摘要
An intelligent approach to robot navigation by landmark tracking using computer vision is proposed. This approach is based on the concept that a human can reach the destination by tracking the specific landmark in a previous environment. Only a monocular image of the landmark taken by the robot is required. Techniques of image processing and pattern recognition are integrated to resolve the problems of landmark tracking and understanding. Simulated results to confirm the feasibility of this approach is also included.
原文 | 英語 |
---|---|
頁(從 - 到) | 79-89 |
頁數 | 11 |
期刊 | Mathematical and Computer Modelling |
卷 | 26 |
發行號 | 4 |
DOIs | |
出版狀態 | 已出版 - 08 1997 |