Development of the automatic system for radiographic film interpretation (II) (The automatic identification of kinds of weld defects) (In Japanese: English Abstract) : Katoh, Y.; Okumura, T.; Itoga, K.; Harada, T.; Sugimoto, K.; Michiba, K.; Kawano, S. Journal of JSNDI, Vol. 41, No. 4, pp. 196–206. (Apr. 1992)

• Publisher: Elsevier Science
• Year: 1996
• Tongue: English
• Weight: 159 KB
• Volume: 29
• Category: Article
• ISSN: 0963-8695
• DOI: 10.1016/0963-8695(96)84894-8

✦ Synopsis

Digital image processing and artificial intelligence technologies are applied to the radiographic inspection of weld joint. In this paper, the method of categorization of weld defects is introduced. The senior inspectors often use the specific knowledge acquired from their experiments for the identification weld defects in testing. In this study, two methods are used in conjunction for the categorization of weld defects. One of them is based on the data base of the defect's feature, and the other is based on the knowledge base acquired from imerviews with inspectors. It is shown that the results inferred by the expert system compare very well with the inspectors' judgements, and weld defects identification ought to be inferred by using the knowledge obtained from those other than films as well as those obtained directly from a film.