Research On The Defect Detection And Recognition Method Of Metal Mobile Phone Backplane Based On Machine Vision

With the continuous improvement of people's living standards,smart phones have gradually spread to everyone.In recent years,smart phones made of metal materials have been popular because of their excellent quality.Among them,metal backplane,as an important part of smart phones,will inevitably produce surface defects after a series of manufacturing processes,which will inevitably affect the quality of mobile phones.Therefore,the detection of appearance defect is an essential link.At present,the main detection method is to use manual visual inspection.It has the disadvantages of slow speed,low efficiency and different detection standards,which is difficult to meet the needs of industrial production.Machine vision inspection technology has the advantages of high efficiency,automation,and intelligence,which can well overcome the shortcomings of manual detection.Based on machine vision technology,this paper takes the surface defects of metal mobile phone backplane as the research object,designs a defect detection and recognition system for metal mobile phone backplane,and studies and implements the defect detection and recognition method for metal mobile phone backplane.The main research work is as follows:(1)This paper introduces the research background and significance of this subject,summarizes the machine vision technology,introduces the hardware,software system and specific application cases,and analyses and summarizes the domestic and international research status of machine vision technology in the field of surface defect detection and the specific direction of mobile phone backplane surface defect detection.(2)This paper introduces the research object,analyses the defect types and detection standards,gives the overall scheme of defect detection and recognition system for metal mobile phone backplane,introduces the design and implementation of mechanical structure,visual imaging system and electrical control system in detail,and carries out hardware analysis and selection for transmission device and imaging system components.(3)This paper studies the defect detection method of metal mobile phone backplane,introduces the overall flow of defect detection algorithm.For image region of interest extraction,an image rotation correction method based on contour offset angle and an image segmentation method based on pixel point gray projection are studied,the division of different regions of mobile phone backplane are implemented.For non-logo region,a defect detection method based on sliding window and Gaussian mixture model is proposed,and for logo region,a defect detection method based on normalized correlation coefficient template matching is implemented.The experimental results show that the defect detection method studied in this paper can accurately detect the defect of metal mobile phone backplane.(4)This paper studies the defect recognition method of metal mobile phone backplane.Based on the analysis of classical recognition methods,a defect recognition algorithm of mobile phone backplane based on cascaded support vector machine is studied and implemented.The algorithm extracts six features of defect,long-short axis ratio,duty cycle,area,circumference,roundness and gray gradient peak value.The radial basis function is used as the kernel function to train the SVM classifier,and three SVM binary classifiers are cascaded to recognize and classify four kinds of defects.The experimental results show that the recognition accuracy of the proposed defect recognition algorithm reaches more than 95%.(5)This paper designs and develops a software system for defect detection and recognition of metal mobile phone backplane,the overall design scheme of the software system is given,and the interface of the software system and the workflow of the software are introduced in detail.