基于Tetrolet⁃Radon变换与Schur分解的盲水印算法  

Blind watermarking algorithm based on Tetrolet⁃Radon transform and Schur decomposition

作　　者：吴捷[1,2] 马小虎[1] WU Jie;MA Xiaohu(Soochow University,Suzhou 215006,China;Taizhou Polytechnic College,Taizhou 225300,China)

机构地区：[1]苏州大学,江苏苏州215006 [2]泰州职业技术学院,江苏泰州225300

出　　处：《现代电子技术》2021年第5期41-46,共6页Modern Electronics Technique

基　　金：国家自然科学基金(61402310);江苏省自然科学基金(BK20141195)。

摘　　要：为了增强水印算法的鲁棒性及提取水印的便利性,提出一种基于Tetrolet⁃Radon变换与Schur分解的盲水印算法。首先对原始图像进行一级Tetrolet变换,再将得到的低频子带进行分块Schur分解,从分解得到的上三角矩阵中选择特征值最大的元素嵌入水印,并利用最大特征值元素和均值之间的稳定关系来设计密钥,利用密钥实现水印的盲检测。提取水印信息时,利用Radon变换对含水印图像进行几何校正后再进行水印提取操作。实验结果表明,文中算法在保证水印不可见性的前提下,能够有效抵抗各种类型的攻击,并且在绝大多数攻击下NC值均大于0.98。In order to enhance the robustness of the watermarking algorithm and its convenience of watermark extraction,a blind watermarking algorithm based on Tetrolet⁃Radon transform and Schur decomposition is proposed.In the algorithm,the original image is subjected to one⁃level Tetrolet transform,and then the obtained low⁃frequency subband is subjected to partitioning Schur decomposition.The element with the largest eigenvalue is selected from the upper triangular matrix obtained by the decomposition.And then,watermark is embedded into the element.The stable relationship between the maximum eigenvalue element and the mean value is utilized to design the secret key,which is used to realize the blind detection of watermark.Radon transform is used to perform geometric correction of the watermarked image,and then extract the watermark information.Experimental results show that the proposed algorithm can effectively resist all kinds of attacks on the premise of ensuring the invisibility of the watermark,and its NC value is greater than 0.98 under most attacks.

关 键 词：盲水印算法 Tetrolet⁃Radon变换 SCHUR分解 密钥设计 水印提取 盲检测 

分 类 号：TN915.08-34[电子电信—通信与信息系统] TP393[电子电信—信息与通信工程]