Research On The Measurement Method Of Box Aperture And Center Distance Between Holes Based On Machine Vision

Vision measurement is widely used in modern production fields because it can obtain enough information,integrate with the digital design information of products,and feedback information from the processing control system to form closed-loop control of the product production process.In recent years,how to apply visual measurement technology to achieve online and in-situ measurement of products has attracted widespread attention.This paper researches the visual measurement methods of box aperture and center-to-center distance.It is applicable that a calibration device for calibrating the equation of the hole end face was designed,which overcomes the problem of measuring 1/2 contour holes,and specific size parts need to be equipped with a specific center distance measuring fixture.First,in order to accurately extract the elliptical contour of the box hole edge and the calibration device end face on the image plane from the image,three sub-pixel edge detection algorithms are analyzed in this paper.Based on the feature that the object to be detected is ellipse,this paper proposes the evaluation method of sub-pixel edge detection algorithm,designs the evaluation experiment,and introduces the camera calibration link,which more truly simulates the actual measurement state.Through analysis and comparison,it is determined that Zernike moment method is more suitable for the measurement requirements of this paper.Secondly,this paper analyzes and compares the difference between elliptic geometric fitting and algebraic fitting,and also compare their derivation algorithms under different constraints.The traditional ellipse geometry fitting algorithm is optimized to calculate the orthogonal contact points,and the calculation speed of the algorithm is improved.The method of simulation experiment is used to analyze and compare the calculation accuracy of various algorithms from the aspects of anti-noise ability and ability to fit the complete ellipse by using local contour data points.Thirdly,the paper proposed a visual measurement model of box aperture and center distance.Two measurement methods are designed according to whether there is a chamfer at the edge of the box holes.And during these measurement process,a solution was given based on the structural characteristics of the calibration device in order to improve the accuracy and efficiency.(1)The coordinate point set in the region of interest is mixed with the coordinate point set in the non-interest region.(2)The set of coordinate points in the region of interest needed to be grouped.Finally,We established a platform for visual measurement,and the measurement model proposed in this paper was also verified by experiment.The errors found in the experiment were analyzed and improvement measures were proposed.This paper has researched significance on the application of visual measurement technology of box aperture and center distance in engineering practice.