基于图像小波变换低频系数的数字水印算法

提出一种基于小波变换且具有图像自适应性的数字水印算法.通过小波变换将图像分解成不同空间与频率的系数,这些系数之间存在一种树结构关系,利用树结构关系对最低频系数进行分类,对不同类采用不同的嵌入对策,将水印信号嵌入在最低层逼近系数中,而现有基于小波变换的算法将水印嵌入在各个方向的中频或高频系数中.实验结果表明,该算法具有不可见性,对于通常的图像失真具有较好的鲁棒性,说明最低频成分并不是水印的禁区.

Watermarking the Lowest Approximation of Wavelet Transformation of Images

An image adaptive digital watermarking algorithm based on discrete wavelet transform (DWT) is proposed. With wavelet transform, the image is decomposed into different spatial and frequency coefficients, among which a tree Structure exists. Using the tree structure, the lowest approximation coefficients are classified into two groups. The watermark is embedded into the lowest approximation components in different way respective to its class while other DWT methods embed watermarks into the middle or high frequency components. Experimental results show the algorithm is invisible and robust to certain common distortions. Thus the lowest frequency component is not prohibition for watermarking.