A system for automated visual inspection of ceramic tiles

The paper deals with an image processing system for visual inspection of ceramic tiles aimed at automated surface quality analysis. Such a system is currently under development within a project devoted to defining and implementing an advanced equipment for high-resolution color-image acquisition, based on a newly developed sensor, implementing an appropriate image analysis algorithm for color measurement, texture characterization, and defect detection in ceramic tiles, and integrating them into an automated inspection system. Results show that good performances in defects detection can be obtained by using a set of adaptive rank-order filters driven by some local features, such as anisotropy or local business, even when textured tiles are concerned.<
>