Transparent grippers for robot vision

While existing grippers execute manipulation tasks, they occlude parts of the objects they grasp as well as parts of the workspace from vision sensors. We present the concept of a transparent gripper that enables vision sensors to image an object without occlusion while it is being manipulated. The physics of refraction, total internal reeection, lens eeects, dispersion and transmit-tance are analyzed to determine the geometry and material properties of a practical transparent gripper. Algorithms are presented that compensate for image shifts caused by refraction. The experiments demonstrate the proposed gripper to be an eeective solution to a variety of problems, including 3D object model generation. A robot generally interacts with the physical world by means of a gripper mounted on its end-eeector. The level of task complexity that the robot can handle is determined to a large extent by the gripper it is equipped with. For this reason, grippers and the process of grasping have been intensely studied in recent years 10]]9]]6]. This has led to the design and fabrication of a panoply of grippers that range from simple parallel-jaw ones to sophisticated anthropomorphic hands. While the gripper determines the mechanical dexterity of a robot, sensory feedback determines the level of intelligence a robot can achieve. Vision serves as a powerful component of such a feedback system. It provides a richness of information that can enable a robot to handle uncertainties inherent to a task and react to varying environments. As all grippers available today are made of materials that are opaque, a robot vision system has no way of observing an object in its entirety, while it is being manipulated or even simply held by the gripper. We attempt to bridge the gap between these two seemingly unrelated areas, namely, gripper design and visual sensing. We seek to develop a gripper that would enable a robot to \see" while it \grasps." Our solution is a transparent gripper that allows a vision sensor to see through to the object it holds. A transparent gripper has, among others, two pertinent applications. First, it allows vision sensors to monitor an assembly or manipulation task in progress. This improves the perceptual capabilities of the robot system, and hence allows it to recover gracefully from errors. The second application involves the task of automatic generation of 3D object models. In both machine vision as well as in computer graphics, there is a tangible need …