一种基于自适应量化嵌入器的隐写方法

感知质量和容量一直是图像隐写算法设计中难以折中的两个要素,如何取得更好的综合性能已吸引了很多研究者的兴趣.提出一种基于自适应量化嵌入器的隐写方法.这一方法通过改变某一载体信号所在子空间的索引来调制秘密信息,自适应特性体现在决定着子空间尺寸的分割码本的设计上.将这种隐写方法用于以2×2图像块为单元的数字图像隐写算法设计中,并将分割码本设计得具备HVS特性以取得较好的折中效果.实验表明,算法在感知质量和容量方面具有较好的综合性能.

A Steganographical Method Based on Adaptive Quantization Embedder

The capacity and the imperceptibility are the two important factors which are hard to compromise in the design of image steganographical algorithm.This problem attracts the interests of lots of researchers about how to obtain the better synthetical performance.In this paper,a steganographical method based on adaptive quantization embedder is proposed.In the method,the secret information is modulated by modifying the index of some cover signal on the sub-space containing it.And,the adaptation is reflected by the design of the partition codebook which determines the size of the sub-space.The method is also applied to an image steganographical algorithm with 2×2 pixels block as the embedding unit,and a partition codebook is built with HVS characteristics.Experimental results show the proposed method has better synthetical performance on the capacity and the imperceptibility.