| With the rapid development of the manufacturing industry and the development of the Made in China 2025,circular arcs are constantly emerging.How to realize intelligent,innovative,green and accurate and rapid measurement of circular arc contours has also become an urgent problem to be solved.Due to the special shape of some circular arc workpieces and the complex spatial position,two-dimensional measurement methods are difficult to measure because of their limitations.Due to its advantages of non-contact,simple structure and high precision,the three-dimensional measurement of binocular vision has become a hot topic in recent years.This article intends to conduct a related study on a series of methods for binocular vision measurement of circular arc contours of workpieces.The main research contents of this article include:(1)A complete set of binocular visual arc contour measurement system was designed to solve the problems of special circular arc contours and complex spatial positions.By analyzing the basic principle of binocular vision,the hardware part of the three-dimensional measurement system was firstly constructed,and then the software part of the binocular visual arc contour measurement system was completed.Finally,the accuracy experiment and error analysis of the three-dimensional measurement system were completed.(2)In view of the special contours of workpiece arcs and the difficulty of extracting contour features,this paper bases on the edge detection of Sobel operators on the basis of the characteristics of edge contours of workpieces,and carries out relevant improvements to extract the edge contours of arcs.Firstly,the fusion gray value is generated by the adaptive convolution operation and the DLBP texture feature,and then the edge points detected by the Sobel operator are further filtered out by the fusion gray value.At the same time,the normal distribution of the histogram of polar coordinate distribution is processed,on the one hand,the interference of the noise point is eliminated when the background is complex,and on the other hand,the outer contour feature and the inner contour feature can be further distinguished.(3)For the problems of high similarity and low precision of edge contour features,the paper improves the shape context matching algorithm to ensure the matching precision of edge contour points.Firstly,the weights of DLBP texture features are introduced to overcome the effects of noise and illumination changes.Secondly,the similarity measure of joint polar lines is proposed to reduce the complexity of the search space and enhance the real-time performance of the algorithm.At the same time,matching point pairs have higher accuracy.(4)Aiming at the problem that the traditional space arc fitting has low robustness and low fitting accuracy,this paper proposes a robust spatial arc fitting method.Firstly,based on the Lagrangian multiplier method,the objective function is established based on the constraint of the plane condition,and then the scheme of the fitting equation directly based on the spatial arc is derived.Secondly,the error tracking point is eliminated by using the RANSAC method,thereby improving the fitting accuracy.(5)This paper introduces the camera's imaging model and stereoscopic vision model,and completes the calibration of the camera's internal and external parameters.At the same time,in the process of digital image acquisition,noise interference and the camera placement position deviation and other issues,this article preprocesses the original image,including image stereo correction,image denoising and image background processing.The experimental results show that the system has good stability and high measurement accuracy.The relative error of ordinary arc workpiece measurement is about 0.003,the relative error of complex arc workpiece measurement is about 0.01,and the repeatability accuracy of the system is less than 0.2mm.This paper has strong theoretical research significance and engineering practice value. |
PDF Full Text Request