Image-based exploration obstacle avoidance for mobile robot

Obstacle avoidance is a key component of autonomous systems. In particular, when dealing with large robots in unmodified environments, robust obstacle avoidance is vital. This paper presents a new image-based exploration algorithm for a mobile robot equipped only with a monocular pan-tilt camera to autonomously explore the natural scene structure in indoor environments. The algorithm inferred and computed the frontier information directly from the segmentation images and classified each super-pixel as belong either to an obstacle or the ground plane. The method used the distance and orientation information of the frontier to control the robot to avoid collisions. Experimental results on a mobile robot in an unmodified laboratory and corridor environments demonstrate the validity of the approach.