摘要
基于Contourlet变换对其高频系数进行统计分析,从理论和实验方面证明Contourlet变换的高频系数符合广义高斯分布,采用极大似然估计的方法对Contourlet变换系数的边缘分布和联合分布进行拟合,提出一种Contourlet变换域乘嵌入水印及盲检测框架。实验证明,所提出水印检测器具备良好的检测性能。
Based on Contourlet transformation knowledge foundation, the Contourlet transformation high frequency coefficients is carried on the statistical analysis. The theory and the experiment proves that the Contourlet transformation high frequency coefficient conforms to the generalized Gaussian distribution, adopting the estimation of marginal and joint statistical distribution of Contourlet coefficients via the method of maximum likelihood estimator. Embedded multiplicative watermarking and blind detection frame is presented in the Contourlet transformation domain. Experiments prove watermarking detector with a good detection performance.
出处
《计算机工程》
CAS
CSCD
北大核心
2009年第5期150-152,共3页
Computer Engineering
关键词
CONTOURLET变换
广义高斯分布
乘嵌入水印
盲检测
Contourlet transform
generalized Gaussian distribution
embedded multiplicative watermarking
blind detection