摘要
研究图像奇异值分解的数字水印算法的不可见性与稳健性问题,针对目前水印算法存在高虛警或不可见性差的问题,为提高水印的不可见性并同时保证其具有较强的稳健性,提出了一种基于Torus映射和图像奇异值分解的半盲图像数字水印算法。算法首先将载体图像分块,对每个子块进行奇异值分解并提取出每块的最大奇异值组成方阵,通过Torus映射变换将该方阵置乱并分块,再通过量化方式将二值水印信息通过这些分块嵌入载体图像。仿真证明,加水印图像的峰值信噪比高且对常见的图像处理都具有良好的不可见性和稳健性。
This paper studies the issues of invisibility and robustness for the SVD based watermarking.Given that many methods have the problems of high false alert or poor invisibility,this paper proposes a digital watermarking scheme based on Toral Automorphism and SVD in order to improve the watermark＇s invisibility and keep its good robustness as well.The original image is divided into blocks and SVD is performed to each block.Then,the largest SVs of each block are grouped secretly by using Toral Automorphism.Each watermark bit is embedded in the largest SVs of each group by using quantification.The experiment based on this scheme demonstrates that the watermark is of good invisibility and robust for common digital image processing.
出处
《计算机仿真》
CSCD
北大核心
2011年第6期125-128,176,共5页
Computer Simulation
关键词
自同构环
数字水印
奇异值分解
盲水印
Toral Automorphism
Digital watermarking
SVD
Blind watermarking
