| Project/Area Number |
16360197
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (B)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
System engineering
|
|---|
| Research Institution | University of Tsukuba |
|---|
Principal Investigator |
YASUANGA Moritoshi University of Tsukuba, Graduate School of System and Information Engineering, Professor, 大学院システム情報工学研究科, 教授 (80272178)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
YAMAGUCHI Yoshiki University of Tsukuba, Graduate School of System and Information Engineering, Assistant Professor, 大学院システム情報工学研究科, 講師 (30373377)
YOSHIHARA Ikuo University of Miyazaki, Faculty of Engineering, Professor, 工学部, 教授 (20322315)
|
|---|
| Project Period (FY) |
2004 – 2006
|
| Project Status |
Completed (Fiscal Year 2006)
|
| Budget Amount *help |
¥10,300,000 (Direct Cost: ¥10,300,000)
Fiscal Year 2006: ¥3,200,000 (Direct Cost: ¥3,200,000)
Fiscal Year 2005: ¥4,000,000 (Direct Cost: ¥4,000,000)
Fiscal Year 2004: ¥3,100,000 (Direct Cost: ¥3,100,000)
|
|---|
| Keywords | Multimedia / Pattern Recognition / Internet / Reconfigurable LSI / FPGA / Tracking / Surveillance Camera / High Speed Signal Propagation / 入出力 / CCDカメラ |
|---|
| Research Abstract |
The goal of this research is to develop a new pattern recognition system based on DDI (Direct Data Implementation) that is proposed by the head investigator of the research. Fundamental research on the DDI has been promoted already under other grants from the Ministry of Education, Culture, Sports, Science and Technology, and Japan Science and Technology.
During this research project, we combined the previous results from the fundamental research on the DDI with the PNN (Probabilistic Neural Networks) and developed a new hardware system. The PNN is one of configurations of neural networks in which each neuron-element plays a role of a kernel function of the probabilistic distribution function in the pattern recognition. The PNN is expected to be applied usefully to the new pattern recognition hardware combined with the DDI. Under this idea, we have developed a hardware system to recognize hand shapes, or hand-charades. This research has been carried out mainly for the electronic circuit development. During this research project, it was also found that the research results could be combined with the pan-tilt camera to develop a novel surveillance system, which is very important equipment recently for social safety. On the basis of this concept, we have started developing the bio-inspired camera system that mimics the eye-system and provides new surveillance functions. The DDI and the PNN meet the requirement of the bio-inspired camera because high speed pattern recognition is indispensable for it.
The fundamental idea of the DDI and the PNN is not limited to the image recognition but to other patterns such as speech and signal spectrum. We have not applied the DDI and the PNN to these applications yet, and they are the next goals of the research.
|
