Research On Measuring Technology Of Shaft Workpiece Geometric Quantity Based On Machine Vision

With the rise of intelligent manufacturing technology,higher requirements are placed on intelligent automated measurement of industrial sites.The traditional measurement methods based on contact measurement have the disadvantages of high labor intensity,low precision,and easy damage to the workpiece.Automated,accurate,non-contact measurement of workpiece geometries can be achieved using machine vision-based measurement methods.Based on the analysis of the research status of machine vision measurement technology at home and abroad,this paper designs a high-precision machine vision measurement system,focusing on the key technologies of machine vision measurement,including lens distortion correction,pixel equivalent calibration,sub-pixel edge extraction and Parallax image stitching,etc.Firstly,According to the measurement requirements,the mechanical structure,hardware design and selection and measurement software scheme of the system are determined.Secondly,in order to improve the system accuracy,edge detection and sub-pixel edge positioning technology are studied.Several common edge detection algorithms are analyzed in detail,and the superiority of Canny edge detection algorithm is determined.The Canny algorithm is combined with the sub-pixel edge detection technology based on pixel neighborhood.An improved noise image sub-pixel edge extraction algorithm which based on iterative reconstruction is proposed.The improved algorithm is applied to the image processing of the axis workpiece,which can obtain better edge detection effect.Then,for the long-axis workpiece,the large parallax image is studied.The stitching algorithm proposes an APAP stitching algorithm with similarity constraints,which not only ensures the stitching accuracy of the image,but also solves the problem of shape distortion of the non-overlapping regions of the image.Using this algorithm to splicing the stepped axis images with partially overlapping regions,a good stitching effect can be achieved.Finally,according to the above technical principle,the software and hardware system is built and the length of the stepped shaft and the tapered surface of the cone are measured.It is available that the shaft diameter measurement error is between ±0.03 mm and the angle measurement error is between ±0.3°,the effectiveness of the machine vision based precision measurement system for workpiece geometry was verified.