基于矩阵分解和混沌置乱的数字水印算法

为了提高运算效率，同时保证算法的不可见性和鲁棒性，提出了一种基于矩阵Schur分解的盲水印算法．首先利用混沌原理对水印信息置乱加密，然后将分块载体图像进行离散余弦变换（DCT），利用矩阵分解理论得到对称矩阵，将对称矩阵作Schur分解，通过量化调制完成水印的嵌入．结果表明，该算法运算量小，并且具有良好的不可见性和鲁棒性．

Digital Watermarking Algorithm Based on Matrix Decomposition and Chaotic Scrambling

In order to reduce computational complexity, and also to meet the demand of invisibility and robustness of the digital watermark, a novel blind watermarking algorithm based on Schur decomposition of matrices was proposed. Firstly, the chaotic sequence was used to scramble watermark, then the original image was subdivided block by block, and every block was manipulated by discrete cosine transform （DCT）. After that, symmetric matrices were obtained by matrix splitting method, which were decomposed to get diagonal matrices by Schur decomposition. Finally, the watermark was embedded into diagonal matrices with quantization method. Results showed that the proposed new algorithm not only had less computational complexity, but also had good performance on invisibility and robustness.