基于离散小波变换和奇异值分解的数字水印算法  

Digital Watermarking Algorithm Based on Discrete Wavelet Transform and Singular Value Decomposition

作　　者：祁清[1] 崔新春[1] 马秀峰[2] 

机构地区：[1]曲阜师范大学信息技术与传播学院,日照276826 [2]曲阜师范大学印刷学院,日照276826

出　　处：《中国印刷与包装研究》2012年第6期10-16,共7页China Printing Materials Market

基　　金：教育部人文社科项目--基于生物特征水印的数字版权保护关键问题研究(No.11YJCZH021);山东省教育科学规划课题--以数学建模竞赛为引领的大学生创新能力培养机制研究(No.2010GG006);山东省高等学校科技计划项目--数字水印中的优化算法研究(No.J10LG04)

摘　　要：当前数字作品的版权保护问题备受关注,数字水印是保护数字作品版权的有效手段。本研究针对大容量水印嵌入问题,提出一种基于离散小波变换和奇异值分解的灰度图像数字水印算法。首先,利用离散小波变换将原始图像分为LL、LH、HL和HH四个子带;其次,通过奇异值分解获得LL子带图像的对角矩阵S^l以及水印图像的对角矩阵S^w;然后,将矩阵S^l与S^u进行叠加来完成水印嵌入。该算法的突出特点为嵌入的水印容量大,且算法具有较强的鲁棒性。实验证明,嵌入水印后的图像在经受高斯噪声、椒盐噪声、旋转等攻击后,利用该算法仍能提取到有效的水印图像。Nowadays, copyright protection of digital works has gained more and more attention. Digital watermarking is an effective way to protect copyright of digital works. In this paper, the problem of embedding the larger capacity watermarking into an image was studied, and a robust algorithm based on Discrete Wavelet Transform （DWT） and Singular Value Decom- position （SVD） for watermarking grayscale images was proposed. Firstly, a host image was decomposed by DWT into four sub -bands, namely, LL, LH, HL and HH. Next, the diagonal matrix Sl of the low -frequency LL image and the diagonal matrix Sw of the watermarking image were gained by SVD respectively. Then, Sl and SWwere superposed together to accom- plish watermarking embedding. The outstanding features of the proposed algorithm are that it provides larger watermarking capacity and is more robust than others. Experiments showed that the proposed approach is robust against most of usual at- tacks, such as Gaussian Noise, Salt ＆ Pepper, rotation, etc.

关 键 词：数字图像 水印算法 离散小波变换 奇异值分解 

分 类 号：TN911.73[电子电信—通信与信息系统] TP309.7[电子电信—信息与通信工程]