一种基于多级DCT和SVD的鲁棒数字水印算法

摘要

从提高数字水印鲁棒性的角度出发,算法在充分考虑传统离散余弦变换不能有效抵抗旋转几何攻击的基础上,提出一种基于多级离散余弦变换(DCT)和奇异值分解(SVD)的数字水印算法.将离散余弦变换与奇异值分解相结合,实现变换域算法与空间域算法的综合使用.通过峰值信噪比(PSNR)和归一化相关系数(NC)对算法的性能进行评价.实验结果表明,该算法可以有效地抵抗有损压缩和一些通常的图像处理操作,鲁棒性较好.%From tbe perspective of improving the robustness of digital watermarking, the algorithm takes into consideration Ihe insufficient resistance against rotated geometry attack with traditional discrete cosine transform, and presents a digital watermarking algorithm based on multiple level discrete cosine transform (DCT) and singular value decomposition ( SVD). By combining DCT and SVD together, it realizes the comprehensive application of transform domain and space domain algorithms. Through peak signal to noise ratio (PSNR) and normalized correlation coefficient ( NC) the paper evaluates the performance of the algorithm. Experimental results show that this algorithm can effectively resist lossy compression and a few common image operations with quite good robustness.