Image Processing Laboratory

Ryuichi Oka Professor	Keitaro Naruse Assistant Professor	Yu Nakajima Research Associate

First and foremost, the Image Processing Laboratory engages in research and development of image-based pattern recognition including some areas of Artificial Intelligence, database organisation, retrieval, and robotics. More specifically, as can be seen from the background information and the recent research publications of the members of the laboratory, 2-dimensional Continuous Dynamic Programming for non-linear image registration without mode, object tracking in a time-varying image, image retrieval from video data, character recognition of telop in a video image and extracion of the Internet community are our current focus. Related to the recent progress of the web, a huge amount of multimedia data without index becomes available to store in our PC's. However, no sophistcated methodology to manage such data has been developed so far. So that we are solicited to attach an index to each data. Our research aim is to develop algorithms to realize automatic annotation to real word data for integrated retrieval of multimedia information. The algorithms include self-organisation and transformation among representation of multimedia and feature extraction and recognition of real data. Real world data includes video, still image, speech, music, sound, and text each of which has not been indexed by labels. An example of multimedia integration retrieval called "CrossMediator" developed by ten-year project (1992-2002) RWC of METI (Japan). Some parts of CrossMediator have been in the commercial market by through a private company. Prof. R. Oka was a chief of group which developed "CrossMediator". Our laboratory will pursue to develop more sophisticated functions which might reveal a new generation of the Internet. Speech retreival system used in CrossMediator is used for developing new types of commercial products. At the present time we use two kinds of index for searching video data. One is used to obtain the target media from Youtube by using index. The other one is used for searching parts of the target data which are not indexed. One of the recent progress is developing a set of algorithms for spotting recognition and tracking objects in a time-varying image. These algorithms enable to realize automatic anonotation of video image capturing moving objects. Another recent progress is developing a spotting algorithm for song wave retrieval. A large amount of song wave data is available for anyone along with the progress of storage hardware of music data. But the progress of technology for accessing the song wave data seems quite slow so that we must use conventional and not sophistcated tools. Our new software can provide users a convenient way to find and rerieve the song wave data by singing an arbitrary part of it which he/she wants to hear. The Internet is regarded as a network composed of virtual communities. Visualization of the community is becoming an important research target. Our lab is developing a sophistcated algorithm based on so-called Associated hyper-linked word space (A-space). We use the algorithm to detect clusters each of which corresponds a virtual community. Visualization of each virtual community is realized by showing the content extracted from the web pages included the community.

Refereed Proceeding Papers

[naruse-01:2007]	Masao Kubo, Keitaro Naruse, Hiroshi Sato, and Takashi Matsubara. The Possibility of an Epidemic Meme Analogy for Web Community Population Analysis. Proceedings of the 8th International Conference on Intelligent Data Engineering aud Automated Learning - IDEAL 2007, December 2007. LNCS, 4881:1073-1080, Springer, 2008.
	In this paper, the mechanism of topic spreading in Web BBS is modeled by the one of epidemics. The behavior of the model quite coincides with the one of actual data in Web. As the application of the modeling, it allows us to estimate the number of users of Web BBS.
[naruse-02:2007]	Keitaro Naruse and Masao Kubo. User Social Interaction in Web BBS as Sandpile. In 3rd International Nonlinear Sciences Conference, 2008.
	This paper presents a user social interaction model inspired by a sandpile phenomenon in the real world. The developed model fits well in the logs in Web BBS, and it provides us an estimation method of the number of users in each of Web BBS.
[oka-01:2007]	Takashi Kawai and Ryuichi Oka. Segmentation-free detection the similar images with pixel coresspondence. In Proceedings of meeting on Image Recognition and Understanding, page 532. IEICE, July. 2007.
	It proposes a new method in our project though recently, there are some kinds of detection of the pattern of the image. It compares images by the accumulation distance by using each color of two images difference. In addition, the accumulation distance obtained from two images is replaced with 3D space and a similar pattern of the image is extracted from a different aspect.
[oka-02:2007]	Yuuichi Yaguchi, Yoshiyuki Watanabe, Keitaro Naruse, and Ryuichi Oka. Speech and song wave search in the Web : System Design and Imprementation. In CIT2007, pages pp.270-275. IEEE, Oct. 2007.
	We propose a song wave retrieval method. Both song wave data and a query wave for song data are transformed into phoneme sequences by frame-wise labeling of each frame feature. By applying a search algorithm, called Continuous Dynamic Programming (CDP), to these phoneme sequences, we can detect a set of similar parts in a song database, each of which is similar to a query song wave. Song retrieval rates hit 78 percent in four clauses from whole databases. Differences in each query from song wave data and speech wave is investigated.
[oka-03:2007]	Kenta Iseki and Ryuichi Oka. Extraction of Character Area from Natural Scene Images. In Proceedings of meeting on Image Recognition and Understanding, page 475. IEICE, July. 2007.
	To detect character regions from natural scene images is dificult because of variations in text and complex backgrounds. In this paper, a new connected component (CC) based method for color segmentation using a k-means algorithm is described. After clustering the colors on the input images, four selection rules for getting character-like regions are suggested. And the proposed method was tested on the ICDAR 2003 Robust Reading Competition data set. According to the results of the evaluation, the proposed method worked well on detecting character-like regions.

Unrefereed Papers

[naruse-03:2007]	Ide and Naruse. Quadruped Gait Generation for Humanoid Robot using Virtual Muscle Oscillation Method. In Proc. of 2008 spring JSPE Conference, 2008.
[naruse-04:2007]	Naruse and Sekii. Animation by Semi-automatic Generation of Control Point Trajectory in Spatial Key-framing. In Proc. of Robotics mechatronics conference, 2007.
[naruse-05:2007]	Naruse. Robot motion deformation by virtual muscle oscillation method. In Proc. of 2007 fall JSPE Conference, 2007.
[naruse-06:2007]	Naruse and Kubo. Analysis of Message Interaction in Web by Sand pile Model. In Proc. of SI2007 Conference, 2007.
[naruse-07:2007]	Unno and Naruse. Statistical analysis of conversation spread on the Web. In Proc. of 2008 spring JSPE Conference, 2008.
[naruse-08:2007]	Ishii and Naruse. "Investigation of color visual impression on web and psychology reection system. In Proc. of 2008 spring JSPE Conference, 2007.
[naruse-09:2007]	Anzai and Naruse. Pulse-taking method of consumer by topic extraction from blogs. In Proc. of 2008 spring JSPE Conference, 2008.
[naruse-10:2007]	Mizushima and Naruse. Prototype of education kit of a modular robot using a muscular model. In Proc. of 2008 spring JSPE Conference, 2008.
[oka-04:2007]	Hiroshi OHNISHI, Yuichi YAGUCHI, Keitaro NARUSE, and Ryuichi OKA. A Mining Method for Linked Web Pages Using Associated Keyword Space With Text Data. In Technical Report of SIG-W12, pages SIG-WI2-2007-7-4, July 2007.
	The purpose of this paper is to introduce a novel method for mining knowledge from linked Web pages. Our past research was focused on link structure and text data in Web pages separately. In this research, however, we treat link structure and text data simultaneously and equivalently, then create an affinity matrix about Web pages. Once the affinity matrix was obtained, we use Associated Keyword Space, or ASKS, method to allocate each items in N dimensional association space. Experimental results confirm that association space with link structure and text data could project textual characteristics in the space.

Grants

[naruse-11:2007]

Keitaro Naruse. Analysis of social interaction of users in Web Ministry of Education Scientific Research Fund, 2007-2009.

Academic Activities

[naruse-12:2007]	Keitaro Naruse, 2007. Registration Chair,2007 IEEE International Conference on Computer and Information Technology, (CIT2007).
[naruse-13:2007]	Keitaro Naruse, 2007. Session Organizer, SICE System Integration Division Conference.
[naruse-14:2007]	Keitaro Naruse, 2007. Paper Reviewer, 2007 IEEE International Conference on Computer and Information Technology, (CIT2007).
[naruse-15:2007]	Keitaro Naruse, 2007. Paper Reviewer, JSAI.
[naruse-16:2007]	Keitaro Naruse, 2007. Technical Program Committee, 2007 IEEE International Conference on Computer and Information Technology, (CIT2007).
[oka-05:2007]	Ryuichi Oka, May 2007. Reviewer of submitted papers.
[oka-06:2007]	Ryuichi Oka, April 2007. Senior Member of editorial committee.
[oka-07:2007]	Ryuichi Oka, May 2007. Reviewer of submitted papers, the Acoustic Society of Japan.
[oka-08:2007]	Ryuichi Oka, April 2007. Reviwer of submitted papers, the Seventh Inter. Conf. on Computer and Information Technology.

Ph.D and Other Theses

[naruse-17:2007]	Kenta Ide. Graduation Thesis: Quadruped Gait Generation for Humanoid Robot using Virtual Muscle Oscillation Method, University of Aizu, 2007. Thesis Advisor: Keitaro Naruse.
[naruse-18:2007]	Hiromi Anzai. Graduation Thesis: Topic Extraction from Blogs for Consumer Pulse-Taking, University of Aizu, 2007. Thesis Advisor: Keitaro Naruse.
[naruse-19:2007]	Sayuri Ishii. Graduation Thesis: A System Reecting Color Visual Impressions on the Web, University of Aizu, 2007. Thesis Advisor: Keitaro Naruse.
[naruse-20:2007]	Taku Unno. Graduation Thesis: Statistical analysis of conversation spread on the Web, University of Aizu, 2007. Thesis Advisor: Keitaro Naruse.
[naruse-21:2007]	Yoshiaki Mizushima. Graduation Thesis: Prototype of education kit of a modular robot using a muscular model, University of Aizu, 2007. Thesis Advisor: Keitaro Naruse.

Others

[oka-09:2007]	Ryuichi Oka, Hironobu Takahashi, Takuichi NIshimura, and Toshiro Mukai. An artcle titled Realtime Gesture-Speech Interface Realized on a Personal Computer of Notebook Size (pages: 206-215) is included in a book titled Spoken Language Systems edited by Seiichi Nakagawa and Michio Okada and Tatsuya Kawabata (Ohmsha,2005).
[oka-10:2007]	Ryuichi Oka. Telphone speech retrieval system was released to the markect by Neix Inc. The system uses the technology developed by R. Oka and Mediadrive Inc.

2009 THE UNIVERSITY OF AIZU ALL RIGHTS RESERVED