Image feature based navigation of nonholonomic mobile robots with active camera

Abstract

This paper proposes an image feature based navigation method for active camera mounted mobile robots that are affected by nonholonomic constraint. Since visual servo used for usual manipulators can't be applied to the nonholonomic mobile robots as it is, visual servo is applied to tracking control of the active camera to track targets in the center of image plane. Moreover, posture of nonholonomic mobile robots from targets is controlled by image features of targets that are related to relative distance and angle between robots and targets. It is confirmed that the proposed method can navigate a mobile robot in front of a target through simulations.