| With the deep integration of the new generation information technology and manufacturing industry,the traditional manual detection mode has been unable to meet the fast-paced and flexible production needs.In the automatic production line,the visual system needs to cooperate with the industrial robot when necessary,and match with the PLC control system of the production line to achieve online detection,identification,positioning and other functions.Based on the principle of image processing algorithm and the key technology of visual positioning,this paper focuses on the application of machine vision system in automatic production line.In the research process,combining machine vision and industrial robot,a set of industrial robot vision guidance system for an automatic production line is designed and developed,which focuses on solving the problem of low recognition and positioning accuracy in the process of robot hand eye calibration,as well as realizing the on-line detection of workpiece.The main research contents and conclusions are as follows:(1)According to the detection requirements of each station in an automated production line,hardware system is built in the laboratory environment,which mainly includes the selection and installation of industrial cameras,optical lenses,lighting sources and other equipment.The influence of different imaging models on camera calibration results is explored,and the principle of camera calibration algorithm is revealed.On this basis,the camera calibration algorithm is designed based on Halcon vision software.The results show that the barrel distortion exists in the selected lens,and the non telecentric poly centric imaging model of face scan is more suitable for the actual optical imaging model.The average error of camera calibration using the polycentric imaging model is the smallest and the center distance of mark point is closer to the actual value.Finally,the internal and external parameters of the selected camera are obtained while the barrel distortion of lens is corrected by the phase machine calibration Finally,the vision positioning system is calibrated accurately.(2)The strong positioning algorithm based on edge detection was used for hand-eye calibration of industrial robots,the detection principle of traditional positioning algorithm was explored.The experiment of traditional localization algorithm based on threshold segmentation and template matching is designed and implemented in the laboratory environment and actual working condition environment.In order to solve the problem of poor positioning accuracy and anti-interference of traditional positioning algorithm,a strong positioning algorithm based on edge detection is proposed to precisely locate the mark points in the nine point calibration image.On the basis of the positioning algorithm based on threshold segmentation,the nine point calibration image is segmented by using the best pre calculated threshold;in order to eliminate the gray-scale interference of the background area,the image is further segmented The binary operation and feature detection and other image processing operations are carried out.After the mark point is accurately segmented from the image,the edge detection algorithm is used to extract the contour of the mark point,and the mark point central coordinates are accurately obtained after five iterations of fitting the mark point through the least square method based on Tukey.Through experiments,the effectiveness of the strong positioning algorithm based on edge detection is verified,the positioning accuracy of hand eye calibration is improved,and it has strong anti-interference ability.(3)The principle of image processing algorithm is studied in depth,and the corresponding image processing algorithm is designed according to the shape characteristics of workpiece.In order to improve the robustness of the algorithm and the efficiency of the system,the gray-scale processing of the workpiece image can reduce the dimension of the color image matrix,while greatly improving the operation speed;in order to filter out the noise points in the workpiece image,the analysis and comparison of the neighborhood average filtering,median filtering,Gaussian filtering and bilateral filtering and other filtering denoising algorithms are made.The experimental results show that when using the bilateral filtering algorithm to smooth the image,it can not only filter the noise,but also protect the edge pro file information of the workpiece.In order to realize the online detection of the workpiece,the edge detection of the workpiece contour is carried out.The least square method based on Tukey is used to fit the straight line of the workpiece contour in the reference image,and the template matching algorithm based on shape is used to identify and locate the template area in the measured image.Affine transformation matrix is established by using the position and orientation data obtained by template matching,the fitted edge line is affine transformed into the image to be tested,the edge contour of the workpiece in the image to be tested can be obtained by performing edge detection again,and finally the position and orientation data of the workpiece can be calculated by the mathematical formula.(4)According to the project requirement of an automatic production line,the visual guidance system is designed.Based on WinForm desktop application program,a visual guidance system including motion master control software and three visual positioning software is designed and developed by using C#language and Halcon visual software.In order to realize the data communication between motion master control software,each visual positioning software and PLC,a set of communication protocol is written.After many times of debugging,the visual guidance system designed in this paper can assist the motion control system to complete the workpiece loading and unloading,gluing,precision assembling and screw fastening and other production line tasks,which verifies the reliability and effectiveness of the visual guidance system. |
PDF Full Text Request