Location and Collaboration Robot Arm System via Multi-view

As an indispensable factor in the automated production process, the robotic arm needs to adapt to the human-machine interaction scene, so as to achieve more frequent and high-precision interaction tasks and complete the closed-loop human-machine interaction process. In this paper, a collaborative robot arm system is designed, which combines artificial intelligence algorithm and uses computer vision technology as the motion guidance to control the robotic arm to complete the human-machine interaction tasks. The bounding box of the object is obtained by using the Faster-RCNN deep learning algorithm, and the location and principal axis direction of the object are obtained by combining the ellipsoid fitting algorithm to realize the accurate grasp of the robotic arm. At the same time, the haptic feedback module is designed by using the linear vibration motor to respond to the events in the interaction process, so as to complete the intelligent closed-loop interaction process and ensure the production safety of human and machine.