Robust visual servoing: bounding the task function tracking errors

This article deals with the well known problem of robust visual servoing (VS), which is formulated by the use of the general task function approach. Robustness is provided with respect to calibration parametric errors not only in terms of stability, but also in terms of tracking errors boundness. As a result, the proposed approach allows the user to choose a bound on the tracking error so as to guarantee the visibility of the target during the task execution. Most importantly, the error boundness is obtained while avoiding the use of large gains. Rather, one uses bounded gains, together with the modulation of the desired velocity along the desired path. An application of this general approach is developed and experimentally validated.