An image based method of finding better walking strategies for hexapod on discontinuous terrains

摘要

The adaptive gait planning is an important aspect for walking robot traversing on rough terrain. Hexapod robot is robust among all legged robots walking over rough terrain. For rough terrain, many studies have been performed on robot vision system to recognize the surrounding terrain. But those studies are based on simple rule of forbidden walking regions of the robot. This paper presented a special image based method to recognize the surrounding terrain. Moreover, a number of parameters used to establish, modify, and analyze the terrain. A hexapod robot walking gait model also designed to test the effectiveness of proposed method. Simulation results show that the method can handle large databases with image processing technique to find out the optimum strategy of walking on rough terrain.