Design of Adaptive Steganographic Schemes for Digital Images

Most steganographic schemes for real digital media embed messages by minimizing a suitably dened distortion function. In practice, this is often realized by syndrome codes which oer near-optimal ratedistortion perfor-mance. However, the distortion functions are designed heuristically and the resulting steganographic algorithms are thus suboptimal. In this paper, we present a practical framework for optimizing the parameters of additive distortion functions to minimize statistical detectability. We apply the framework to digital images in both spa-tial and DCT domain by rst dening a rich parametric model which assigns a cost of making a change at every cover element based on its neighborhood. Then, we present a practical method for optimizing the parameters with respect to a chosen detection metric and feature space. We show that the size of the margin between sup-port vectors in soft-margin SVMs leads to a fast detection metric and that methods minimizing the margin tend to be more secure w.r.t. blind steganalysis. The parameters obtained by the NelderMead simplex-reection algorithm for spatial and DCT-domain images are presented and the new embedding methods are tested by blind steganalyzers utilizing various feature sets. Experimental results show that as few as 80 images are sucient for obtaining good candidates for parameters of the cost model, which allows us to speed up the parameter search