基于高斯混合模型的自适应隐写安全性分析

信息隐藏安全性研究可用以指导隐写算法设计,对信息隐藏来讲具有重要的意义。近年来,为了提高算法的安全性,自适应隐写算法层出不穷,但自适应隐写算法安全性的理论研究并未同步展开。以自然图像的高斯混合模型为基础,论证了在图像复杂度不同的区域采用不同嵌入率的自适应隐写算法的安全性高于非自适应隐写算法,并对论证结果的实际意义进行了分析,得出几点对隐写算法参数设计及图像载体的选择具有指导意义的结论。

Security Analysis of Adaptive Steganography Based on Gaussian Mixture Model

The research on security of data hiding can be used as guidance to design a steganography or enhance the statistical imperceptibility of an information hiding system.In recent years,adaptive steganography aiming at improving security has emerged endlessly.But corresponding theoretical analysis of adaptive steganography has not followed the research on arithmetic yet.Based on the Gaussian Mixture Model(GMM) of real images,the security of adaptive steganography which hides secret with different embedding rate...