An Optimization Strategy for Improving Security in Steganography
Xiancheng Wu, Shunquan Tan · 2018
To improve the security performance of steganography, many adaptive steganographic algorithms have been presented recently. These algorithms enhance the security mainly by pixel-level's adaption while paying no attention to the image-level's adaption. Image-level's adaption means that different images have different characteristics and thus should be considered with different steganographic strategy. In this paper, we propose an optimization strategy to achieve image-level's adaption based on the texture. In the proposed method, we treat different images with individual steganography strategy according to their textures. For each cover image, a bank of Wavelet filters is assessed and the one to achieve the smallest change of the image, with respect of features used in steganalyzer, after embedding the message is adopted. The experiment result shows that the proposed method improves the security significantly compared with the state-of-arts adaptive steganographic algorithms.