Surface Defect Detection of Mobile Phone LCD Screen Based on Improved PSPNet
The screen of mobile phone is a key component of smartphones,the quality of screen directly affects the experience of users.Therefore,the screen defect detection of mobile phone has become an important part of industrial production.However,the surface defect detection of mobile liquid crystal display(LCD)screens has the problems of low detection accuracy and a large number of model parameters,which can not meet actual industrial production needs.To solve this problems,this paper studies existing defect detection algorithms and classical semantic segmentation models,and proposes an improved mobile phone LCD screen defect detection model based on pyramid scene parsing network(PSPNet).The model adopts the MobileNetV3 as the feature extraction network,which effectively reduces the model parameters.the multi-scale pyramid pooling module is used to further integrate the multi-scale contextual information,improving the feature extraction ability of the model.It also effectively addresses the issues of small defect si-zes,blurred boundaries,and significant differences in same defect size in screen images.Meanwhile,the attention mechanism is intro-duced to increase the robustness of the model.Experimental results show that the accuracy of the improved model is significantly bet-ter than that of other traditional semantic segmentation models in the surface defect detection of four LCD screens:SQ,Mura,TP,and Line.
mobile phone LCD screendeep learningdefect detectionPSPNetmulti-scale pyramid pooling
万方数据
维普
