基于Krawtchouk不变矩和SVD-LWT的双重水印研究

为提高数字图像水印的鲁棒性和透明性,提出了一种基于Krawtchouk不变矩、奇异值分解(SVD)和提升小波变换(LWT)的彩色图像双水印算法。该算法首先从载体RGB图像中提取绿色分量进行两级提升小波变换,利用Krawtchouk不变矩对平移、旋转和缩放的不变性以及Krawtchouk矩良好的局部特性和重构性,计算其低频分量的Krawtchouk低阶矩不变量构造水印系统;然后对其变换后的中频分量奇异值进行分解,嵌入混沌置乱后的水印图像的奇异值作为另一水印系统。仿真实验结果表明,该算法不仅有良好的不可见性,而且对常见攻击和几何攻击都具有很好的鲁棒性,并具有一定的应用价值。

Double watermarking examination based onKrawtchouk moment invariants and SVD and LWT

In order to improve the robustness and transparency of the digital image watermarking, this paper proposed a color image double watermarking algorithm based on Krawtchouk moment invariants, singular value decomposition SVD and the lifting wavelet transform LWT. Firstly, the algorithm extracted green component from the RGB image for two level ascension wavelet transform. The invariance of Krawtchouk moment invariant to translation, rotation and scaling, and more, good local characteristics and reconstruction of Krawtchouk, which could calculate Krawtchouk moment invariants of the low frequency components to structure one watermarking system. And then, the transformation of the intermediate frequency component singular value decomposition, embedded chaos after scrambling watermark image of the singular value as another watermarking system. Simulation results show that this algorithm not only has good invisibility, but also has very good robustness to the common attack and geometric attacks, and it has application value.