[article]
Titre : |
Digital system for detecting, classifying, and fast retrieving corrosion generated defects |
Type de document : |
texte imprimé |
Auteurs : |
S. T Kyvelidis, Auteur ; N. Kouloumbi, Auteur ; L. Lykouropoulos, Auteur |
Année de publication : |
2001 |
Article en page(s) : |
p. 67-73 |
Note générale : |
Bibliogr. |
Langues : |
Américain (ame) |
Tags : |
Matériau revêtement Peinture Corrosion Défaut Détection défaut Classification Système numérique Analyse image |
Index. décimale : |
667.9 Revêtements et enduits |
Résumé : |
The in-situ assessment of the degradation of a coating system is often difficult, especially in cases such as underwater piping systems. Furthermore, the results of the optical inspection are qualitative and subjective. The use of digital image processing and storing of corrosion images in databases can eliminate such problems and enhance the assessment of the coating degradation characteristics.
Digital defect detection systems have been used in the industrial domain for the last 20 years to automatically detect and classify defects on various surfaces, such as wood, steel, and textiles, using image processing techniques and decision-making theory.
The proposed system processes digitized images of corroded surfaces by applying similar technologies (digital filters, texture analysis, segmentation techniques, and fuzzy decision algorithms) to automatically identify and classify corrosion generated defects, such as blisters, rust, etc. Furthermore, it employs state-of-the-art database techniques specialized for image content (query by example, content-based retrieval) to effectively store and retrieve these images according to their specific features. |
Permalink : |
https://e-campus.itech.fr/pmb/opac_css/index.php?lvl=notice_display&id=5727 |
in JOURNAL OF COATINGS TECHNOLOGY (JCT) > Vol. 73, N° 915 (04/2001) . - p. 67-73
[article]
|