摘要:his paper aims the development of autonomous obstacle avoidance algorithm for automated guided vehicle with a Kinect sensor. To do this task, the followings are executed. Firstly, depth data is obtained from the Kinect sensor, and the mean filter approach is developed to filter out noises. Secondly, Otsu and U-V parallax based method is designed to identify background and obstacles. Finally, According to barrier information, corresponding fuzzy control rules is generated for the path planning. Our experiments show that the proposed method is fast, and gives promising results.