Adaptive fractional differential method based on CSGV to extract image texture feature

Abstract

As the sensitivity of the fractional differential algorithm for detail image texture extraction and the difficulty of the best fractional differential order fingding, a novel adaptive fractional differential method is proposed, which can adaptively select the fractional differential order according to the mask window size, definition of fractional differential equations, the composite sub-band gradient vector (CSGV) obtained from the sub-images through a wavelet decomposition of a texture image, and human visual property. The fractional differential operator mask based on G-L formula is designed and realized by employing the adaptive order. The evaluation parameters of image texture feature extraction such as the image information entropy and multi-scale structural similarity (MS-SSIM) are used for quantitative analysis of the extraction method in experiment The experiment results show that for grey texture image this method is able to extract image texture and edge details completely, which approximate the results of optimal fractional differential order and more satisfies human visual sense. It is an effective approach to extract fine texture features of images.