Development of Mobile Robot System Based on Computer Vision

In order to solve the adaptability problem of traditional logistics robots' grabbing and handling operations in automated production lines, and to realize the rapid identification, positioning, grabbing and handling of target objects, a design method of omnidirectional mobile logistics robot with binocular vision is proposed. The system integrates the application of omnidirectional mobile platform, material grabbing robot, binocular visual recognition and other technologies. The mechanical mechanism principle and kinematics model of the robot are analyzed. Image recognition and target contour extraction are completed by the visual recognition algorithm. The target object is positioned by camera calibration. Finally, a set of physical prototypes was designed and tested. The experimental results show that the system can effectively complete item identification, grabbing and handling.