基于DWT和SVD的彩色图像数字水印算法研究  

Study of Digital Watermark Algorithm on Color Image Based on DWT and SVD

作　　者：梁欣[1] LIANG Xin(Xi'an Satellite Control Center,Xi'an 710043)

机构地区：[1]西安卫星测控中心

出　　处：《计算机与数字工程》2019年第8期2014-2017,共4页Computer & Digital Engineering

基　　金：国家青年科学基金项目“微处理器硬件木马旁路检测技术研究”（编号：61602505）资助

摘　　要：结合离散小波变换和矩阵奇异值分解优点提出基于二者的彩色图像数字水印算法,先将载体和水印图像进行颜色模型转换,对载体亮度分量进行二级小波分解,然后选择子带做奇异值分解,把水印分量嵌入到奇异值分解的对角阵中,最后进行小波反变换和格式转换得到含水印图像,同时对水印嵌入Cb、Cr分量时的情况进行仿真实现。通过仿真实验证明在满足不可见性条件下,算法对常见攻击表现出较强鲁棒性,对于提取水印的对角线失真现象提出改善方法且改善效果显著。Combination of discrete wavelet transform and matrix singular value decomposition advantages,the color image digital watermarking algorithm is put forward based on the two.First the carrier and the watermark image are converted to the color model,then to the secondary carrier of luminance component wavelet decomposition is performed,and then the subband is selected to do singular value decomposition,and then the watermark component is embedded in the diagonal matrix of singular value decomposition.By the wavelet transformation and image format conversion,the watermark image is obtained,at the same time watermark embedding Cb,Cr component is simulated and realized.Through the simulation experiments prove that under the condition of satisfy invisibility,algorithm of common attack shows the strong robustness,to extract the watermark diagonal distortion phenomenon also puts forward improving methods and improve the effect is remarkable.

关 键 词：数字水印 离散小波变换 矩阵奇异值分解 鲁棒性 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]