DEFECT SEGMENTATION TECHNIQUES FOR CERAMIC WALL TILE

Abstract

This paper discussed the image processing algorithms that can be used to segment defects from the acquired ceramic tiles image. The paper proposed two gray-level threshold algorithms, a fixed threshold algorithm and an adoptive interpolating threshold algorithm. The two thresholding algorithms are tested across a set of test images and their performance compared using the area under receiver operator characteristic curves. The experimental results showed that the fixed threshold algorithm produces the receiver operator characteristic curve that encloses the greatest area and provides promising results in preference to the adaptive threshold algorithm.