Research On Automatic Positioning Bending Machine Based On Machine Vision

For domestic sheet metal bending process and bending equipment,bending workpieces of different sizes and shapes must be processed by different specific molds,and have the characteristics of low utilization rate,low bending precision and low efficiency.On the basis of bending machine which is existed,to the traditional sheet metal bending process equipment technical reform and technical innovation,developed a high precision automatic positioning bending machine based on machine vision system,the use of machine vision recognition technology to realize automatic positioning in the process of sheet metal processing,remove the multifarious specific mold in the process of sheet metal processing,use of visual image processing to realize accurate positioning in the process of sheet metal processing and automatic processing.Specific research contents are as follows:(1)By comparing the advantages and disadvantages of several commonly used sheet metal bending schemes and combining the actual requirements of the factory,a set of fully automatic bending equipment is designed.According to the bending scheme,the overall structure design of the bending machine is completed.According to the related technical parameters of the bending machine,the model selection of the camera,lens,light source and other main equipment on the visual system is completed for the established machine recognition system.(2)A series of pre-processing operations are performed on images of refrigerator shesheet workpieces by camera.The weighted average method is used to gray-scale image processing,the median filtering algorithm which is the best of three traditional filtering algorithms can denoise the gray-scale image.The foreground and background of refrigerator sheet image result have been separated by using close-open operation of morphology in image after binarization.The residual image and the dark spot in the image of sheet are solved effectively.(3)Edge detection was carried out for the pretreated thin plate graph,and the image was extracted by the five operators of Canny,Sobel,Roberts,Prewitt respectively.Compared with several edge detection operators,the image processed by the Canny operator could better reflect the edge information of the image.The edge line is fitted by the least square method inthe image detected by the edge,the deviation between the edge and the edge information of the template is found,and the Angle of deviation is calculated.(4)In order to improve the accuracy of automatic positioning of thin plate,hand-eye calibration experiment is conducted on the mechanical hand.Establish Eye-to-Hand hand-eye calibration model of visual system planning calibration plate movement space,using if camera calibration method analysis model,determine the camera calibration process,complete the image pixel coordinates and mutual transformation between the world coordinates of the manipulator,the camera calibration experiment: connect the 7*5 black and white squares and white board consisting of a board placed in 14 different azimuth of calibration template as a calibration object,identified 24 in each picture inside pixel point overall average error is 0.0962,satisfies the requirement of camera calibration precision.The manual calibration of vision system takes two thin plate workpieces with different positions and postures as objects,and extracts the coordinates of the center point.The error between the measured coordinates after the manual calibration is less than 1mm,which meets the requirements of calibration precision.(5)The combination of research and development of machine vision and bending specimen,respectively in order to pick up the automatic positioning bend prototype of machine vision and not carry the traditional bending machine for machine vision test base,in100,the size is 600 mm * 300 mm,and folding size of 50 mm fridge sheet workpiece as an object,is divided into five groups of experiments,each group of 20 pieces of sheet,experimental prototype equipment an average success rate of 99.00%,significantly higher than the traditional forming success rate of bending machine,industrial requirements.