Design of mobile garbage collection robot based on visual recognition

Abstract

This design mainly studies the intelligent mobile garbage collection robot based on visual recognition technology, which can carry out path planning, traverse the given area, scan and identify and pick up recyclable garbage. The system is composed of navigation unit, target identification unit and sorting control unit. The navigation unit is based on ROS distributed framework, uses lidar to collect the environment information of given area, realizes SLAM function based on scanning matching algorithm, and carries out path planning through optimal path algorithm and traverses the selected area. In the process of robot traversal, the target recognition unit passes through MobileNetv3-SSD deep learning algorithm detects and classifies the target from the image acquired by the camera, obtains the coordinate position and angle information of the target, and controls the manipulator to carry out the garbage grabbing task.