RELATIVE END-EFFECTOR CONTROL USING CARTESIAN POSITION BASED VISUAL SERVOING

This paper presents a complete design methodology for Cartesian positi on based visual servo control for robots with a single camera mounted at the end-effector, Position based visual servo control requires the explicit calculation of the relative position and orientation (POSE) o f the work-piece object with respect to the camera, This is accomplish ed using image plane measurements of a number of known feature points on the object, and then applying an extended Kalman filter to obtain a recursive solution of the photogrammetric equations, and to properly combine redundant measurements, The control is then designed by specif ying the desired trajectories with respect to the object and forming t he control error in the end-effector frame, The implementation using a distributed computer architecture is described, An experimental syste m has been built and used to evaluate the performance of the POSE esti mation and the position based visual servo control, Several results fo r relative trajectory control and target tracking are presented, The r esults of experiments showing the effect of loss of some of the redund ant features is also presented.