Abstract: In order to make an automatic guided vehicle (AGV) identify the guidance path, a method about simplified image processing, including image segmentation, filter, border detection, distortion recti fying and an improved way to get the relative position of the body is presented. The method can reduce noise disturbance and the influence of the distorted image and make AGV perform the real-time image processing. Meanwhile, an optimal control strategy for path tracking controlling of the AGV is presented. In order to testify the correctness of the anal- ysis of theory, the image of lane marker is processed and then the experiment of the path tracking is carried out. The results show that AGV with the method and the optimal control strategy could carry out the path tracking accurately and reliably.

Key words: Automatic guided vehicle Optimal threshold, Path tracking Image processing