| Project/Area Number |
08455119
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (B)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Intelligent mechanics/Mechanical systems
|
|---|
| Research Institution | The University of Tokyo |
|---|
Principal Investigator |
INOUE Hirochika The University of Tokyo, Dept.of Mechano-Informatics, Professor, 大学院・工学系研究科, 教授 (50111464)
|
|---|
| Project Period (FY) |
1996
|
| Project Status |
Completed (Fiscal Year 1996)
|
| Budget Amount *help |
¥6,700,000 (Direct Cost: ¥6,700,000)
Fiscal Year 1996: ¥6,700,000 (Direct Cost: ¥6,700,000)
|
|---|
| Keywords | robotics / computer vision / intelligent robot / vision processing / autonomous robot / route map / teaching |
|---|
| Research Abstract |
"View Sequence" is a sequence of views taken by the camera of the mobile robot along a route. In this research, we developed a model for representing a route and a method for real-time navigation based on the view sequence. The results of the research are summarized in following three points :
(1) Development of a route representation based on view sequence
In the teaching run by a human, the robot automatically takes sequential views. By smooting and shrinking the image, the route representation using view swquence becomes compact. This representation is easy to aquire and suitable for representing a route in the unstructured indoor environment.
(2) Development of the recognition method based on the matching of the views
The matching of views is performed based on correlation of the images. As the results, the vertical and horizontal displacement and the distortion between two views, and consequently the position of the robot and the presence of obstacles are recognized.
(3) Development of guiding a robot along a route
The action of the robot and the control values (steering angle and speed) are determined using the result of matching. The autonomous navigation based on the view sequence is realized using the methods described above.
|
