Design And Research Of LCD Display Defect Intelligent Correction System

Under the background of the increasing demands of consumers for display products,the competition in the display panel industry is becoming fiercer and fiercer.How to improve the benefits of their products becomes the top priority for panel manufacturers in the future.Liquid Crystal Display(LCD)has become the leading flat panel display product with its excellent display performance.Under the current production technology conditions,LCD products produced by factories have many different defects.With the improvement of uniformity standard,the previous chromaticity uniformity correction methods cannot meet the accuracy requirements gradually.Meantime,major manufacturers still use manual evaluation method as the main method to evaluate the correction effect of LCD at the current stage,which is inefficient and subjective.This paper studies the relevant methods of LCD display defect correction,and solves the above problems by designing and implementing a defect correction system.The main work content of this paper is as follows:1)The LCD image display principle and uniformity correction principle are introduced.According to the LCD correction principle,the overall design scheme of the system is proposed,and the selection of system hardware and the design of software modules are analyzed.2)The chromaticity uniformity correction method of local LCD color response model based on Taylor series expansion is introduced.According to the chromaticity measurement principles of color camera and monochrome camera,the chromaticity measurement errors of the two are analyzed through UT transformation,and the conclusion is drawn that color camera cannot meet but monochrome camera can meet the requirements of industrial high-index measurement.On the basis of the above,an LCD chromaticity uniformity correction method based on channel separation is proposed.It estimates the color response model parameters of the LCD by using the sample data of the LCD monochrome channel collected by the monochrome camera,and realizes chromaticity uniformity correction by using the estimated model.From the experiments,the method in this paper can meet the requirements of industrial high-index.3)Firstly,various defects in LCD production are summarized,and then an LCD grade evaluation method based on brightness grading is proposed by simulating manual correction effect evaluation method.According to the characteristics of different brightness of various kinds of defects,the defects are detected hierarchically through brightness and range,and the number of defects to be detected by each detection unit is reduced.For some rare defects,two data augmentation methods are proposed,and transfer learning is used to reduce the amount of training data needed to be collected.Experimental results show that LCD defects can be accurately identified and evaluated by this method.4)Each functional module of LCD display defect correction system software is designed in detail,and the robustness,yield and productivity of the system are tested in the module factory.The test results show that the system designed in this paper meets the requirements of relevant technical indexes of industrial production.