| As the main medium for displaying information,LED display has been expanding in its application field,and at the same time,the requirements for its display quality in all walks of life have been increasing.However,as an important technical means to ensure the display quality of LED display,the detection method of display quality of LED display has some drawbacks,such as labor consumption,inefficiency and unilateral results,which can no longer meet the requirements of LED display manufacturers for display quality detection.Therefore,a complete set of LED display quality detection system based on CCD camera is designed to detect the display quality of LED display quickly and accurately.The main contents of this thesis are as follows:（1）Based on the study of the quality detection index of LED display and the imaging principle of LED display on a CCD camera,the corresponding design requirements are put forward and the system design scheme is formulated accordingly.Then,the factors that affect the image quality of the CCD camera,including dark current,distortion and molar lines,are studied,and the corresponding correction methods are studied.Finally,the pre-processing experiments of the CCD camera and the results are analyzed.（2）A fast positioning algorithm for LED light beads is presented.Firstly,a fast ROI cropping method for LED display images is presented to solve the slow speed of conventional ROI cropping.Then,aiming at the slow threshold selection problem of standard OTSU algorithm,an OTSU image segmentation algorithm based on improved sparrow search algorithm is proposed by introducing good point set theory,particle swarm algorithm,sinusoidal cosine algorithm and t-distribution into the standard sparrow search algorithm,and combining it with OTSU algorithm.Next,after using mathematical morphology to process noise,the LED beads are quickly labeled and sorted by the8-direction neighborhood growth algorithm,and the positioning of the LED beads is finally completed.Finally,ROI clipping speed experiment,ISSA-OTSU algorithm experiment and LED bead positioning experiment are carried out,and the results verify the speed of the algorithm proposed in this thesis.（3）Optimize the detection method of display quality of LED display.Firstly,the advantages and disadvantages of traditional calculation methods of LED bead brightness are analyzed.A new calculation method of LED bead brightness is proposed,which combines the centroid of LED bead with improved pixel averaging method,and the brightness value is corrected by using a modified method based on Gaussian filter and data iteration.Then the traditional angle of view detection method in the LED display quality detection index is analyzed,and the angle of view measurement method of LED display based on CCD camera is proposed.Finally,the brightness extraction and angle detection experiments of LED beads are carried out to verify the validity of the proposed method.（4）By completing the five modules: camera module,LED display module,detection module,data management module and rights management module,the software of LED display quality detection system is completed.Testing speed experiment,display quality test experiment and software function experiment on the actual production line verify that the detection performance and function of the system can meet the actual detection requirements. |
