The Research Of Digital Watermarking Algorithm Based On Non-sub Sampled Contourlet Transform

With the rapid development of digital multimedia and Internet technology, information digitization and dissemination network became a major feature of the current information age, Real-time and accuracy of information dissemination have been significantly improved, but it also brings some reality problems related to information security and copyright protection. As an important means of addressing the digital copyright protection, digital watermarking discipline emerges as the time requires, this emerging interdisciplinary subject has become one of the research hotpots quickly, and has been widely used.Firstly, this paper introduced the background, classification, features, applications, and research status and other related content of the digital watermarking, classified the robust digital watermarking algorithm and the digital watermarking algorithm against geometric attacks, analyzed defects and difficulties of the existing watermarking algorithm, and gave the standard of performance evaluation of watermarking system. Secondly, the performance of Contourlet transform domain and non-sampling Contourlet transform domain in the field of digital watermarking were compared through a classical watermarking algorithm, and the result of non-sampling Contourlet domain is more suitable for digital watermark embedding and extraction. Finally, many digital watermarking algorithms of non-sampling Contourlet domain were studied in detail. In order to improve the robustness that resist to many types of common attacks and geometric attacks and the embedding capacity of image watermarking algorithm, the idea of non-sampling Contourlet transform, singular value decomposition, scale invariant feature transform and multiple embedded were introduced into digital watermarking technology, this paper proposed two new watermarking algorithms, the first algorithm embedded a certain amount of watermarking into the maximum singular values of blocks in low and high frequency sub-bands of the NSCT domain, this idea compromised three mutual restraint requirements of digital watermark invisibility, robustness and embedding capacity, it improved comprehensive performance of digital watermarking algorithm. The second algorithm highly kept robustness, but also significantly improved the embedding capacity and achieved the blind watermark extraction by improved the selection strategy of embedding. Experimental results showed that two algorithms had good robustness, they can effectively resist JPEG compression, noise, filtering and various geometric attacks.