Vision-based obstacle avoidance system for autonomous mobile robot in outdoor environment

Abstract

In this paper, we propose an obstacle avoidance system on based vision sensor for an autonomous mobile robot. For real-time turn angle correction, we obtain the state equation of a mobile robot from input-output continuous data. Each individual image pixel is classified as belonging either to an obstacle or to a non-obstacle based on its color property (HSI color model). The HSI color model is less sensitive to illumination changes than the RGB color model. According to proper conditions and reference zones, we chose a path area, a navigation point, and a turn angle. This method uses a single color camera.