Increasing the Tracking Region of an Eye-in-Hand System Using Controlled Active Vision

This paper presents a control strategy which allows a manipulator/camera system to track objects with planar motion while simultaneously avoiding kinematic singularities by moving in directions along which the tracking task space is unconstrained. The projection of a cartesian manipulability gradient on these directions of motion is used to determine the magnitude and direction of the required motion for singularity avoidance, while the controlled active vision paradigm is used for tracking. The algorithms developed have been experimentally verified on an eye-in-hand system. Results demonstrate the effectiveness of the method by showing that the tracking region of a manipulator tracking objects with planar motion can be increased by a factor of between two and three. The overall accuracy of the tracking system is also improved.