Uncalibrated visual servoing of mobile manipulators with an eye-to-hand camera

摘要

Most research on the visual servoing of mobile manipulators used the assumption that the depth information is already known. In this paper, a visual servoing scheme without the prior knowledge of depth is proposed to regulate a mobile manipulator to a desired pose at eye-to-hand setup. To deal with the nonholonomic constraint of wheeled mobile platform, the Jacobian matrix was combined with the depth-independent matrix to construct the image Jacobian. To avoid the sophisticated process of calibration, a parameter updating rule based on the depth independent matrix is developed to estimate the intrinsic and extrinsic parameters. Lyapunov theory is used to prove the asymptotic convergence of the image errors. Simulation have been implemented on a 6-DOF mobile manipulator. Simulation results show the effect of the algorithm.