利用矩阵范数实现的公开水印技术

Public Watermarking Using Matrix Norm

分析和讨论了基于奇异值分解的水印算法的本质,提出了一种无需利用奇异值分解的公开水印算法.该算法基于图像子块的矩阵范数及其量化,步骤简单,与基于奇异值分解的水印算法相比大大减小了计算量;并利用三种不同的矩阵范数进行了大量的实验.实验结果表明,该算法具有很好的水印透明性,对常见攻击(如重采样、平滑、加噪声和JPEG压缩等)具有较强的稳健性.

In this paper, having analyzed the essence of the SVD based watermarking algorithms, we propose a novel public watermarking algorithm without SVD. Basically our approach is based on the matrixnorms of sub-images. Compared with the SVD based algorithms, the new algorithm is quite simple and can be easily implemented with a lower computational cost. We give various numerical experiments for the new algorithm with different kinds of norms. The experimental results with three kinds of norms reported in this paper show that this algorithm has a good transparence of the embedded watermark image and is robust to attacks such as re-sampling, smoothing, noise adding, and JPEG-compression.