基于伪随机序列调制的混合域自适应盲水印算法

在多小波变换和DCT变换混合变换域中,提出了一种基于伪随机序列调制的彩色图像盲水印算法。该算法先将原始彩色图像由RGB空间转换到YIQ空间．选取Y分量多小波变换后的中频区域作为水印的嵌入区域。水印嵌入前先对待嵌入水印区域8×8分块,然后对所有子块按照纹理和能量特征进行模糊分类,最后将置乱后的二值水印信息用两个互不相关的伪随机序列进行扩频调制．嵌入到8×8子块DCT变换后的交流系数上。提取水印时,通过比较两个伪随机数列和待检测系数相关性的大小来提取水印,不需要原始图像参与。实验表明,该算法对噪声、JPEG压缩和几何剪切等常见的攻击操作具有很强的鲁棒性。

Adaptive blind watermarking algorithm based on pseudo-random sequence modulation in mixed-domain

A blind watermarking algorithm based on pseudo-random sequence modulation for color images is proposed in multi-wavelet transform and DCT Hybrid transform domain. In this method, the original color image is converted from RGB space to YIQ space, and muhi-wavelet transform is performed on Y component, the mid-frequency regions are selected as the embedding region. The regions are divided block by block with size 8×8 pixels before watermarking embedding, then all the sub-blocks are classified in accordance with their texture and energy characteristics, finally, the scrambed binary watermark, which are spread spectrum modulated with two unrelated pseudo-random sequence, are embedded in the AC coefficients of 8 × 8 subblocks in DCT domain, the embedding strengths are different according to the results of fuzzy classification. The watermark is extracted by comparing the relativity between two random sequences with the coefficients to be detected, which do not need the original image. Experiments results show that the algorithm can resist common attacks such as noise, JPEG compression and geometric shear with strong robustness.