N.: Leather features selection for defects’ recognition using fuzzy logic

Abstract: In the present work are investigated 12 histogram and statistical features for analysis of leather surface images. A research of the features suitability for surface defects detection is done. For the image analysis was used the quadtree decomposition method – a technique that partitions an image into homogeneous blocks. This method gave the possibility of investigation the changes of the features values depending on the area size. It is proposed a scheme for defective regions identification using fuzzy set theory