基于压缩感知的数字图像水印算法

压缩感知理论作为新一代信息处理理论基础,在很多领域得到了应用,但在数字水印方面的应用还有待发展。利用压缩感知理论,在图像稀疏化后的观测域中实现水印嵌入过程。用小波变换将原图像稀疏化,选择一个固定高斯随机矩阵对其进行观测,在观测压缩域中嵌入与提取水印,用匹配追踪算法恢复稀疏信号。实验结果表明,该算法满足透明性、鲁棒性和安全性要求;同时抗提取实验表明,该算法能有效抵御非法提取水印信息,增加了安全性。

Digital image watermark algorithm based on compressive sensing

The compressive sensing theory(CS),as a new generation of signal processing theory foundation,has been applied in varieties of areas.But CS theory has yet to be further developed in the digital watermark application.In this paper,a new algorithm of digital image watermark based on CS is proposed to embed watermark in the observation domain of the sparsed image.Gaussian stochastic matrix is selected to observe images sparsed by wavelet transform and realize watermark embedding and extracting in the observation domain.And then a Matching Pursuit algorithm is used to restore sparse signal.The experimental results indicate that the algorithm meets the needs of transparency,robustness and security.At the same time,the anti-extraction experiments show that the algorithm can resist illegal extraction of the watermark efficiently,and the security is guaranteed.