The Research On Intelligent Recognition Of Welding Line Defects On X-ray Film

With the development of machine industry and the appearance of automatic welder, X-ray detection of welding line has being faced with more opportunity and challenge. Judging X-ray film's quality is an important part in X-ray detection. Traditionally, the judging work is done by judging worker, the man analyses many defects of welding line on X-ray film and identifies the quality grade of film. The traditional method has many disadvantages, for example, bad effect from man's quality, low efficiency, fraying film and so on. It does not meet requirement of productivity gradually. So many new judging technologies of film have been presented. Intelligent judging technology is a hot research field.Solved the digitization of X-ray film, Digital image processing become an important part in the intelligent film judging. Much research about the digitized film has been done in this paper, these research are preprocess of image, analysis of image, extracting defect's characteristic information and classified defect.Preprocess of image is the base of following image processing. After analyzing image's noise model, self-adaptive median filtering is presented in this thesis. High frequency intensification filtering aims at X-ray film's character of low gray contrast and faint edge, it intensified object's edge in image, to a certain extent it widen gray distributing. Edge detector is an effective method to line out defect. The paper analyzed the theory of some edge detecting algorithms, and presented a new edge detecting method that is based on morphological gradient. Emulation results indicated that the edge detection on X-ray digital image is one effective method. Selection and extraction of defect's characteristic parameter are precondition for defect classification, it directly affect the result. Studied character of defect, author presented an eigenvector that factually reflected defect's essential properties and their respective computation. On solution of defect's classification, the self-organized and self-adaptive back propagation (BP) neural network algorithm is used to intelligently recognize the defect of welding line on X-ray film.The above-mentioned solution to intelligent defect recognition of welding line on X-ray film passed emulation and has a perfect effect.