Learning to grasp familiar objects based on experience

摘要

Stably grasping objects for a specific task is a hot research topic in robotics due to multiple degrees of freedom of hand kinematics, incomplete visual sensing of objects, and various shapes of objects. This paper proposes an effective grasp planning method by integrating the crucial grasp cues (positions and orientations of thumb fingertips and the wrist) from human experience. This approach greatly reduces the search space of hand kinematics and meanwhile, represents important human grasp intentions. Then, for various shapes of objects which are partially observable in the visual sensing, the presented approach learns the “thumb” grasp point employing a SHOT shape descriptor based on objects' category level. This method recognizes the grasp point according to the shape affordance at each point on the object, which performs the grasp point generalization on the familiar objects. Finally, we verify the developed methods via both simulations and experiments by grasping various shapes of objects.