一种基于视觉模型的DCT域公开水印算法  

DCT Based Public Watermarking Algorithm with Visual Model

作　　者：刘红军[1,2] 杨胜[1] 夏太武[2] 

机构地区：[1]湖南大学计算机与通信学院,长沙410082 [2]邵阳学院信息工程系,邵阳422000

出　　处：《计算机科学》2009年第7期281-283,共3页Computer Science

基　　金：国家自然科学基金项目(60603053)资助

摘　　要：提出了一种基于视觉模型的DCT域公开水印算法。算法在充分考虑人类视觉系统屏蔽特性的基础上,通过改变DCT分块内DC分量与低频AC分量之间的大小关系以自适应嵌入水印,水印提取不需要原图像。实验结果表明,算法保证水印不可见的同时,对有损压缩和一些通常的图像处理操作是鲁棒的。A lossy compression-resisting public watermarking technique was presented. A watermark with visually recognizable pattern was embedded into the host image by modifying the polarity between DC value and low-frequency AC value in each 8× 8 DCT block,and the watermark was adapted to the image by utilizing the masking characteristics of the human visual system（HVS）, watermark extraction processes don＇t need the original image. The algorithm provides very good results both in terms of image transparency and robustness. The experimental results show that the proposed technique successfully survives lossy compression and common image processing operations.

关 键 词：有损压缩 离散余弦变换 数字水印 人类视觉系统 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] TP309[自动化与计算机技术—计算机科学与技术]