碩士 / 國立臺北科技大學 / 電機工程系碩士班 / 89 / The problem of automatic guided vehicle (AGV) is to detect the obstacle ahead correctly. Using ultrasound sensors or infrared sensors can detect the approximate distance, but to obtain accurate distance largely depend on the reflection rate of various materials of the obstacles and the shape of the obstacles. The size of the obstacle is also unknown. With visual image, we can understand the type of the obstacles. Combining with proper method to detect the distance, we can control the AGV avoiding obstacle in advance.
A novel vision-based obstacle detection method is proposed in this thesis. By projecting the laser spot onto the obstacle, the distance between AGV and obstacle as well as the width of the obstacle are obtained based on trigonometric calculation. By practically measuring the location of each sample spot in Cartesian space and image space, an artificial neural network (ANN) is used to approximate the nonlinear mapping from Cartesian space to image space. The obstacles are located by the proposed ANN as the AGV proceeds and takes obstacle avoidance strategies in effect.
| Identifer | oai:union.ndltd.org:TW/089TIT00442008 |
| Date | January 2001 |
| Creators | Cheng-Tung Chang, 張正東 |
| Contributors | Leehter Yao, 姚立德 |
| Source Sets | National Digital Library of Theses and Dissertations in Taiwan |
| Language | zh-TW |
| Detected Language | English |
| Type | 學位論文 ; thesis |
| Format | 169 |
Page generated in 0.0093 seconds