A sensory robotic grasping system for object recognition and shape recovery

Describes a sensory robotic grasping system which is composed of an ADEPT robot, a servo gripper system, an image grabbing and processing system, a microcomputer and an optical soft rubber tactile sensing system. The system can stably grasp and effectively recognize machine parts as well as breaking objects, such as eggs, fragile glass containers, etc. The authors also present the object recognition and model-based shape recovery using two groups of tactile image sequences acquired from the two tactile sensors while the sensory gripper grasps objects. The algorithms have been implemented and tested on actual data obtained using this sensory robotic grasping system.<
>