法は,独自圧縮方式であるため,圧縮データフォーマットを公開しなければ情報漏洩の問題は解決できる.(2) 蓄積した膨大な圧縮データからカラーヒストグラムに基づいた画像照合を行い,人物等の類似画像を高速に検索する画像検索技術を開発し,すべての画像の画素単位での類似検索について,その検索効率を明らかにした.(3) 画像圧縮技術および画像検索技術を組み合わせたシステム全体をFPGA上に作成し,本技術のハードウェア規模を明らかにした.

空間充填曲線による画像情報のモデル化と画像圧縮・検索への応用

2007年度

研究成果概要: 生活・安全などにおいて、画像の圧縮伝送・蓄積に関する研究開発が社会的なニーズとともに益々重要となっている。例えば、遠隔監視では、セキュリティ対象画像を24時間圧縮伝送し、記憶装置に大量の圧縮データを蓄積させている。しかし、これに... 生活・安全などにおいて、画像の圧縮伝送・蓄積に関する研究開発が社会的なニーズとともに益々重要となっている。例えば、遠隔監視では、セキュリティ対象画像を24時間圧縮伝送し、記憶装置に大量の圧縮データを蓄積させている。しかし、これに関わる研究開発者は国際標準化方式である MPEG (例えばMPEG4 AVCなど)などを利用しているが、(a)離散コサイン変換(DCT)、動き補償などを利用した低域周波数通過型の圧縮方式であるために画質がぼける、(b)検索したい人物画像をMPEG圧縮による蓄積データの中から高速に検索できない、(c)圧縮データフォーマットが公開されているための情報漏洩、などの問題が指摘されている。本研究では、これらの問題を解決すべく、MPEGとは異なる、空間充填曲線を用いた画像圧縮方式を検討し、高速検索の実現も含めて、画像圧縮技術の高画質化、高解像度化、高速検索が可能なものを実現することを目的とする。本研究によって、具体的に次のことが明らかになった。(1)画像上の各部分領域を任意形状とし、空間充填曲線を用いて圧縮に適した画像情報のモデル化を行った。これは、空間充填曲線を複数組み合わせることによって、輪郭情報保存性のみならず一般人の主観的画質評価に優れた新たな圧縮手法を開発した。本手法は、独自圧縮方式であるため、圧縮データフォーマットを公開しなければ(c)の情報漏洩の問題は一部解決できるが、セキュリティ面でより頑健なものにするために走査法に着目した暗号化についても検討した。(2)蓄積した膨大な圧縮データからの特徴抽出および画像照合を行い、人物などの類似画像を高速に検索する画像検索手法を開発した。色情報等に基づくヒストグラムによる検索によって、数十倍の高速化が可能となった。なお、MPEGでは圧縮データが周波数領域での表現になっており、DCTなどの逆変換を行う必要があり、高速検索が困難である。

空間充填曲線による画像圧縮・検索に関する研究

2008年度

研究成果概要:本研究では、これまでに構築してきた空間重点曲線を用いた画像圧縮法に対して、計算効率のよい画像検索を行うことを目的として検索手法を開発した。これまでの画像圧縮法によって圧縮されたデータ(以下、圧縮データ)から探索対象物(例えば人物、...本研究では、これまでに構築してきた空間重点曲線を用いた画像圧縮法に対して、計算効率のよい画像検索を行うことを目的として検索手法を開発した。これまでの画像圧縮法によって圧縮されたデータ(以下、圧縮データ)から探索対象物(例えば人物、車など)を何らかの特徴で表現できる高速画像検索の実現を目指して、空間充填曲線を用いた画像検索方式を確立している。この特徴は、圧縮データから空間情報が抽出できるので、様々な記述子を用いて表現することができる。本研究ではフーリエ記述子とPCA(Principal Component Analysis)を用いた新たな方式を提案した。画像MPEG7の形状データベースを用いて、MPEG7に採用されているCurvature Scale Space記述子、Wavelet記述子等の従来手法との類似検索などの比較を行い、その検索効率をRecall-Precision指標等により明らかにした。次に、画像上の各局所領域を予測符号化に適した画像情報のコンテキストモデル化を考察した。コントラストがありエッジが保存できれば、その画質はくっきりした明瞭性に優れたものとなる。すなわち、高コントラスト、エッジ保存型であり、平坦な部分は滑らかに変化するようなモデル化を行うことが必要となる。各局所領域に対して、予め与えられた単純予測器を適応的に選択することによって圧縮効率を上げるものである。これにより、国際標準化方式と比較した評価によって、本手法の有効性を明らかにした。画像検索手法は、蓄積した膨大な圧縮データに対して検索を行うため、特に高速化が要求される。このため、本研究では、FPGAを利用して、高速アルゴリズムを実現し、ソフトウェア性能よりどの程度高速化されるかを明らかにした。

コミュニケーション・ロボット「二宮くん」の研究開発

2010年度

研究成果概要:2009年6月図書館、介護施設等の本の読み聞かせを代行あるいは支援してくれる、本の読み聞かせロボット(ブック・リーダ・ロボット)「二宮くん」(にのみや・くん)を試作した。「二宮くん」は、これまでにない、前に置かれた本を読みあげてく...2009年6月図書館、介護施設等の本の読み聞かせを代行あるいは支援してくれる、本の読み聞かせロボット(ブック・リーダ・ロボット)「二宮くん」(にのみや・くん)を試作した。「二宮くん」は、これまでにない、前に置かれた本を読みあげてくれる、親しみのある人型ロボットである。これまで「二宮くん」の展示会出展などを通して、用途が本の読み聞かせのみではなく、コミュニケーション・ロボットとしての役割を果たすともっと面白いロボットになるのではないか、という意見を多数頂いた。本研究では、このようなニーズに応えるためにコミュニケーション・ロボットとしての機能を実現することを目的とする。2010年6月にも、研究開発を行ってきた本の読み聞かせロボット「二宮くん」を展示会に出展した際に、歌を歌わせること、対話ができること、という要望が強かったが、今年度はまず、本を読むだけでなく、楽譜を読ませる機能を追加して、歌を歌わせることを試みた。画像処理により、五線譜部と歌部の区別を行い、ドレミの音符の認識を行った。その結果、音符は90%以上の精度で読みとることができた。歌詞の認識も行う必要があり、音符と連動できるようにした。次に、家庭という空間を想定した場合に、本などをベースにしたコミュニケーションを行うことを念頭に、顔認識機能の追加を行った。その結果、複数走査によるパターンコード表現に基づいた顔認識アルゴリズムを確立した。公開されているORLデータ、FERETデータなどの様々な画像データに対して顔認識のアルゴリズムを評価した結果、従来手法LBP、LDP等に対して数%の認識率向上が得られた。また、音声認識方式の開発、対話テキスト解析方式の導入を検討しているが、今後コミュニケーション・ロボットのプロトタイプシステムを構築し、人間と比較した評価を行うことによって改善を図る。

読み聞かせロボットの高精度画像認識に関する研究

2013年度

研究成果概要:二宮金治郎をイメージしてデザインを考え、薪のかわりにパソコンを背負うというコンセプトにより、感情を込めて本の読み聞かせを行うロボット「二宮くん」を開発している。「二宮くん」は、目にカメラを搭載しているため、一般照明の影響、本のフォ...二宮金治郎をイメージしてデザインを考え、薪のかわりにパソコンを背負うというコンセプトにより、感情を込めて本の読み聞かせを行うロボット「二宮くん」を開発している。「二宮くん」は、目にカメラを搭載しているため、一般照明の影響、本のフォントの多様性により文字認識がうまくいかず、認識率があまり良くないという問題があった。本研究では、「二宮くん」の高精度認識を実現することを目的としてこれらの問題を解決するため、次のような研究成果を得た。(1)照明等による影響による画像の変動を吸収するために、カラーコントラスト補正法を開発した。これはRGBカラー空間の輝度情報に着目し、大域処理と局所処理の2段階の階調マッピングを実現する階層的ヒストグラム表現を用いるものである。NASA Langley Research Centerのデータセットを使った評価実験の結果、提案手法は代表的な従来手法よりPSNRによる評価において約10%、エントロピー評価においても同等の画質改善ができることを確認した。(2)本などの画像から文字部分を高精度に抽出し認識する必要がある。本研究では、Gaborフィルタを拡張し、新たに局所Gaborテクスチャ特徴を提案した。これは画像中における局所領域のコーナー点の周辺の色とテクスチャを用いた特徴量である。色情報はHSV色空間から得られる画像中の色彩と色相の情報から抽出し、テクスチャ情報は明度情報からガボールフィルターを用いて抽出する。また、テクスチャ情報は離散フーリエ変換を適用し、画像の回転に対して不変の値にすることで、画像中のテキストにおいて頑健な特徴量を抽出した。文字認識・文書理解国際会議にて一般公開されている画像データセットを用いてテキスト領域の検出性能を評価した場合、提案手法の性能は他手法の性能よりも再現率において約10%、F値において数パーセント高いことを示した。これらの研究成果により、独自に収集したデータセットを使った認識実験の結果、従来の文字認識法に比べて数パーセントの認識精度の向上を図ることができた。

悪天候下の深層学習による道路環境の画像認識技術に関する研究

2017年度

研究成果概要:高齢者用の運転支援などを実現するため、エコ電気自動車の周囲環境認識技術について研究している。本研究では、まず車体周囲の3次元環境情報を取得する上で、ステレオビジョンの計測精度を高めることを検討した。これまで奥行き情報を取得する上で...高齢者用の運転支援などを実現するため、エコ電気自動車の周囲環境認識技術について研究している。本研究では、まず車体周囲の3次元環境情報を取得する上で、ステレオビジョンの計測精度を高めることを検討した。これまで奥行き情報を取得する上で物体間の境界がぼけてしまうという問題があったが、これを上下左右に延びたクロス形状のフィルタを設計し、これを2段階で適用する新たな方法論を確立した。次に、悪天候下や夜間走行を行う上で近赤外カメラも利用して自動車周辺の状況把握に深層学習を利用することを検討した。その結果、エッジがクリアに再現され、安定したカラー領域が生成される新たな近赤外カラー化モデルSNetを開発した。

多次元医用画像の高効率解析アルゴリズムに関する研究

2017年度

研究成果概要:脳科学分野などでは画像データの高密度化・高次元化が顕著であり、ハードウェアの進歩だけでは実時間処理が困難な現状がある。本研究は、多次元画像を対象とし、高速フィルタ処理、高速特徴抽出法などのアルゴリズムの開発を行った。まず画像処理に...脳科学分野などでは画像データの高密度化・高次元化が顕著であり、ハードウェアの進歩だけでは実時間処理が困難な現状がある。本研究は、多次元画像を対象とし、高速フィルタ処理、高速特徴抽出法などのアルゴリズムの開発を行った。まず画像処理における基本的処理として用いられるバイラテラルフィルタ設計に関して、特異値分解を用いた効率的な手法を提案した。従来法と同等の近似精度を維持し、およそ5割少ない計算量で実現した。次に、病理組織の染色画像を対象として、Secant Normal Votingに基づく細胞のセグメンテーション手法を考案した。本手法は、従来問題となっている重複領域などに対しても良好な細胞核検出が実現できることがわかった。

多次元高精細バイオ画像データの高速画像処理アルゴリズムに関する研究

2016年度

研究成果概要:多次元高精細バイオ画像データにおけるアライメントを含む高速3次元再構成法について、アルゴリズム面での改善を含めて実時間の処理が可能となる手法を研究開発した。まず核磁気共鳴画像法(Magnetic Resonance Imaging...多次元高精細バイオ画像データにおけるアライメントを含む高速3次元再構成法について、アルゴリズム面での改善を含めて実時間の処理が可能となる手法を研究開発した。まず核磁気共鳴画像法(Magnetic Resonance Imaging)による3次元ボリュームデータのアライメントに対して、高速かつ高精度な画像処理アルゴリズムを確立した。従来手法と比較して数%~10%の高速化を図ることができた。また、肝臓などの細胞画像の3次元アライメントによる画像再構成アルゴリズムを確立し、従来手法に比べて高精度化および高速化を図ることができた。

ヒルベルト曲線を利用した多次元データ管理システム

2003年度

研究成果概要: 本研究では,画像表現形式のまま移動や拡大等の処理ができ,範囲検索が高速にできるような多次元データ管理手法として、ヒルベルト曲線を用いて線形4分割木構造表現を行う手法を検討した.これは,空間内においてデータの有無を1,0の2値デー... 本研究では,画像表現形式のまま移動や拡大等の処理ができ,範囲検索が高速にできるような多次元データ管理手法として、ヒルベルト曲線を用いて線形4分割木構造表現を行う手法を検討した.これは,空間内においてデータの有無を1,0の2値データで表現し,データがある領域をアドレスと領域サイズの組を一つのデータとして表し,そのデータの組をヒルベルト曲線が辿る順に並べて、多次元データを表現する手法である.ヒルベルト曲線はデータ間の近傍保存性が良く,データのクラスタ抽出ができるという特徴を持つ。範囲検索を高速に行えるMD木による範囲検索では,検索する領域を分割し,それぞれの検索領域を木構造の根から順にノードを辿って検索する.これに対し本手法は,1次元データで管理していることから,検索する領域を1次元データの先頭から検索領域が見つかるまで探索し,検索できれば,分割された他の検索領域は1次元配列上で近傍に存在するはずであるため,その近傍を検索するだけでよい.また,本手法は領域をアドレスで管理していることから,検索はアドレスの比較により簡単に行うことができる.本研究では,ヒルベルト曲線を用いたデータ表現から,表現形式を変えずに,移動,拡大・縮小,回転,抜き取り,共有,交わり,投影,制約,結合,差といった10種の処理を行う手法について述べた.2次元データである2値画像を用いて,本手法の有効性を確認した.また、MD木によるデータ表現において範囲検索の検索速度を比較し、本手法はMD木に対して20~30%検索効率がよいことがわかった.さらに,本手法の応用として,実時間描画のためにデータベースから高速な検索が必要である仮想都市空間のデータ管理への本手法の適用可能性について検討した.今後は,高速に範囲検索が行え,データ表現のまま種々の演算処理ができることから,仮想都市空間のデータ管理への適用をさらに検討する予定である.

複数の空間充填走査を用いた画像圧縮技術の開発

2004年度

研究成果概要:本研究では,予めテンプレートとして与えられた複数の空間充填曲線を利用して新たな画像圧縮技術の開発を目的として研究を行った.具体的な成果は次の通りである.(1)画像上の複数の特徴的領域(例えば,平坦な領域)を充填しながら連続的に走査...本研究では,予めテンプレートとして与えられた複数の空間充填曲線を利用して新たな画像圧縮技術の開発を目的として研究を行った.具体的な成果は次の通りである.(1)画像上の複数の特徴的領域(例えば,平坦な領域)を充填しながら連続的に走査するような新たな空間充填走査方法を検討し,必要なテンプレートについて圧縮効率と計算量の関係を導いた。(2)静止画像および動画像を対象とし,木構造表現に基づいた画像の階層的分割を行い,個々の分割領域に複数の空間充填走査を適用して画像を表現し,画像記述に適した効率的な画像圧縮技術を開発した.これは,予測係数を逐次更新する予測符号化方式ALCM(Activity Level Classification Model)を改良し,コンテキストモデル符号化と組み合わせた予測符号化方式である.複数のカラー標準画像を用いた圧縮実験の結果、提案手法は国際標準化方式JPEG2000およびJPEG-LSを越える圧縮性能を得た.また,提案手法は,符号化・復号化処理において国際標準化方式に比べて比較的簡単な処理で済むため,ハードウェア規模がこれらの国際標準化方式に比べて小さく実現できることを確認した.(3)近年標準化されたJPEG2000や次世代のMPEG7などに代表される画像圧縮技術では,単なる画像圧縮のみの手法ではなく,画像の検索,内容抽出なども効率よく行うことができる技術が求められているが,圧縮データストリーム上での部分画像検索が行うことができない.そこで,本研究では圧縮データストリーム上で部分画像検索を行うことができる圧縮方式を検討した.今後は,画像圧縮技術および画像検索技術を組み合わせたシステム全体を構築し,本技術の検索性能およびハードウェア規模を明らかにする予定である.

現在担当している科目

科目名開講学部・研究科開講年度学期
画像情報処理大学院情報生産システム研究科2019春学期
イメージメディア研究(修士) 春大学院情報生産システム研究科2019春学期
イメージメディア研究(修士) 秋大学院情報生産システム研究科2019秋学期
イメージメディア演習A大学院情報生産システム研究科2019秋学期
イメージメディア演習B大学院情報生産システム研究科2019春学期
イメージメディア演習C大学院情報生産システム研究科2019春学期
イメージメディア演習D大学院情報生産システム研究科2019秋学期
マルチメディア特論大学院情報生産システム研究科2019秋学期
パタン認識大学院情報生産システム研究科2019秋学期
情報セキュリティ論大学院情報生産システム研究科2019秋学期
イメージメディア研究(博士) 春大学院情報生産システム研究科2019春学期
イメージメディア研究(博士) 秋大学院情報生産システム研究科2019秋学期

教育内容・方法の工夫

九州大学全学共通教育自己点検・評価報告書

2000年04月-2000年09月

詳細

概要:全学共通教育科目「情報処理基礎演習」の学生による授業評価(アンケート調査)において、授業改善の結果、良い評価を受けた。特に、「学問への理解の広がり、力がついた」および「勉学に対する意欲がわいた」という学生が多く、アンケート調査を行った当該科目の13クラス中上位の評価を得た。

マルチメディア機器を活用した授業方法改善

2000年04月-

詳細

概要:情報処理基礎演習、画像インタフェース特論、画像情報処理の講義では、パーソナルコンピュータとPCプロジェクタおよびビデオを使って、できるだけ視聴覚的に改善を行った。

作成した教科書・教材・参考書

教科書作成

2003年03月

詳細

概要:題目は画像処理-画像表現・圧縮・フラクタル-

参考書作成

1998年12月

詳細

概要:題目は空間充填曲線とフラクタル