Laboratory of Electronics and Information Technologies (LETI) National Engineering School of Sfax, B.P.W, Sfax,
(Received 14 May 2002; accepted 9 April 2003)
The detection and identification of defects in fairence tiles during manufacture is necessary for an objective evaluation of quality, and for rapidity of processing. To this end, we are working on the automatic classification of surface defects observed on faience tiles.
The range of production is large and tile imprints are diversified. This leads to a complexity in the choice of the classification method and in the parameters to adopt.
After a subdivision of tiles into several types according to imprint, we propose in this paper, for each type, the methodology and the processing realized in order to classify tiles as 1st, 2nd, or 3rd quality.
Paper ID: JTE12364J