Researchers Of Static Image Watermarking Resist To Geometric Attacks

Digital watermarking technology, which is a new aspect of information security domain, can protect data copyright, authenticate data origin and integrality in open internet environment by embedding other information into host data. Geometric attack which is an effective way with its simpleness and strong offensive, is considered as the bottleneck of using robust static image watermarking technology in commercial applications. Study of image watermarking resist to geometric attacks is signification.Scholars designed a lot of methods to resist to geometric attacks, including of constructing geometrically invariant field, embedding template, using image geometric moments and image feature, etc. Such ways were experimented and analysis in this paper, we explain that the algorithm using image feature is a better choice to resist to geometric attacks base on comparing these methods above, and then choice Harris detector and scale invariant feature transform (SIFT) to resist to geometric attacks. We also modify the common expressions using image moments to compute rotation parameter, improve the estimate precision and reduce the difficulty of numerical analysis.Bas algorithm can achieve synchronization orientation between watermarking embedding and extracting by using Harris detector. In this paper, we consider the effect of geometric attacks especially scale attacks on Harris detector in real implement, add the arithmetic operators of feature choice, delete such feature points which are not stable in geometric transform, improve the performance of Bas algorithm. A new algorithm of computing the image affine transform parameters base on Harris detector and Delaunay grid algorithm is proposed, which can reduce storage of assistant information.Scale invariant feature transform is a new way to extract image feature proposed in 2004. We achieve a watermarking algorithm in wavelet domain base on image reverting by using SIFT, and get better results. At the end, zero-watermarking using SIFT is discussed, the new proposed zero-watermarking method can resist to mostly geometric attacks.