Research On The Theory And Application Of Geometric Attacks In Digital Watermarking

The rapid development of computer and multimedia network has changed ourlives significantly. Real-time transmission of the files can be easily realized with thehelp of Internet. Meantime, the prevalence of online downloading softwares, such asBitcomet and eMule, has further accelerated the share and distribution of themultimedia contents. Without canonical monitoring, the multimedia products, suchas image, audio and video, have been distributed and downloaded randomly, whichhas greatly infringed the legal rights of the authors. The digital watermarkingtechnique appears accordingly and has been regarded as a promising way to protectthe copyright of digital products. However, most of the existing watermarkingschemes can only resist traditional signal processing attacks, such as added noise,image compression and filtering. When the watermarked image is subject togeometric transformations, such as rotation, scaling and translation (RST), it is hardto detect the watermark successfully. As a result, geometric attack is regarded as thebottleneck of the practical application of digital watermarking. It is of greattheoretical and practical importance to design digital watermarking schemes with theability to resist geometric attacks.In this dissertation, we mainly do researches on image watermarking schemeswith the ability to resist geometric attacks. The related techniques are first studiedincluding the invariant moment theory, image normalization and feature pointdetection. Then watermark synchronization schemes are presented as well as severalwatermarking algorithms. The main research work and contributions are as follows.1. In-depth researches are conducted on geometric attacks, including theclassification, their effects on the watermarking systems and the esistingcountermeasures. First, geometric attacks can be grouped into global attacks andlocal attacks according to their different principles. Second, the different effectsof geometric attacks on spatial domain watermarking shcmes and transformdomain schemes are analyzed respectively, with image rotation and scaling as anexample. Finally, we classify the current countermeasures into three kinds,namely inverse transformation based methods, invariant domain embedding andfeature based synchronization. The author also analyzes these methods and pointout the advantages and disadvantages respectively. 2. An image adaptive zero watermarking scheme with the ability to resistgeometric attacks is presented based on the pseudo Zernike moment. Theproposed method employs the rotation invariant property of the magnitudes ofpseudo Zernike moments, and generates the original watermark sequence byapplying odd-even quantization on the pseudo Zernike magnitudes that arecomputed over the original image. During watermark detection, the sameparameters are used to apply odd-even quantization on the pseudo Zernikemagnitudes of the distorted image, producing the extracted watermark sequence.Then the similarity between the original watermark and the extracted one iscomputed and used to judge the presence of the watermark. This scheme is azero watermarking scheme so that it causes no degradation on image quality.Besides, it can resist both geometric attacks and traditional attacks, such asrotation, scaling, JPEG compression, added noise and filtering.3. Two watermark synchronization schemes are proposed based on the Harriscorner detector. The key idea is to determine the regions for both watermarkembedding and extraction by combining Harris feature points and imagenormalization. Based on the proposed synchronization schemes, two robustimage watermarking schemes are presented, which embed the watermark indiscrete wavelet transform (DWT) domain and spatial domain respectively.Besides, the original image is not needed in watermark detection. Experimentalresults show the effectiveness of the synchronization schemes. Besides, theproposed schemes can obtain good invisibility and they can resist traditionalsignal processing attacks and geometric attacks, simultaneously.4. A new framework for image watermarking system is presented based on thescale space feature points. The principles that watermarking schemes shouldfollow under this framework is analyzed in detail, for example how to choose thefeature points, how to determine the local regions, selection of embeddingdomain, watermark embedding and extraction strategy design, etc. Then tworobust image watermarking schemes are proposed, which can resist bothtraditional signal processing attacks and geometric attacks. The first scheme firstpartitions the local region in a rotation invariant pattern and embeds a binarywatermark sequence in the Nonsubsampled Contourlet Transform (NSCT)domain. The second scheme embeds a binary watermark image into the localregions adaptively in DWT domain. Extensive simulation results show that thesetwo schemes can resist both signal processing attacks and geometric attacks. Comparisons with the existing schemes also demonstrate the superiorities of theproposed scheme.5. A new watermark embedding and detection strategy is proposed, whichcombines image feature points and invariant moments. The basic idea is to usefeature point based watermark synchronization, and employs the pseudo Zernikemoment vectors computed from the local image regions as the watermark signal.Then the watermark is embedded by modifying the moment with predeterminedorder. Watermark detection can be achieved using the proposed minimumdistance decoder. As the watermark is embedded into local image regions, moresatisfactory image quality can be obtained compared to the existing pseudoZernike moment based watermarking methods. Besides, the robustness togeometric attacks is improved significantly. Simulation results using Stirmark4.0 have demonstrated the effectiveness of the proposed schemes.6. A new robust video watermarking scheme is presented using the space-timeinterest points. Space-time interest points represent interest events or inherentstructures of the video so that they cannot be easily destroyed when the video isdistorted. Based on this observation, we propose to use the space-time interestpoint for watermark synchronization. Namely the embedding and extractionframe index and positions are determined by these features. The proposedscheme can solve the probem of time-dimensional watermark synchronization invideo watermarking, and it can not only resist traditional image processingattacks but also video oriented attacks, such as frame removal, frame averagingand frame swapping. Simulation results demonstrate the effectiveness of theproposed scheme.