小波域中的鲁棒性盲水印算法

Improved robust blind watermarking algorithm in wavelet domain

提出了自适应的鲁棒性图像水印方案。用密钥选择嵌入的位置,增加水印的安全性,并只对有嵌入要求的图像块进行离散小波变换,这样算法的计算复杂度只与水印容量有关,与宿主图像无关,相对减少了计算的复杂度。在水印嵌入过程中,对相应的有嵌入要求的图像块进行分类来决定不同的嵌入强度,达到了自适应的嵌入要求。为了更好地增加方案的鲁棒性,采用基因算法对阈值进行了优化。实验结果表明,此方案具有很好的透明性,对通常的图像处理具有很好的鲁棒性。

This paper presented a robust adaptive image watermarking scheme. In order to increase watermark security, key was used for the choice of embedded location, and only a pair of requirements of the embedded image blocks is applied discreted wavelet transform, This algorithm complexity was only determined to the watermark capacity, and had nothing to do with the host image. So the scheme reduced the computational complexity relatively. In embedding process, it classified the requirements of the embedded image block to determine the different embedding strength, so that reached adaptive embedded requirements. In order to further increase the robustness of the program, this paper use genetic algorithm to optimize the threshold. Experimental results showed that this scheme had a good transparency and good robustness to common image processing.