Research On Digital Video Watermarking Algorithm Based On H.264

With the VCD, DVD, EVD, VOD and other digital video products are increasingly rich, internet technology has greatly accelerated the widespread dissemination of multimedia information. However, due to digital reproduction does not cause deterioration in the quality, unauthorized users can easily access, copy, distribute digital multimedia products, which greatly damaged the economic interests of copyright owners. Compressed domain digital video copyright protection technology is not yet mature, which makes copyright protection of digital products as an effective means of video digital watermarking technology research more urgent. Therefore, the digital video watermarking technology has become the focus and hot spots of multimedia information security research. On the basis of depth study of the key technologies based on H.264 coding standard and existing classic video watermarking algorithm, this paper presents two kinds of robust watermark algorithms based on H.264 for video copyright protection, and a large number of experiments have been carried out to verify the validity of the algorithms. The main research results of this paper are as follows:(1) A robust blind detection video watermark algorithm based on H.264 compression standard on the domain of DCT is proposed. This algorithm first scrambles the primitive watermark image by the improvement scrambling plan. Then in H.264 frame predictive coding stage, the watermark information embedded into the I-frame-by-bit for each macro block of 16 4×4 sub-blocks of the low-frequency AC coefficient of DCT transform, and find a few low-frequency coefficients of the median based on storing algorithm as a specific embedding position, according to the model two-by-bit watermark embedding rules of the algorithm, embedded the watermark into the median bit-by-bit. Finally, simulation of the watermark embedding and extraction in the H.264 reference model has been carried out, and using the Gauss noise, the low-pass filtering, the frame deletion, the heavy quantification code and so on several kind of video frequency methods of attack to analysis and test the performance of the algorithm, and compared with the related algorithm, the results show that the algorithm is simple and effective, and has higher robustness and un-perceptibility.(2) A design of weight difference video watermarking based on motion vector is proposed. This algorithm is fully taken into account human eye visual characteristic, select the threshold value of the larger macro blocks as a watermark carrier, through the judgments of each MV difference between horizontal and vertical components to the embedded watermark. The algorithm embeds watermark not only in the difference D>0 and D<0, but also when the difference D=0. By choosing to change the horizontal component of MV to achieve the embedding of watermark, on the one hand, it guaranty the embedded capacity, on the other hand, it avoid the horizontal component to be done in a big change, thereby, the impact of video quality is reduced. In this paper, experimental analysis of the visual distortion, before and after the embedded watermark PSNR, the extracted watermark and other aspects, have been carried on, and compared with the related algorithms. The results show that the algorithm is simple, fast and effective features to meet the requirements of real-time processing of video, and has better robustness and security.