Objective Ultrasonic Characterization of Welding Defects Using Physically Based Pattern Recognition Techniques
Computer-based methods for analysing ultrasonic data to distinguish between different defect types have been based on a variety of techniques such as adaptive learning , artificial intelligence  and statistical pattern recognition . The uncertain classification reliability of these techniques when applied to a range of realistic defect types has, however, often been a significant practical limitation to their use.
KeywordsFatigue Porosity Welding Peaked
Unable to display preview. Download preview PDF.
- 1.M. F. Whalen, L. J. O’Brien, and A. N. Mucciardi, Proceedings of the DARPA/AFML Review of Progress in Quantitaive NDE, AFWAL-TR-80-4078, 1980.Google Scholar
- 2.L. W. Schmerr, K. E. Christensen, and S. M. Nugen, in Review of Progress in Quantitative NDE, edited by D. O. Thompson and D. E. Chimenti (Plenum Press, New York, 1987), Vol. 6A, pp. 879–887.Google Scholar
- 3.J. L. Rose, J. Nestleroth, L. Niklas, O. Ganglbauer, J. Ausserwoeger, and F. Wallner, Materials Eval., 42, 433–438, 443 (1984).Google Scholar
- 5.S. F. Burch, and N. K. Bealing, paper presented at 21st Annual British Conference on NDT, Newcastle, Sept. 1986.Google Scholar
- 6.B. J. Smith, Brit. J. NDT, 28, 9–16 (1986).Google Scholar