摘要
提出了一个基于DWT的有意义稳健图像水印算法 .1)采用基于最小距离的水印重同步技术 ,能抵抗RST ,shearing ,generallineartransformation等几何攻击 ;2 )采用多分辨率匹配和粗细搜索相结合的方法 ,有效降低算法的计算量 ;3)应用二维交织技术和BCH编码 ,具有纠正随机错误和突发错误的能力 .把一个 5 36bits信息构成的字符串水印嵌入到 5 12× 5 12× 8bits的图像中 ,所实现的水印在抵抗StirMark攻击中达到了较好的性能 .相对于目前的一些水印算法 ,具有隐藏数据量大且好 ,对抗几何变换具稳健性等优点 .
An image watermarking algorithm in DWT domain is proposed which is robust against geometric distortion. We hide 536 information bits in a 512×512×8 image. The merits of the proposed algorithm are as follows: 1) By introducing distance between the attacked image and the original image, we can resynchronize the data extraction based on the minimum distance; 2) Using multi-resolution matching and coarse-fine searching to prune the searching space, the computation of our algorithm is drastically reduced; 3) With BCH coding and 2-D interleaving technology the algorithm can correct some random errors and bursting errors in detection. The watermark thus generated is invisible and performs well in StirMark test. Compared with other watermarking algorithms reported in the literature, the algorithm is more robust, especially against geometric distortion, while having much higher bit rate.
出处
《自动化学报》
EI
CSCD
北大核心
2003年第5期709-715,共7页
Acta Automatica Sinica
基金
SupportedbyNationalNaturalScienceFoundationofP .R .China (6 9975 0 11,6 0 172 0 6 7and 6 0 1330 2 0 ) ,"86 3"Pro gram (2 0 0 2AA14 4 0 6 0 ) ,NaturalScienceFoundationofGuangdong (980 4 4 2 ,0 2 175 8) ,FoundationofMinistryofEducation
关键词
图像水印算法
几何变换
BCH编码
水印重同步技术
Algorithms
Error correction
Feature extraction
Image coding
Optimization
Robustness (control systems)
Wavelet transforms