题 目：Real-time Multi-view Image Analysis and Its Application to Human Motion Sensing
主讲人：Rin-ichiro Taniguchi 日本九州大学教授
时 间：9月25日(周六)下午2：30
地 点：玉泉校区曹光彪东楼502
报告简介：
In this talk, I would like to talk about one of the hottest topics in computer vision. Computer vision is a research to develop computer/machine systems which can understand input images and which can recover 3D real world information from the images.
The difficulty in computer vision is its ill-poseness. Images, which are captured by a camera, are 2D information, and, thus, they are reduced information from a 3-D space. Mathematically, recovering 3D information from 2D data can not be solved without any ambiguities.
To overcome this problem, we have employed multi-view image analysis,where multiple cameras are employed to disambiguate the ill-posed problem. One of the biggest problems in the multi-view image analysis is its computation cost, which becomes very severe especially when real-time processing is required. Therefore, we have developed a parallel processing system, i.e., a PC cluster for vision processing, to realize real-time multi-view image analysis.
As its practical application, we have constructed a vision-based real-time human motion sensing system, where we have developed a human posture analysis algorithm from a limited number of cues, which can be robustly acquired from input images.
个人简介：
Rin-ichiro Taniguchi was born in 1955. He graduated from Department of Computer Science and Communication Engineering, Kyushu University in 1978 and from Graduate School of Engineering in 1980. He received Dr. Eng. from Kyushu University in 1986.
He joined Graduate School of Engineering Sciences, Kyushu University as a research associate in 1980. He has been working as a professor for Graduate School of Information Science and Electrical Engineering,Kyushu University since 1996.Currently, he is the department chair of Department of Intelligent Systems, Graduate School of Information Science and Electrical Engineering.His research interests covers image processing, computer vision and parallel processing. Especially, his current research issues are real-time vision processing and its application to multi-media communication and human-computer interaction. He served PC members, chairs, and secretaries of international/domestic conferences, workshops and societies, including ICPR, QCAV, FCV. He also received several awards from academic societies in Japan.
计算机学院(软件学院)
2004.9.23
主讲人：Rin-ichiro Taniguchi 日本九州大学教授
时 间：9月25日(周六)下午2：30
地 点：玉泉校区曹光彪东楼502
报告简介：
In this talk, I would like to talk about one of the hottest topics in computer vision. Computer vision is a research to develop computer/machine systems which can understand input images and which can recover 3D real world information from the images.
The difficulty in computer vision is its ill-poseness. Images, which are captured by a camera, are 2D information, and, thus, they are reduced information from a 3-D space. Mathematically, recovering 3D information from 2D data can not be solved without any ambiguities.
To overcome this problem, we have employed multi-view image analysis,where multiple cameras are employed to disambiguate the ill-posed problem. One of the biggest problems in the multi-view image analysis is its computation cost, which becomes very severe especially when real-time processing is required. Therefore, we have developed a parallel processing system, i.e., a PC cluster for vision processing, to realize real-time multi-view image analysis.
As its practical application, we have constructed a vision-based real-time human motion sensing system, where we have developed a human posture analysis algorithm from a limited number of cues, which can be robustly acquired from input images.
个人简介：
Rin-ichiro Taniguchi was born in 1955. He graduated from Department of Computer Science and Communication Engineering, Kyushu University in 1978 and from Graduate School of Engineering in 1980. He received Dr. Eng. from Kyushu University in 1986.
He joined Graduate School of Engineering Sciences, Kyushu University as a research associate in 1980. He has been working as a professor for Graduate School of Information Science and Electrical Engineering,Kyushu University since 1996.Currently, he is the department chair of Department of Intelligent Systems, Graduate School of Information Science and Electrical Engineering.His research interests covers image processing, computer vision and parallel processing. Especially, his current research issues are real-time vision processing and its application to multi-media communication and human-computer interaction. He served PC members, chairs, and secretaries of international/domestic conferences, workshops and societies, including ICPR, QCAV, FCV. He also received several awards from academic societies in Japan.
计算机学院(软件学院)
2004.9.23
