基于小波变换域的数字图像嵌入和提取方法  

Digital image embedding and extracting method based on wavelet transform domain

作　　者：刘琰 周理 LIU Yan;ZHOU Li(School of Information Science and Engineering,Fujian University of Technology,Fuzhou 350118,China;School of Mechanical and Electrical Engineering,Hunan City University,Yiyang 413000,China)

机构地区：[1]福建工程学院信息科学与工程学院,福州350118 [2]湖南城市学院机械与电气工程学院,湖南益阳413000

出　　处：《沈阳工业大学学报》2019年第1期68-72,共5页Journal of Shenyang University of Technology

基　　金：国家自然科学基金资助项目(11302051);福建工程学院校级科研启动项目(GY-K-11-26;GB-K-11-10;GY-H-0902)

摘　　要：针对数字图像水印的嵌入和提取问题,提出了一种基于小波变换域的数字图像水印的嵌入和提取算法.该算法结合了离散小波变换、Arnold置乱变换和奇异值分解.通过Arnold置乱变换对数字水印图像进行变换,对原始图像进行二次离散小波变换,之后分别对置乱后的水印图像和原始图像小波变换中的低频部分进行奇异值分解,对两者的奇异值矩阵进行加性操作实现水印的嵌入过程.结果表明:该算法能够准确实现水印的嵌入和提取功能;嵌入的水印具有良好的隐身性,人眼不能感觉出水印嵌入带来的变化;算法具有较强的鲁棒性,经过椒盐噪声、高斯噪声、JPEG压缩、高斯平滑和裁剪操作等污染及攻击后,都能较好地恢复水印信息.Aiming at the embedding and extracting problem of digital image watermark,a digital image embedding and extracting method based on wavelet transform domain was proposed. The proposed algorithm combined the discrete wavelet transform, Arnold scrambling transform and singular value decomposition. Through the Arnold scrambling transform,the digital watermark image was transformed,and two discrete wavelet transform was performed for the original image. In addtion,the singular value decomposition were carried out for the watermark image after the scrambling and the low-requency part in the wavelet transform of original image,and the additive operation was performed on the singular value matrix of both watermark image and original image to realize the embedding of watermark. The results show that the proposed algorithm can accurately achieve the function of watermark embedding and extraction. The embedded watermark has good invisibility,and the human eye can not feel the change caused by watermark embedding. The algorithm has strong robustness,and can recover the watermark information after the salt and pepper noise,Gauss noise,JPEG compression,Gauss smoothing and cutting operation.

关 键 词：数字水印 小波变换 ARNOLD变换 奇异值分解 水印嵌入 水印提取 鲁棒性 隐身性 

分 类 号：TN391[电子电信—物理电子学]