Iterative estimation of the relative spatial poses of a robot end effector and target object

A technique for finding the relative spatial pose between a robotic end effector and a target object to be gasped without a priori knowledge of the spatial relationship between the camera and the robot is presented. A previously developed technique is used to determine the pose of the end effector and a target object relative to the camera. The transformation between the camera's and the robot's coordinate frames is computed dynamically. The spatial transformation between the robot end effector and the target object is computed with respect to the robot's coordinate system. The algorithm was demonstrated using a 5-degree of freedom robot and an RGB camera system. The camera can be dynamically positioned without concern for an assumed spatial relationship between the camera and robot allowing optimization of the view of the object and the end effector. Further, the iterative nature of the grasping algorithm reduces the effects of camera calibration errors.