Person following control for a mobile robot based on color invariance corresponding to varying illumination

Abstract

In this paper, we present a method of person following control for a mobile robot using visual information. Color information is often used for object tracking. Color information of objects varies greatly under illumination changing environment. In such conditions, the robot controlled by visual information may lose sight of a person. In this paper, we consider a robust person following method by color invariance and image-based control. Color invariance shows robust features of colored objects in terms of changing illumination conditions. At first, we estimate the lowest positions of both feet of a tracked person through particle filters based on color invariances. Then, we control the velocity of the robot to track the person by using the image-based controller. Experimental results using an actual robot demonstrate the effectiveness of the proposed method.