基于提升小波和奇异值分解的灰度水印算法  

Gray image watermarking algorithm base on LWT and SVD

作　　者：蒋华[1] 张敏[1] 

机构地区：[1]桂林电子科技大学计算机与控制学院,广西桂林541004

出　　处：《计算机应用研究》2009年第8期3028-3030,共3页Application Research of Computers

基　　金：国家自然科学基金资助项目(50175070)

摘　　要：以提升小波变换和奇异值分解的理论为基础,提出了一种新的基于LWT和SVD的灰度图像水印算法。该算法核心思想是先对载体图像进行分块;然后对每块二级LWT后的中高频带继续LWT;再对选取的各频带进行SVD,选取相应的奇异值组成新的矩阵,对新矩阵按规则分块,并再次SVD。通过两次分块、两次LWT和四重使用SVD构造矩阵的方法,有效地将抽取的奇异值重新分配和组合。最后将Logistic混沌置乱后的灰度水印信息加载到组合后的矩阵中。该算法以保证鲁棒性和透明性的良好平衡为前提,提高了嵌入的信息量。仿真实验表明,该算法不仅具有很好的透明性,而且对JPEG压缩、滤波、加噪等常见的图像攻击具有很强的鲁棒性,其整体性能明显优于现有同类的水印算法。This paper proposed a gray image watermark algorithm based on LWT and SVD. Firstly, divided the original image into many blocks. Then implemented the middle-high frequency of two-level LWT with each block by LWT, and applied SVD to the selected sub-frequency, divided the new matrix constructed with singular values into many blocks by rule, which were decomposed with SVD again. Through twice partition blocks, twice LWT and the use of quartie SVD, the matrix would re-allocation and combination for extraction＇ s singular value after SVD, and finally embedded the watermarking image with Logistic chaos scramble into the matrix. The algorithm achieved optimal balance between robustness and imperceptibility, and increased embedded capacity. Experimental results show that the proposed method has a good transparence and robustness a- gainst attacks, such as JPEG compression, filtering, adding noise and other image processing. The method＇ s performance is much better than the existing and similar watermark algorithms.

关 键 词：数字水印 提升小波变换 奇异值分解 混沌置乱 透明性 鲁棒性 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]