Visual Identification And Sorting Of Multiple Workpieces Based On Industrial Robot

Shield machine is a kind of construction machine widely used in tunnel construction.Its production mode is a typical multi-variety and small-batch production mode,and it requires a wide variety of parts.In the production process of the parts,the automation of the processes represented by the sorting and cutting of the workpieces is relatively low,which becomes a bottleneck that restricts the increase in the productivity of the shield machine.In this paper,aiming at the problem of low manual operation efficiency in the process of workpiece sorting,starting from the automation and intelligent upgrade requirements of the sorting process,based on the machine vision technology,a template matching method based on DXF files is proposed,and in-depth research on image splicing,contour extraction and template matching are conducted.The main research work of this paper are as follow:(1)According to the requirements of visual recognition and sorting of workpieces,a machine vision platform based on KUKA robot was designed and built,HALCON was selected as image processing software,and the light source,lighting method and the sorting method of workpieces were explained,and the calibration of the camera and robot tools were performed.(2)In the aspect of image preprocessing,the image with distortion is corrected,and the image mosaic technology is studied.The image mosaic technique based on Harris point detection algorithm is used.At the same time,in order to ensure that the points in the mosaic image can be used to guide the robot to perform workpiece positioning and sorting,the transformation matrix between the image coordinate system and the robot coordinate system is solved.(3)For workpiece recognition and contour extraction,color space conversion is performed,and image smoothing and contour extraction methods are studied.The Gaussian filter is used to smooth the noise of the image,and the contour extraction of multi-workpieces under complex background is studied.The contour extraction method which based on threshold segmentation and image morphology processing is applied and smooth workpiece profiles are obtained.(4)For template matching and workpiece positioning,DXF file is selected as matching template,matching methods based on pyramidal hierarchical search and similarity measures are applied and the target workpieces are accurately positioned.Picking experiments are conducted and the workpiece recognition and positioning method proposed in this paper is proved to be useful,which can meet the requirements of rapid and accurate sorting of workpieces in the shield machine production process.