【摘要】 在水印算法中用奇异值分解可以提高图像的抗几何失真性,但它对椒盐、高斯、滤波等攻击的抵抗能力却很弱。对原有的基于SVD的水印算法进行改进,提出基于W-SVD的水印算法,即在小波分解的基础上进行奇异值分解。实验证明,改进后的水印算法在保持强的抗几何失真性的同时对椒盐、高斯、滤波等攻击的鲁棒性也有提高。更多还原

【Abstract】 Use singular value decomposition (SVD) in the watermark algorithm may enhance the anti-geometry distortion of the image, but it is very weak to some attacks such as salt, gauss, filter and etc. This paper improves to the original watermark algorithm based on SVD, proposing a new watermark algorithm based on wavelet and singular value decomposition (W-SVD), namely carries on the singular value decomposition in the foundation of the wavelet decomposes. The experiment proved that, the improved watermark algorithm not only maintains the strong anti-geometry distortion ability but also greatly enhances the robustness to the attacks such as salt, gauss, filter and etc.更多还原