Automatic Detection Of Optical Fiber Panel Translucent Image Shadow Defects

Optical fiber panel has some characteristics of clear as well as image transmission.Optical fiber panel as an image intensifier optical input and output window is the most typicalapplications. And it plays an important role in improving the quality of imaging device. It isused more and more widely as the key element of contemporary electron optical imagingdevice. However, the defects of optical fiber panel affect its image quality, it is verymeaningful to research and develope a variety of defects detection system to improve thetransmission quality of optical fiber panel.According to the national standard which stipulates the method in the shadow defectsinspection, this topic researches and develops an automatic system with which the shadowdefects can be tested effectively. First, we detect the shadow detection based on edge defectsthrough the research on the defined of the shadow defect and the gray gradient detectiontheory of knowledge. Secondly, we use the Canny operator to achieve the shadow defect edgedetection on the basis of previous. And we have identified the Canny operator deficiencies inthe shadow defect edge detection process by comparing the experimental results andtheoretical analysis and the way of improvement. We achieve the shadow defect edgedetection by improved Canny operator. Third, we obtained a clear, complete, close, thinshadow defect edge image by using a series of edge optimization algorithm for shadow defectedge for further optimization. Fourth, through the corresponding calculation method, the mainfeatures, parameters and indexes of the shadow defects is offered, at the same time, thequantitative analysis of it is also achieved. The defects can go automatically by the organicintegration and overall control of the software system as well as the hardware system.In this paper, the design of optical fiber panel shadow defect automatic detection systemcan realize the automatic defect detection shadow. The results of these experiments haveshowed the advantages of this automatic system, which is easy to operate, quick and accurateto process data, stable and reliable in performance. The system is valuable in practice.