A Robust Person Tracking and Following Approach for Mobile Robot

Abstract

The ability to detect and follow a specific person is considered to be a key prerequisite of service mobile robots in indoor environments. A novel vision-based tracking and following method for a mobile robot is proposed, which combines the clothes color features of the person's upper body and the contour of the head-shoulder. Horizontal-Projecting Probability Histogram (HPPH) of torso based on particular color clothes of upper body is proposed for extracting the pattern features, and the contour matching method by the head-shoulder is integrated to reduce the false positive detection rate, which significantly improves the tracking reliability. The proportional compensation method is utilized to enable the robot to follow the targeted person. Experimental results validate the robust performance of the approach.