Visual servoing of rigid robots with modified Cartesian-based encoding

摘要

Deals with issues concerning visual servoing of robotic manipulators in fixed-camera configurations employing modified Cartesian-based task encoding. It is shown that in the absence of measurement noise precise 6-DOF positioning sometimes can possibly be achieved despite camera model imprecision. Based on the modified Cartesian-based encoding approach, the encoded task can be implemented without knowing the precise camera model and which, if accomplished, implies that the original 6-DOF positioning task is accomplished with precision as well. A visual servo controller that is capable of driving the encoded error to zero is proposed. The overall closed-loop system is exponentially stable provided that camera model imprecision is small.