Research On High-precision Image Measurement System

With the fast-growing of computer digital technology,the technological standards of production technology and manufacturing are getting higher and higher.Visual measurement as a kind of non-contact,high precision,fast measurement technology has been used in many industries extensively.Especially in some precision components such as chips,small workpiece detection,with its increasingly prominent advantages,it has become a research hotspot in academic research and industrial application at home and abroad.This paper studies the high-precision image measurement system based on the measurement requirements,and completes the design and implementation of the hardware acquisition system and the software algorithm system.The main tasks of this subject are as follows:First,the working principle of the acquisition system hardware including lens,camera and light source is studied,and the effect and influence of them on the imaging quality of the workpiece are summarized,such as lens distortion and depth of field,camera resolution,light source type and lighting mode.Combined with the measurement demand and cost and other factors,the hardware selection and construction of the acquisition system and image contrast experiment are carried out.According to the imaging effect,the final hardware system design scheme is determined.Secondly,the software algorithm system based on the detection of small workpieces is studied,and the image measurement software is developed.The image measurement software is developed in C++ on the Windows,and Qt is selected as the design framework of the interface interactive software.Based on the function and demands of the highprecision image measurement system,the software architecture includes a display interaction(presentation)layer,an image algorithm layer,a business logic layer,and a core tool layer.Finally,the actual implementation plan of each module is formulated,and the experimental test shows that the software system of this subject not only ensures that the measurement function requirements are met,but the intelligent interaction also provides the convenience of measurement.Thirdly,in the processing of subpixel edge contour extraction,this paper combines the improved Canny algorithm with the gray value model's subpixel edge extraction algorithm.First,the improved Canny operator of the guide filter instead of the Gaussian filter is used to extract the pixel-level edge of the measured object.Secondly,on the basis of pixel level edge,the edge extraction algorithm based on gray value model is used to provide high precision sub-pixel edge contour for measurement.At the same time,in order to solve the measurement accuracy problem of the circle with high curvature in the small workpiece,the discussion of the circular edge is added to the edge extraction algorithm based on the gray value model,and the influence of the pixel gray value of different distances on the value of gray value A and B on both sides of the edge is considered.Fourthly,a high-precision image measurement system is built and calibrated to collect and measure the target parameters of the tiny workpiece.Through data analysis,it can be concluded that the system studied in this topic has better accuracy in the measuring scene of small workpiece and can achieve the speed of 0.1s to complete a single measurement part,and the algorithm has a wider application range.