Research On Geometrical Invariant Video Watermarking Algorithms

With rapid development of the digital multimedia technology and network technology, and the popularization of the network, people can conveniently download, store and distribute multimedia information. It brings convenience to the entertainment of people but threat to copyright protection for multimedia works. Increasingly deterioration of problems about piracy and copyright attracts attention from both government and research institution in the world.Digital watermarking technology plays an important role in multimedia content protection and related fields. Currently, most open literatures on watermarking focus on static image watermarking technology, but nowadays this trend seems to be changing. With the promising development of high speed network and digital TV network, circulation of video media become more and more convenient and fast, making it urgent to protect video copyright. The huge data size, diversity of compress standard and distinctiveness of data structure of video bring in distinctiveness and complexity to the research about video watermarking, leading to much higher demands in performance than image watermarking.Firstly, embedding and detection of video watermarking should be of low complexity to satisfy the real-time requirements in application; secondly, higher performance is required on robustness, especially the robustness against geometric attacks, which is one of the most difficult problems in digital watermark field. Therefore, video watermarking should not only be robust to spatial geometric attacks but also satisfy real-time requirements. Besides, video watermark should take temporal geometric attacks into consideration, such as frame replacement, frame deletion, etc..To summary, research on Real-time video watermarking technology against geometric attacks is a challenge, which satisfys the requirements of practical application.To solve this problem, a real-time video watermarking scheme against geometric distortions is proposed for DCT (Discrete Cosine Transform)-encoded compressed video data (GIW-FDCT). The full DCT coefficients have proven to be invariant to scaling and local geometric attacks. Therefore the watermark is embedded into watermark minimal sequences (WMS) by modulating the low-frequency full DCT coefficients. To meet the requirement of real-time performance, a fast inter-transformation is employed to construct full DCT directly from block DCTs. To resist various video format conversions, the watermark detection is performed in spatial domain. This does not affect the real-time performance because compressed video can be easily decoded fast by employing some codec or decoding chips. Furthermore, the proposed scheme can resist rotation attacks by employing a rotation compensation strategy. Experimental results demonstrate that the proposed scheme is real-time, transparent and robust to signal distortions, geometric distortions.Next, a new blind video watermarking scheme against geometric distortions based on DT-CWT (Dual Tree-Complex Wavelet Transform) is proposed, denoted as VW-DTCWT. The scheme is most suitable for DCT-encoded compressed video. The 2-D Discrete Fourier transform (DFT) coefficients are constructed directly from the block DCTs by exploiting a fast inter-transformation. Then a RST (rotation, scaling, and translation) invariant domain is generated by performing the Log-Polar Mapping (LPM) and the DT-CWT on the 2-D DFT magnitudes.To resist the local cropping, we propose a novel robust video watermarking scheme based on local affine invariant features in the compressed domain (LVW-HA). In order to synchronize the watermark, we use local invariant feature points obtained through the Harris-Affine detector which is invariant to affine distortions. To decode the frames from DCT domain to the spatial domain as fast as possible, a fast inter-transformation between block DCTs and sub-block DCTs is employed and down-sampling frames in the spatial domain are obtained by replacing each sub-blocks DCT of 2x2 pixels with half of the corresponding DC coefficient. The above-mentioned strategy can significantly save computational cost in comparison with the conventional method which accomplishes the same task via inverse DCT (IDCT). Then we extract Harris-Affine feature points in the down-sampling frame. A RST invariant domain is generated as the embedding domain by performing the SPM (Square-Polar Mapping) and 1D-DFT on the local patches centered on feature points. Experimental results demonstrate that the proposed scheme is transparent and robust to signal-processing attacks, geometric distortions including rotation, scaling, aspect ratio changes, linear geometric transforms, cropping and combinations of several attacks, frame dropping, and frame rate conversion. Finally, a real-time video watermarking scheme against geometric distortions based on histogram in DWT Domain is proposed for DCT-encoded compressed video data (VW-HDWT). Watermark resistance to geometric distortions is achieved by using the invariance of the histogram shape of the low-frequency sub-band of DWT domain. The watermark bits are embedded into watermark minimal sequences (WMS) by modulating the ratio of each two neighboring bins in histogram shape of DWT domain. To meet the requirement of real-time performance, a fast inter-transformation is employed to construct DWT domain directly from block DCTs. The experimental results show that the proposed scheme meets real-time reqirement and are robust to many geometric distortions.