Research On Recognition Technology Of Inner Edge Of Coil Of Strip Based On Machine Vision And Coding Trajectory Planning

To prevent loosening coil of strip during transportation,the traditional processing method is to weld the inner edge of coil steel manually,but this method can not meet the increasing demand of production line.In order to improve production efficiency,research on the technology of automatically identifying the inner edge position can provide technical support for robot welding and accelerate the development of coil of strip processing industry.On the other hand,based on machine vision,the coding area is planned on the end face of the coil,and the end effector is driven by the robot to perform the coding work.In view of the above problems and requirements,this paper uses the PC host computer for image processing and the location of the point of weld and the coding area.The main work is as follows:Analyzed the overall task of the system and designed the overall plan,and selected the monocular camera as the image acquisition system;the camera arrangement scheme of the optical axis of the camera facing the end face of the coil is determined,and the camera is fixed on the worktable,forming an "eye-to-hand" system with robot;the camera was calibrated and corrected for distortion,and the error caused by the measurement using visual technology was analyzed.the hand-eye system was calibrated,and the image coordinate system and the robot base coordinate system transformation matrix are calculated,and the error and error source generated by the hand-eye system calibration are analyzed.Since the acquired image contains different noises,the collected image is grayed out and filtered to reduce the influence of irrelevant factors;using the template matching based on variable shape,the rapid recognition and positioning of the end face of the coil in the image is realized;the influence of several common edge detection methods is analyzed,and the Canny algorithm with bilateral threshold is used as the edge detection method.After extracting the required inner edge of the coil,the corner points are used to obtain the two end points of the inner edge of the coil.The overall image processing flow and scheme were designed,and the pose calculation model of the end point of the inner edge of the coil was established.In the end,it is necessary to spot weld the inner edge of the coil and to carry out the coding on the end face,kinematics analysis of the robot was performed using the D-H coordinate method.The advantages and disadvantages of the polynomial interpolation algorithm are analyzed in the joint space,and the position interpolation calculation is performed in Cartesian space.A simplified robotic arm model is built inMATLAB to simulate the desired trajectory in three dimensions.Finally,the MFC framework was used to develop the identification system for the inner edge of the coil.The error generated by the image processing process is analyzed.The robot system was established in Robot Studio,which verified the feasibility of the welding scheme and planned a reasonable welding trajectory.In the simulation robot system,a database of character trajectories was prepared,and a series of trajectories could be obtained through different inputs.