SVD和DWT数字水印算法的应用研究

Digital Watermarking Algorithm Based on SVD and DWT

研究数字图像版权保护问题,由于数字媒体在网上易被复制篡改。针对单纯奇异值分解或小波变换水印算法均存在抵抗攻击差的难题,为了更好的保护数字图像版权,提了一种奇异值分解与小波变换相结合的数字水印算法。首先对水印图像进行置乱处理并对原始图像进行分块,从中找到符合要求的最佳水印嵌入子块,然后对所选择的最佳子块进行小波变换,对子块的低频系数进行奇异值分解,最后将水印嵌入各子块的奇异值中进行仿真。结果表明,水印算法能够很好的抵抗多种攻击,水印具有很好的鲁棒性和不可见性,克服了奇异值分解和小波变换水印算法缺陷,为设计提供了依据。

Digital media copyright protection has become an increasingly important issue,and digital watermarking algorithms have attracted much attention.There are different faults in simple pure using singular value decomposition or wavelet transform watermarking algorithm,therefor,a digital watermarking algorithm combination with singular value decomposition and wavelet transform is proposed.The algorithm firstly scrambles the watermark image and partitions the original image regions according to the requirement of finding the best watermarking embedding sub block,then chooses the best sub block wavelet transform and pairs of low-frequency coefficients of the block singular value decomposition,ultimately embeds the watermarking into each sub-block singular value.Experimental results show that the watermark algorithm can resist many kinds of attacks watermarking very well,has good robustness and not visibility,and solves the contradiction between the robustness and the invisibility.