Image-based visual servoing of nonholonomic mobile robots

Abstract

In this paper, a novel image-based visual approach for the position control of a nonholonomic mobile robot is presented. The mobile robot is endowed with a fixed camera, and visual feedback is used to control the robot pose with respect to a rigid object of interest. After introducing a three dimensional state space representation of the camera-object visual interaction model fully defined in the image plane, a closed-loop stabilizing control law is designed, based on Lyapunov's direct method. The image-based control scheme, which uses a discontinuous change of coordinates, ensures global asymptotic stability of the closed-loop visual system. Moreover, in the case of known height of the object, global stability is formally proved using an adaptive control law. Experimental results obtained with a tank model validate the framework, both in terms of system convergence and control robustness.