基于图像归一化和IWT的鲁棒可逆水印算法  

Robust Reversible Watermarking Algorithm Based on Image Normalization and IWT

作　　者：左兴元 张正伟 李芬芬 李瑶 ZUO Xingyuan;ZHANG Zhengwei;LI Fenfen;LI Yao(Faculty of Computer and Software Engineering,Huaiyin Institute of Technology,Huai’an Jiangsu 223003,China)

机构地区：[1]淮阴工学院计算机与软件工程学院,江苏淮安223003

出　　处：《信息与电脑》2022年第3期87-90,共4页Information & Computer

基　　金：国家统计局科技计划项目资助(项目编号:2018LY12)。

摘　　要：为提高现有可逆水印的鲁棒性,提出了一种基于图像归一化和整数小波变换(Integer Wavelet Transform,IWT)的鲁棒可逆水印算法。首先对待嵌入水印进行置乱处理;其次对载体图像进行归一化处理并确定特征区域,将此特征区域分块并对所有分块进行IWT;再次利用广义差值扩展将水印嵌入到各分块中的中低频,而在高频区域则利用无损压缩算法嵌入辅助信息;最后利用逆IWT和逆归一化(预失真补偿策略)得到含水印的图像。实验结果表明,该算法不仅实现了水印的较高不可见性,同时也提高了水印的抗攻击能力。To improve the robustness of existing reversible watermarking,a robust reversible watermarking algorithm based on image normalization and Integer Wavelet Transform(IWT)is proposed.Firstly,the watermark to be embedded is scrambled;secondly,the carrier image is normalized and the characteristic area is determined,and this characteristic area is divided into blocks and IWT is performed on all blocks;again,the watermark is embedded in each block by generalized difference extension In the high frequency region,the lossless compression algorithm is used to embed auxiliary information;finally,the watermarked image is obtained by inverse IWT and inverse normalization(predistortion compensation strategy).The experimental results show that the algorithm not only achieves high invisibility of watermark,but also improves the anti-attack ability of watermark.

关 键 词：可逆水印 整数小波变换 鲁棒性 归一化 特征区域 

分 类 号：TP309.7[自动化与计算机技术—计算机系统结构]