摘要
提出了基于Contourlet变换的数字图像水印算法。与小波变换不同的是,Contourlet变换采用类似于线段(contoursegment)的基得到一种多分辨、局部化、方向性的图像表示。水印信号通过基于内容的乘性方案加载到Contourlet变换系数。在采用零均值广义高斯分布拟合Contourlet变换系数的基础上,提出采用极大似然估计实现水印的盲检测。依据Neyman-Pearson准则,在给定虚警率的情况下对判决准则进行了优化。实验结果表明在保证水印隐蔽性的前提下,水印对常见的信号处理手段以及几何变换具有很好的稳健性。
A novel robust watermarking algorithm in Contourlet domain was proposed. The Contourlet transform was adopted by virtual of its advantages over the wavelet transform. A flexible multiresolution, local, and directional image expansion was obtained using contour segments. The watermark was inserted through content-adaptive multiplicative embedding. The Contoudet coefficients were modeled as generalized Gaussian distribution （GGD） with zero mean. Then the maximum likelihood watermark detection method was developed. Under the Neyman-Pearson criterion, the decision rule was optimized by minimizing the probability of missing the watermark for a given false detection rate. Experimental results demonstrate that the proposed algorithm is invisible, and robust to signal processing.
出处
《通信学报》
EI
CSCD
北大核心
2006年第4期87-94,共8页
Journal on Communications
