基于对比度à trous小波的Contourlet变换图像水印算法

Image watermarking algorithm based on Contourlet transform with contrast à trous wavelet

提出将基于对比度的à trous小波代替Contourlet变换中的拉普拉斯金字塔(LP)变换,实现对图像的多尺度分析:对变换后的低频区域分块,计算各个分块的噪声可见性函数(NVF)值,对选出的一系列值较小的分块嵌入水印.实验结果表明,水印算法具有较好的视觉效果,同时对滤波、JPEG有损压缩、噪声等攻击具有较好的鲁棒性.

Replacing the Laplace Pyramid（LP） transform by à trous wavelet based on contrast was proposed to achieve multiscales analysis for image.After transforming the low-frequency area,the NVF value of every block was partitioned and calculated.Then,some blocks with lower NVF value was embedded with watermarking information.Experiment results showed that the algorithm can obtain better visual effect and be robust against attacks such as filtering,compression,noise and so on.