Research On The Automatic Optical Detection System And Algorithm Of LCD Display Defect

Defect detection is an indispensable process in the production process of liquid crystal display screen,and defect detection technology is as critical key as to reduce the production cost and improve the quality of the product.The traditional detection method is manual detection,which has such disadvantages as great subjectivity,low detection efficiency and precision.Moreover,labor cost has increased year by year in recent years.This method has been unable to meet the requirements of mass production,low cost,high efficiency and high precision in production line.Therefore,the research of automatic optical inspection method with the advantages of fast speed,non-contact and high accuracy is very important for improving the quality of LCD products and promoting the development of industry.In view of the above problems,this paper will study the automatic optical defect detection algorithm based on machine vision,and design a complete set of detection system.First of all,based on the analysis of the working principle of LCD,the types of display defects and their causes,and the requirements of the system design,the overall design was put forward.Two key parts that affect the quality of the image are discussed emphatically:the lighting module and the image acquisition module.The detailed analysis and selection of the hardware architecture were carried out,and the design of the optical acquisition system was completed.Secondly,the related image processing algorithms were researched,including the filtering denoising,binaryzation,morphology algorithm,image correction and so on.By comparing and analyzing the various methods of each algorithm,combined with the actual demand,the optimal method was selected or the new method was put forward to identify the basis for the follow-up defect detection algorithm.Then each step of the defect detection algorithm was descr:ibed in detail in turn.The reasonable establishment of standard template library and preprocessing work can overcome external interference.The image registration is the core of the algorithm,in view of the characteristics of LCD itself,an registration algorithm based on Fourier Merlin transform and feature matching was proposed,which has high registration accuracy and strong anti-interference ability,and the registration accuracy is further improved by weighted average fusion.The morphological algorithm and the proposed local adaptive gaussian weighted binarization method can make a good segmentation of the defects,thus improving the accuracy of detection.Determine the type and location of defects based on regional characteristic parameters and minimum bounding rectangle method.Finally,the software of defect detection was designed,and its function and specific operation were introduced.Verify the detection algorithm and system detection precision through experiment,and the experiments were all quantified by comparing with the algorithm based on the SURF/SIFT.The experimental results show that the proposed detection algorithm,the accuracy and efficiency of the system are better tihan the contrasted algorithm,and the detection accuracy is 97%.It can meet the actual requirements and has a good application prospect.