Targeted attack and security enhancement on texture synthesis based steganography

We describe an effective and efficient strategy building steganography detector for patch synthesis based steganography, one case of which is reversible texture synthesis based steganography method proposed by Wu et al. (2015). By exploiting the observation that steganography destroys optimization of matching extent between the synthetic patch and optimal candidate patch, we reconstruct the two patches from an overlapped region to extract the existence of optimality, which are distinct between cover and stego images, to form features. Support vector machine (SVM) is implemented for classification. Meanwhile, a variant of Wu et al.'s steganographic method is proposed with reinforced security, by padding redundant regions carrying no message around the periphery of the synthesized image and generating additional candidate patches to increase capacity. Experiments demonstrate that the modified algorithm offers not only better resistance against the state-of-the-art steganalysis methods and steganalytic attack we developed, but also a larger embedding capacity.