The Research Of Extract And Recognize Film Line Defect System

Welding is the base of modern industry, the quality of welding line is directly affected of the products using life. Its have important real meaning and economic value to effectively detect the defect. Radiography is an important method of non-destructive testing (NDT), in industrial production bring out large number of film on hand, if we simply use our eyes to view the films, workloads are huge, the eyes are easily hurt by glare, the result of the film viewing largely influenced by subjective factors, moreover, the films are hard to keep, bring inconveniently to inquiry and check.Computer-Aided film viewing is a sort of brand-new viewing style, it can greatly improve our efficiency, and effectively overcome misjudgment and omission of artificial viewing because of researchers' different technique knowledge ,experience and different external conditions, leading to the viewing objectively, scientifically and standard. But when we input the welding line image into computer, it has much noise, the edge of defect fuzzy and lack contrast and so on, so these weak points make drawing the defect information, segmentation and recognition much difficulty.This article is based on the research of predecessors, because of in the processes of film digitize, image process and defect recognize there have some problem, so I give a system of drawing defect from welding line image and recognize. The main work is as follows.(1) Studying the basic theory of CCD and image card, designing the film digital hardware system, through experiment verification, the image meet require of film viewing. Design the image scale, convenient of convert the numerical image size as actual size. Easy to storage and inquiry the film image, I use image database.(2)Studying various the image process, image segmentation and edge detection arithmetic, through the experiment comparison, give optimized image process flow, lay good foundation of withdraw the character and defect recognize.(3)Studying the common classification and grade of the welding line, and according to different defects we use different character to withdraw the parameter. We define length, width, area, and circumference and relative gray scale as the defects parameter.(4)Studying the classification principle of the support vector machine (SVM), and according to the actual of film, design the SVM classifier. From the experiment, the classifier can classify gas cavity, circinal slag inclusion, long slag inclusion and crack defects. Mainly reach require of computer aided film viewing.