NDTnet 1998 Aug, Vol.3 No.8

Full-Text - Abstract

In this paper, we present an automatic image analysis procedure for weld defects detection and calibration. The main defects types in aluminium welds are considered: cavities, gas porosities and lack of penetration. Excess of penetration has also to be detected, but is considered as a normal structure. The developed procedure is either adapted to radioscopic images or digitized films, and the results are compared in terms of the smallest detected flaw. On another hand, the results are compared with human interpretation of film radiography (before digitization). This study was tested on aluminium welds of 6 mm thickness sheet. Radiographic films are digitized with a 100 µm sampling step. Radioscopic images are acquired with a minifocus X-ray tube (0.4 mm) and an image intensifier device. The obtained pixel is 200 µm in size.

The image analysis proceeds in several steps: 1. global enhancement of the image (maximal gray-level range, detector drift correction); 2. enhancement of specific features: laplacian filter to detect lack and excess of penetration, matching with a mask dedicated to round shaped objects to enhance cavities, neighborhood analysis around the possible cavities to grow only real ones, specific morphological processing for very small porosities; 3. defect characterizationby computation of all features necessary to classify the weld according to internal specification based on NFA 89 220 French standard. Figure 1 shows the image analysis of a digitized film. The smallest cavity is 0.3 mm in diameter.

Figure 1: a) raw image; b) enhanced image; c) final binary image

On digitized films thesmallest reliably detectedporosities are 0.2 mm indiameter whereas therespective diameter forradioscopy is 0.4 mm. "Reliable detected" means avoiding to detect noise. comparing to thefilm interpretation with a magnifying glass, it appears that an operator is able to detectporosities down to 50 µm in size, but is unable to count and calibrate them reliably.

Thecomparison is still under progress, and the question of using or not a lens is of greatimportance (not prescribed by standards).The conclusion at this point of the work is that an automatic image analysis on radioscopic images can solve a lot of problems in detecting lacks of penetration and cavities down to 0.4 mm, and thus could be used as a first and very fast inspection step. Then, for the other welds not rejected by radioscopy, film radiography could be used and digitized at the adequatesampling step if the operator is unable to calibrate the defects.

Abstract Source:
Book of Abstracts, 7th European Conference on Non-Destructive Testing, 26-29 May 1998, ISBN: 87-986898-0-00

Full-Text Source:
Proceedings of the 7th European Conference on Non-Destructive Testing, 26-29 May 1998, ISBN:

Publication Contact:
7th ECNDT, Park Alle 345, DK-2605 Broendby, Denmark, Fax: +45 46 26 70 11, Email: 101373.3414@compuserve.com