Design of Automatic Recycling Robot Based on YOLO Target Detection

Abstract

In order to achieve automatic item grasping and recovery, we propose a system design method based on YOLO v4 automatic recovery robot, using the higher computing power of Jetson Nano and STM32F103ZET6 computing units, processing image information to control the operation of the robot system, with six degrees of freedom PWM robot arm to accurately grasp the items. After system testing, the average item recognition rate exceeds 98.5%, and the recovery success rate exceeds 96%, truly achieving automatic search, grasp, recovery, and return end-to-end operation.