以模數為基礎之可適性藏密技術

ADAPTIVE MODULUS-BASED STEGANOGRAPHY

　　Modulus-based steganographic methods use a modulus to embed multiple message bits in the pixels of the cover image and can be applied to high-capacity applications such as hiding a secret image into a cover image. For security reason, these methods should basically possess the property of imperceptibility. In this paper, we present a steganographic method to meet this requirement. We first propose a generic model for constructing secure modulus-based steganographic methods. Based on this model, an efficient human visual model is used to compute an adaptive modulus for each pixel of the cover image so that the theoretical imperceptibility can be ensured. Experimental results show that the proposed method can efficiently adapt to the human visual system to provide good performance of imperceptibility for high-capacity steganographic applications.