Zero Kullback-Liebler Divergence Image Data Hiding

In this paper, we propose a computationally-efficient data hiding method for images which achieves Cachin´s security criterion: zero Kullback- Liebler(KL) divergence. In order to preserve the statistical properties of the cover medium, we change the order of pixels rather than modify their values to embed the hidden message. We then subject the proposed stego method to a higher-order statistics based universal steganalysis algorithm and a new learning based steganalysis that we propose specifically for this hiding algorithm. Experimental results show that our proposed method can prevent statistical detection, when the embedding rate is smaller than or equal to 10%, which is higher than those of other existing data hiding methods. Hence the proposed method is safe even in a practical sense against a steganalysis method designed specifically against this stego method.