摘要
提出了一种结合混沌和图像分块提升小波变换的多功能数字水印算法。首先,将宿主图像分成8×8的子图且对各子图进行2级提升小波变换,将实现版权保护和内容认证功能的不可见水印分别嵌入部分子图小波域二级子图的低频系数和水平细节系数,由广义Arnold映射随机选择嵌入子图。嵌入算法采用一种新的量化算法,并能进行盲提取。然后,在图像空域嵌入实现版权通知功能的可见水印,嵌入算法采用加权融合方式,嵌入权重系数由四维超混沌系统序列确定。实验结果证明了所提出的多功能水印方案的有效性。
A multipurpose image watermarking scheme based on chaotic systems and sub-image lifting wavelet transformation(LWT) was proposed.Firstly, the host image was divided into 8×8 sub-images and each sub-image was performed 2-level LWT.Then, the invisible watermarks for copyright protection and content authentication were respectively embedded into low and intermediate frequency LWT components of some sub-images that were randomly selected by general Arnold map.A new kind of quantization technique was adopted in the embedding process and the extracting process was blind.Secondly, the visible watermarks for copyright notification were embedded into the image spatial domain based on image fusion method, in which the embedding coefficients were generated by the 4D hyperchaotic sequences.Experimental results were presented, showing that the proposed watermarking scheme is effective.
出处
《通信学报》
EI
CSCD
北大核心
2009年第S2期101-104,110,共5页
Journal on Communications
基金
湖南省自然科学基金资助项目(06JJ5098)~~
关键词
提升小波变换
混沌
不可见水印
可见水印
lifting wavelet transform
chaos
invisible watermark
visible watermark