A Design of Robot System for Rapidly Sorting Express Carton with Mechanical Arm Based on Computer Vision Technology

Abstract

With the rapid development of logistics and express industry, more and more cartons need to be sorted rapidly, which also consumes a lot of human labor. Manipulators are currently used in industrial and medical fields, with high requirements for accuracy, repeatability, and operation stability. With the continued expansion of the social economy and the advancement of science and technology, computer vision technology has become increasingly widely used in a variety of industries. Using computer vision technology in an automation system can improve picture processing accuracy and allow precise recognition. A prototype robot system for picking express cartons is designed, which combines object detection technology based on computer vision and trajectory planning methods in UR3 manipulator grabbing. This design aims to simulate the car with mechanical arm identify the express carton, grab and place it in the storage after moves to the preset grabbing point. Building simulation environment in gazebo, simulating the grabbing and placing process in gazebo by using ROS1 to analysis the kinematics and dynamics control of the manipulator system. The simulation results verify the feasibility of the architecture design and the accuracy of the dynamic controlling system design. Different size express cartons are recognized by using Open CV. Using MoveIt to achieve collision detection and smoother trajectory planning. This prototype design can also provide a reference to future space intelligent robot.