基于SVD的图像水印算法缺陷分析及对策

Drawbacks in SVD-based Image Watermarking Algorithm and Its Countermeasures

介绍了矩阵奇异值分解的基本原理,分析了现有基于SVD的图像水印算法的缺陷。这种缺陷导致了提取出的水印结果并不是先前嵌入的水印,这样就无法应用于版权保护。结合最新的SVD研究成果,提出了一种相对可靠的解决办法。

The fundamental of matrix singular value decomposition is introduced, and the existing method in the SVD-based image watermarking and their major drawback is discussed. This flaw leads to the extracted watermark is not the embedded watermark, so they cannot be applied to copyright protection. Combined with the latest SVD research results, a relatively reliable watermarking algorithm based on SVD is proposed.