Vision based robotic person following under lighting variations

Abstract

The paper addresses the problem of person following using color and shape identification in indoors and outdoors environments with significant light variations. This is mainly achieved by real-time control of camera parameters affecting the light exposure, i.e. camera gain, shutter speed and iris. Several other new strategies are also proposed to recognize and segment the person from the environment, and to reacquire the person when he is lost during difficult walking maneuvers. In addition to the camera exposure control, the camera pan and tilt, and the robot speed and steering are controlled based on the characteristics of the detected personpsilas image such as mass (area), centroid, and shape characteristics. The system has been implemented on a Segway robotic platform that has been equipped with a camera, an onboard computer, loud speakers, game-pad, etc. A video has been prepared that shows indoors and outdoors trials under a variety of conditions.