Robot visual servoing based on total Jacobian

In robot visual servoing, only the conception of image Jacobian is traditionally utilized, which indicates that the image feature varies only with the robot motion. But for a moving object, the image feature can he affected both by the robot motion and by the object motion. In this paper total Jacobian and object Jacobian are defined. The total Jacobian matrix, consisting of image Jacobian and object Jacobian, is estimated using exponentially weighted least square algorithm. No calibration is required. The control scheme is image Jacobian pseudo-inverse appending an item related to the object motion. Results for both two and six degree-of-freedom systems demonstrate the success of the algorithm.