Title: Automatic detection of welding defects

Authors: Roumen K. Kountchev, Stuart H. Rubin, Vladimir T. Todorov, Roumiana A. Kountcheva

Addresses: Department of Radio Communications and Video Technologies, Technical University of Sofia, Sofia, Bulgaria. ' Space and Naval Warfare Systems Center, San Diego, CA, USA. ' T&K Engineering Co., Sofia, Bulgaria. ' T&K Engineering Co., Sofia, Bulgaria

Abstract: A new approach for the automatic detection of welding defects in digital images (X-ray or photos) is presented in this paper. Various algorithms used in the consecutive stages of the image processing – permitting the successful detection of different kinds of defects and ensuring high accuracy – are described. The method is suitable for the analysis of images of pipes, plates, etc. The experimental results obtained with the software implementation of the described algorithms prove the efficiency and reliability of the new approach. This paper points up the advantages of the presented algorithms in comparison with some well-known methods for non-destructive welding control.

Keywords: image processing; image segmentation; noise filtration; background equalisation; welding defects; defect detection; intelligent systems; nondestructive evaluation; NDE; welding control.

DOI: 10.1504/IJRIS.2011.037739

International Journal of Reasoning-based Intelligent Systems, 2011 Vol.3 No.1, pp.34 - 43

Published online: 26 Dec 2010 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article