Texture Classification Based on Lacunarity Descriptors

摘要

The present work presents a novel solution to provide descriptors of a texture image with application in the classification of such images. The proposed method is based on the lacunarity measure of an image. We apply a multiscale transform over the power-law relation of lacunarity and extract the descriptors from a window of the multiscale transform selected whose limits are determined empirically. We compare the classification accuracy of the proposed method with other state-of-the-art and classical texture descriptors found in the literature. We also do a brief theoretical summary of lacunarity definition, explaining its excellent performance comprobed in the results.