| Project/Area Number |
08408012
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (A)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
社会システム工学
|
|---|
| Research Institution | Okayama University |
|---|
Principal Investigator |
OSAKI Hirokazu Okayamna Univ., Faculty of Eng., Prof., 工学部, 教授 (60032942)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
TOKUNAGA Shuichi Takuma National College of Tech., Dept.of Control Eng., Associate Prof., 電子制御工学科, 助教授 (10197874)
MUNESAWA Yoshiomi Okayamna Univ., Faculty of Eng., Assistant, 工学部, 助手 (70274008)
KAJIHARA Yasuhiro Okayamna Univ., Faculty of Eng., Associate Prof., 工学部, 助教授 (70224409)
|
|---|
| Project Period (FY) |
1996 – 1997
|
| Project Status |
Completed (Fiscal Year 1997)
|
| Budget Amount *help |
¥10,300,000 (Direct Cost: ¥10,300,000)
Fiscal Year 1997: ¥2,700,000 (Direct Cost: ¥2,700,000)
Fiscal Year 1996: ¥7,600,000 (Direct Cost: ¥7,600,000)
|
|---|
| Keywords | Intelligent Robot / Image Processing / CAD / Monitaring System / Tele-Operation / Finger Language / Mial Supporting System / 空間位置 / 動画処理 / 広域視野 / 局所視野 / IP |
|---|
| Research Abstract |
In this study, we aim to develop two main themes, one is the recognition method in local area using CAD figure information, and the other is the recognition method in wide area to monitor the motion of human beings, robots and AGVs and so on. In the first theme, we develop the recognition method (CAD/IP) of the subject by combining the CAD figure information and the input image by image processing. In CAD/IP,3D-CAD figure is translated into 2D-CADfigure is translated into 2D-CAD figure, and the subject is recognized by comparing the characteristics of 2D-CAD figure with those of 2D input image from a camera. Other method can be recognized by comparing 3D-coordinates of 3D figure with those which are calculated by combining two or three 2D images from many cameras. Then we develop on intelligent robot system which can recognize many objects, assemble and handle them.
In the second theme, we propose the recognition method of a subject which is located and moved in the wide area, for example, assembly line, work shop, and factory to monitor the state.
The monitoring cameras located in the wide area that are divided into small convex areas where at least two cameras can be seen. The subject is recognized from the image taken into the constant time interval by the usual image processing, and from the optical flow of the ontinuous image by the optical flow processing system. We develop the tele operation robot system by using wide area recognition method.
|
