Research And Application Of USB-C Flatness Detection System Based On Machine Vision

With the rapid development of the electronics industry,the competition in the electronics industry has become more fierce,accelerating the upgrading of products,and the increasing requirements for manufacturing accuracy.Quickly and accurately obtaining the appearance and size data of microelectronic products has become one of the important research contents in the field of measurement.For the pin flatness of the USB Type-C conversion connector(hereinafter referred to as USB-C),when the pin flatness error exceeds the limited range,it will cause short circuit or open circuit failure and affect the function of the workpiece.Therefore,the manufacturer must develop a production line with an intelligent visual inspection system to solve the problems of low manual line efficiency,high error rate,and slow speed of optical image measuring instrument(OGP),which are not suitable for mass production.In order to meet the needs of modern industrial non-contact,high-precision and fast flatness measurement of micro-objects,this paper uses a 3D laser measurement method to develop a USB-C flatness detection system based on machine vision.The main research contents and innovations of this paper are as follows:1.In order to solve the USB-C pin flatness measurement index requirements set by enterprises,this article first introduces the detection object,analyzes the characteristics of the detection object,builds the overall framework of the system,and then selects the visual sensor,software development platform and image processing function library according to actual needs,completes the system feasibility analysis and image acquisition,laying the foundation for subsequent methods and system research.2.In view of the problem of high system misjudgment rate,in order to improve the accuracy of later measurement,this paper studies the image denoising method,by comparing the mean filtering and fractional integral denoising methods,the 3D laser image denoising algorithm of the cubic spline interpolation function is studied,and the effectiveness of the denoising effect is experimentally verified to meet the user's production needs.3.In view of the problem of low system accuracy,in order to improve the detection accuracy,this paper first studies the template matching technology,by analyzing the advantages and disadvantages of the template matching technology based on gray scale,the template matching technology based on geometric features is used to experimentally verify the superiority of the design method;Secondly,the common edge feature extraction methods are analyzed,and the improved Sobel operator's edge feature extraction method is studied.Through simulation comparison,and in practice,the experimental detection verifies that the accurate edge extraction can meet the needs of users.4.In order to verify the feasibility of constructing the experimental platform and the effectiveness of the method,according to the measurement principle and demand analysis,the feature plane composed of the feature points in the feature area will be determined and the reference plane will be selected by the study least square method.Finally,through the development of the main software modules,the accuracy of the physical measurement can reach 0.1 microns,which can achieve the purpose of research,save labor costs,and meet user requirements.The innovation of this paper is to study a denoising method of cubic spline interpolation function and image feature extraction method of improved sobel operator,and apply it to USB-C flatness measurement to meet the user's production needs.The system provides a measuring device system and corresponding measuring means for three-dimensional non-contact,high-precision,multi-dimensional measurement of the geometric dimensions of micro-objects.