Image-based obstacle avoidance and path-planning system

Abstract

This paper proposes an image detection method which is applied to obstacle-detecting and path-planning system of a wheeled mobile robot. The positions of obstacles can be recognized by using medium filter, morphologic processing, canny edge detection, and connected component labeling. And this paper utilizes Parallel Lines Distance Measurement (PLDM) to detect the real world coordinates of obstacles. Because it only needs a single webcam and several laser points for the setting of PLDM, it is easier to set up the experimental environment and the cost for setting is relatively cheaper. At last, we combine image detection method with PLDM to complete the path-planning system of a wheeled mobile robot. Then the path-planning can be shown on the two-dimensional map immediately. Our simulation results, presented at the end of this paper, show that the proposed image detection method is effective in the path-planning system of a wheeled mobile robot.