Design Of Measuring System For Diameter And Ovality Of High Speed Wire Rod Based On Machine Vision

With the rise of the concept of industry 4.0,intelligent production process has been highly valued by manufacturing enterprises and governments at all levels.In steel manufacturing,wire rod is always count for much,which plays an irreplaceable role in daily life and engineering construction.China's economy develop so fast that the market needs of wire rod is also increasing rapidly.Wire rod production is also developing towards high speed and high precision.In the process of high-speed wire rod production,the outer diameter and ovality of wire rod are the important judgment basis of wire rod quality detection.Therefore,based on machine vision,this paper designs a non-contact on-line monitoring system for the diameter and ovality of high-speed wire rod,which can monitor the size and quality of wire rod in the process of high-speed wire rod production in real time.Through the analysis of production line environment and production demand,this paper designs a three camera detection scheme.In this system,LED backlight and collimating lens are used to provide light source for imaging system,and high-speed area array CCD industrial camera and bilateral telecentric lens are used to image wire rod;After the industrial camera and lens scheme are determined,the imaging system is calibrated to obtain the conversion relationship between the world coordinate system and the pixel coordinate system,and the imaging distortion is restored.After the image collected by area array CCD industrial camera is converted to digital image by A/D,the image is preprocessed by histogram equalization,noise filtering and so on;the image after preprocessing is processed by fuzzy processing.In this paper,the method of constrained least square is studied to recover the motion blurred wire image;in order to recognize the edge of the image with noise and fuzzy interference removed,Sobel operator,Canny edge detection and subpixel edge detection algorithm based on Zernike orthogonal moment are used to obtain the wire edge coordinates;after obtaining the wire edge coordinates,Hough transform is used to obtain the wire edge coordinates in polar coordinate system.The two edge point sets of wire rod are screened and classified;after the two edge point sets of wire rod are obtained respectively,the least square method is used for linear fitting of the edge coordinates of wire rod,and the diameter of wire rod is calculated by the linear equation fitted out,and finally the ovality of wire rod is calculated by the diameter of wire rod in three directions.The conversion relationship between pixel coordinates and actual diameter is determined by camera calibration,and finally the wire ovality is calculated by wire diameter in three directions.This paper mainly completed the following work:(1)This paper explains the background and significance of high-speed wire rod diameter and ovality detection,analyzes the current situation of wire rod production and the requirements of dimension quality monitoring of wire rod production,and introduces several commonly used methods of machine vision in dimension detection.(2)Based on the analysis of the accuracy requirements of wire size and quality inspection and the consideration of the cost of equipment,the design scheme of the system based on the area array CCD imaging monitoring is determined.The selection principle and the specific selection of imaging components such as lens,light source and camera are introduced.(3)This paper introduces the calibration process of the imaging system in this system,the conversion relationship between the world coordinate system and the pixel coordinate system,establishes the camera model,calibrates the camera distortion coefficient,establishes the distortion model,and corrects the image distortion after imaging.(4)By means of image processing,histogram equalization,noise removal and other preprocessing of the collected digital image are carried out,and the restoration and reconstruction of the image under the condition of blur caused by the high-speed movement of the target are studied.(5)A complete algorithm for detecting wire diameter and ovality is designed,including image preprocessing,wire edge recognition,edge coordinate classification,edge linear fitting and so on.(6)Three edge detection algorithms are studied,including Sobel operator,Canny edge detection,and subpixel detection algorithm based on Zernike orthogonal moment.The edge location accuracy,anti noise ability and the final calculated data accuracy of the algorithm are analyzed.(7)After the feasibility and correctness of the algorithm are verified by Matlab,the algorithm of image processing in this system is realized by C++ programming language,and the exquisite user software is written by using GUI Library of Qt.The interface is exquisite and interactive,and the algorithm is optimized by CPU optimization instruction set,which reduces the time consumption of image processing and enhances the real-time performance of the system.