The Research On Robust Image Watermarking Scheme Resisting Geometrical Attacks

With the rapid development of multimedia and Internet techniques, the multimedia security becomes an important and challenging research topic. Digital watermarking technique is an effective method for digital content protection such as copyright protection and becomes an important research area in the field of information hiding. Thus the research on digital watermarking is of great importance. For image watermarking, it is especially important to explore robust watermarking schemes that can resist geometrical attacks; in addition, such kinds of schemes are also very difficult to design, because geometrical distortions desynchronize the location and cause much difficulty in watermark detecting or extraction.In this thesis, a systematical introduction of the applications and background of watermarking techniques is given first. Then, watermarking techniques on resisting geometrical attacks are reviewed. We present the consequences of geometric attacks and watermark synchronization, and we also describe several exsisting image watermarking schemes that can resist geometrical attacks.Secondly,we present the watermarking techniques on resisting geometrical attacks. Two novel schemes resisting geometrical attacks are proposed: one is to embed the watermark into geometric-invariant region and the other is to recover the image first before detecting the watermark. Our work is focused on the key techniques of digital watermarking for images, which include blind detection, robustness, and invisibility. The research conducted in this thesis is the focus of current watermarking techniques.The feature-based image watermarking technique is studied. A feature point-based watermarking scheme is proposed based on the analysis of the recent feature-based algorithms. The primary characteristic of the scheme is that it exploits the refined pattern recognition techniques to find the robust image feature. We use a local invariant feature of the image called the scale invariant feature transform. The extracted points are selected to obtain the robust feature points, which are used as a reference for both watermark embedding and detection. Watermark is embedded into each local region centered at the points repeatedly. Experimental results show that our method is robust against geometric distortion attacks as well as signal-processing attacks.Besides, a novel image watermarking scheme based on geometric moments resisting geometrical attacks is proposed. We make use of geometric moments to recover the attacked image by estimating the parameter when detecting the watermarks, original image being not needed. A geometric synchronization problem between the watermark embedding and detection is overcome. Experimental results show its rationality and the effectiveness of the scheme.