一种空域与小波域相结合的水印算法

提出了一种空域与小波域相结合的水印算法．算法采用有意义的二值灰度图像作为水印，首先将水印分成2部分，一部分包含重要信息，另一部分包含次要信息．然后利用二维离散小波变换对原始载体图像进行多级分解，并对含重要信息的水印分量进行Arnold变换，以达到置乱图形和消除空间相关性韵目的，接着将水印数据叠加到载体图像的小波分解系数上，实现水印的嵌入．对于含次要信息的水印分量则通过LSB算法嵌入到经逆向小波变换生成的图像的空域中．实验证明：该算法对／PEG有损压缩、噪声和剪切等常见水印攻击显示出了较强的鲁棒性．

Combinational watermarking algorithm in spatial and wavelet domains

A new watermarking algorithm for digital images was proposed based on spatial and wavelet domains. The algorithm uses binary gray scale image as watermarking. First, the watermarking was divided into two parts, one containing important information and the other less important information. The primary carrier image was made a multi-stage decomposition and watermarking component containing important information was performed a Arnold transform so as to scramble image and eliminate spatial correlation. Afterwards the watermarking data was overlaid on the wavelet decomposition coefficients of the original image to realize watermarking embedding. By LSB algorithm, the other part containing less important information was embedded into the spatial domain of the image generated by inverse wavelet transform. Experiment shows that the algorithm is robust against JPEG compression, image processing noise and cutting attacks.