Research On Hand-eye Calibration And Target Tracking Method For Industrial Robots

The key technologies of intelligent robots include hand eye calibration,target tracking,motion control and real-time feedback.Hand eye calibration and target tracking are important steps for robot to acquire target information and achieve precise movement through camera.The first step in the integration of robots and vision technology is hand eye calibration.The accuracy of hand eye calibration will seriously affects the robot's task.In the traditional robot hand eye calibration method,the calibration board is fixed,and the eye relationship is calibrated based on the motion of the robot.But,the precision of the robot drive chain,the accuracy of the location information between the camera and the hand will seriously affect the accuracy of hand eye calibration.In view of the high precision positioning holes in the flange of the robot wrist,it only needs to calibrate the position relationship between the end effector,the flange and the camera.Therefore,this paper proposes a high precision hand eye calibration method based on the three coordinate measuring machine,which does not need the robot transmission and is not affected by the motion precision of the robot,so the hand eye transformation matrix is solved directly.This paper analyzes the traditional methods of robot hand eye calibration,error sources and methods of improving accuracy.A hand eye calibration method based on three coordinate measuring instrument is mainly designed.Through the coordinate system of three coordinate measuring machine,the coordinate transformation loop of the flange coordinate system,the terminal actuators coordinate system,the camera coordinate system and the robot base coordinate system is made up.A robot grasping model based on UR5 is designed to calibrate the relationship between end effector,flange and camera.The accuracy verification experiment of hand eye calibration method based on three coordinate measuring instrument is designed.The calibration accuracy is verified in two directions:the accuracy of the pixel in the image coordinate system and the physical length in the world coordinate system.The experimental results show that the pixel accuracy can reach 1 pixel and the physical length accuracy can reach 0.21mm.The precision of hand eye calibration is greatly improved.In order to make the robot have the ability to track the target,this paper deeply studies the TLD target tracking method,and studies the operation mechanism of the target detection module,the target tracking module and the learning module.The operation principle of cascade classifier based on variance classifier-ensemble classifier-nearest neighbor classifier is analyzed,and the forward and backward error calculation method and P-N learning method are analyzed.The error calculation process based on TLD target tracking method is designed.The target tracking software system based on the TLD target tracking method is designed.Finally,the TLD tracking method is applied to the workpiece such as nuts and acrylic products,and the target tracking experiment is carried out to verify the effectiveness of the TLD tracking method.