Study On The Detection Of Optical Fiber Panel Shadow Panorama And Checkerboard Defects

Fiber panel is an optical transmission devices which was proposed in the middle of1950s.This optical device has advantage of very good light gathering capability and high resolutionwhich make it the mainstream choice ranks among the field of high-quality optical imagingdevice and system design. However, the optical fiber panel is restricted and influenced byproduction process and working environment, so that the optical fiber panel has many defectsinevitably which seriously affect the image quality. These defects are collectively referred toas the optical fiber panel defect. With the development of science and technology, therequirements of image transmission performance of this device from various application areashave become more and more demanding, the problem of detail accuracy is becoming moreserious, therefore, it is very necessary to study the nature of defects and to realize a waywhich can quantitatively detect the defects of optical fiber panel with high speed and highprecision. This paper carried out part of the research work to meet the actual needs of the testof image transmission quality problem.First，this paper studies the nature and characteristics of some defects of optical fiberpanel, take the test method which made by National Ministry of machinery and electronicsindustry as foundation. Then this paper research and put forward the detection theory andanalysis method of spots, dark band, chicken, shadow panorama, the checkerboard lattice.Secondly, set the hardware system parameters and select software system development tools,complete the work of optical fiber panel transparent image collection. Based on the automaticdetection system in the shadow of defects, realize the detection of whole panorama fiber panelshadow defects with the help of image stitching technology and output the main features andparameters of all the shadow defects on the entire fiber panel. And local shadow detectionresults are compared with existing ones, proved the number of shadow distribution is uniformbut the shadow distribution area is not, so the shadow panoramic detection results are morerepresentative. Finally, this paper put forward a set of checkerboard defect detecting methodwhich fill the technology gaps checkerboard detection after researching techniques of imageprocessing, similarity algorithm and the related theory, and also realize the quantitativeanalysis. At the same time, a group of adjacent composite fiber optical fiber panel weredetected and made quantitative analysis. The results show that the method proposed in thispaper is effective and practical.