Texture classification based on statistical steganographic techniques

Texture based features used for content based retrieval of images and videos should be invariant to various distortions such as noise corruption and compression. In this paper we apply statistical steganography techniques to extract robust texture features. Two texture classification methods, directional steganogaphy histogram (DSH) method and texture decision tree (TDT) method, are presented for texture analysis and classification. Experiments show that the proposed methods can achieve high accuracy rate and also work well even when the query textures are distorted by noise corruption or compression.