In order to improve the target tracking accuracy of the robot visual servo control system,a visual servo control system based on switched Kalman filter was constructed.The principle of visual servo target tracking was studied.The related mathematical model was deduced,and the reason for the tracking error was analyzed.Aiming at the problem of delay introduced by the image acquisition and processing,the velocity of the target was estimated by Kalman filter,and the data was fed forward to the visual servo controller to compensate the tracking error caused by the target motion and delay.For the problem that the performance of the Kalman filter was degraded due to a sudden change of the target motion,a motion monitor was introduced to send a switching signal and reset the Kalman filter when the target motion was suddenly changed.Finally,the algorithm was verified by simulation and experiment.The results indicate that the visual servo system based on Kalman filter can control the tracking error within 1 mm.The switched Kalman filter can effectively decrease the tracking error caused by abrupt motions of target.