Visual Inspection for Fired Ceramic Tile’s Surface Defects using Wavelet Analysis

This paper deals with the visual inspection of ceramic’s tiles surfaces for the purpose of detecting flaws using a wavelet approach. Surface defects in the ceramic tiles are viewed as in-homogeneities in regularity and orientation fields. To improve the homogeneity of batches received by final users and to detect manufacturing defaults, most production lines for ceramic tiles must integrate a visual control stage before the packing operation. The goal of the inspection is not to give a statistical analysis of the production but to classify every tile into quality-constant batches. These tasks are often referred to as visual inspection; Visual inspection procedures have been implemented and tested on a number of tiles using synthetic and real defects. The results suggest that the performance is adequate to provide a basis for a viable commercial visual inspection system. Wavelet decompositions often provide very parsimonious image representations, and this feature has been exploited to devise powerful compression, Denoising and estimation methods. In our work we introduce a hierarchical waveletbased framework for modeling patterns in digital images. This frame work takes advantage of the efficient image representations afforded by wavelets, while accounting for unknown pattern transformations.