Research And Development Of Car Door Grasping And Positioning System Based On Machine Vision

Because of the characteristics of high positioning precision,non contact and good stability,the application of machine vision in the field of industrial automation is becoming more and more extensive.With the rapid development of computer technology in recent years,the combination of machine vision technology and robot is becoming more and more widely.The installation of "eyes" for robots is a main part of flexible production,Robots can use the eyes to realize the functions of measurement,detection,recognition and location,so as to improve the level of intelligence in production.In automobile production,the welding workshop mainly completes the five main parts of the white body(the side wall,the five doors and one cover,the rear floor,the front floor and the engine room).Therefore,the automobile welding production line has a lot of job to do,such as door operation,commonly worked by artificial.This can not be achieved the automation of production sustainable,there exists some defects in the aspects of flexibility,efficiency and precision.And these operations can be completed by the machine vision system with industrial robots,and do not need manual participation.In this thesis,the part of manual operation of the automatic welding production line is studied.Without changing the original coarse positioning basket,the machine vision positioning technology is applied to locate the workpiece.The machine vision positioning technology can well meet the needs of flexible production,achieve automatic operation of the whole process of welding workshop,reduce the cost of the transformation production line,and improve the automation level of welding workshop.This thesis studies the technology of grasping and positioning syst em of six axis industrial robot in the automobile welding workshop.It combines visual technology with industrial robot to achieve robot's grasping and positioning for target workpiece.The main content and innovation of the thesis are summarized as follows:1.This thesis collects the domestic and foreign literature application of machine vision in industrial field,through the investigation found that further research its application in automobile production line is needed,the machine vision applied to automobile production on the workpiece feeding operations,and do the research and development work.2.This thesis studied the composition of the vision of industrial robot,and the key technology of the system is set up about the hardware selection,and the visual system is built by the actual test.In this thesis,the structure of Eye_to_hand is designed according to the conditions and requirements of the site.And also,the hardware selection of the visual system and the design of the visual algorithm are mainly studied.3.The pinhole model of the camera and the camera calibration methods are analyzed and studied,and the calibration methods of the two kinds of hand eye system is analyzed theoretically.Then,based on the requirements of the field staff,a calibration method of hand eye system based on the homography matrix is proposed.This method is easy to understand and implement.4.The related algorithms of image preprocessing was studied,including image filtering,mathematical morphology,edge extraction and other algorithms.Then we studied the workpiece location methods respectively,including image centroid,minimum outer rectangle and template matching.