A computationally efficient approach for Jacobian approximation of image based visual servoing for joint limit avoidance.

摘要

A dynamic Newton's approximation based esti-mation scheme is proposed for image-based visual servoing (IBVS) of a redundant manipulator. The estimation scheme approximates the kinematic Jacobian from joint space to vision space online. The estimated Jacobian is used for redundancy resolution with weighted least norm solution. This work dis-cusses the kinematic limit avoidance with estimated Jacobian for a 7 DOF PowerCube manipulator, hence presenting an approach with reduced computational burden compared to traditional model based visual servoing approaches.