Design Of Robot Teaching System Based On Visual Guidance

With the wide use of robots in industry,improving robot teaching methods has become a major direction of development.At the same time,the rapid development of the field of machine vision has greatly enhanced the degree of robot intelligence.In this paper,robot teaching technology and vision are combined to study the robot visual teaching system,and the following aspects are completed:(1)analyze the development of machine vision teaching technology,and show the importance of target tracking algorithm and robot motion trajectory and velocity planning algorithm.(2)Calibration camera model.By analyzing the parallax principle of binocular camera,the polar line correction of binocular camera is carried out.Then,the spatial cloud information of the image scene is obtained through the semi-global stereo matching algorithm,and the 3D point cloud reconstruction of the image is carried out.(3)For the visual part,the algorithm is divided into two parts: target detection and target tracking.In the target detection,the three-frame difference method is used to detect the tracking target,which solves the problems of cavity and incomplete target detection caused by the traditional inter-frame difference method.An improved Meanshift tracking algorithm is proposed in target tracking.Based on the Meanshift algorithm and the adaptive Kalman filter,the algorithm effectively solves the divergence problem of Kalman filter when the noise characteristics of the system are unknown.(4)For the robot trajectory and velocity planning algorithm,taking the highprecision NURBS curve as the research object,aiming at the defects of the traditional acceleration and deceleration algorithm,an adaptive five-stage acceleration and deceleration algorithm is proposed to realize the flexible acceleration and deceleration control of the system.(5)Analyze the robot visual teaching system,build the software and hardware platform of the system,and verify the effectiveness of the system.