A Secure and Efficient Multi-Object Grasping Detection Approach for Robotic Arms

Robot grasping is one of the most important abilities modern intelligent robots, especially industrial robots. However, existing robot arm’s grasp detection work highly dependent on their edge computing ability, and safety problems in process are not considered enough. In this paper, we propose a new robotic arm model with an edge-cloud collaboration method. With scheme multi-object multi-grasp, our improves mission success ratio grasping. The can only complete compression full-resolution images but also achieve image at limited bit rate. reaches 2.03%; structural difference value higher than 0.91, average speed 13.62 fps. Furthermore, have packaged as functional package ROS operating system, which be easily used actual operations. Our solution fully applied to other robots promote development field robotics.