Research Of Digital Video Watermarking Based On H.264 Standard

With the rapid development and widely application of digital technology, the new style of transmission such as internet network, telecom mobile network and the multimedia terminal such as mobile telephone, STB (Set Top Box), computer expand the transmission area of digital contents (digital text, image, audio and video et al). However, the digital products have the advantage of lossless copy and distribution, and it is ubiquitous that the digital contents which are protected by Intellectual Property Rights random diffused. If there is a lack of the copyright protection of digital contents, it will take serious problem.As a new and effective approach to solve the copyright protection problem, digital watermarking has become a very active research area within the field of information processing and security. The watermark can be all the copyright information such as number, word and icon, and embed in the data of multimedia to protect copyright, communicate secretly and inspect the truth of digital products. Among all kinds of digital watermarking techniques, researchers have paid more attention to video watermarking technique due to its high research value and potential market demand. But currently video watermarking techniques lag far behind image watermarking techniques for the limitations of existing video coding standards and some special attacks against video watermarking techniques. At the present time, video watermarking algorithm are mostly based on MPEG and carried out in DCT (Discrete Cosine Transform) domain. H.264 standard is burgeoning image compression standard and its compressive efficiency is the highest of the current image standards. Its bit stream's structure appears strong adaptability to the network. Watermarking 's research on this standards has important effects on copyright protection in the network video transmit. Consequently, paper do the main research on video watermarking algorithm combining with H.264 standard ' s many excellent capabilities. The emphatic research work and contribution of this thesis are as follows:To begin with, this thesis systematically summarizes the characteristics and the application of video watermarking, and focuses on the research status of video watermarking algorithm and principles. Several traditional video watermarking algorithms are studied, and the analysis and comparison of the performances are given in detail.Then, introduce systemically about the basic kownledge of H.264. It do the important effects on digital television broadcast, video real-time communication, network video stream media transfer and multimedia short message. The encoder and decoder's characters and main function structure of H.264. Its common technologys and special technologys which are different from the common standards. H.264's speciality show that it is predominant in network transmission and video capture areas, and do the research about video watermarking technology on H.264 have a wide market application value.Furthermore, based on the strong correlation of block's coefficients, adjacent blocks and adjacent frames and video serial's movement characters, propose a video watermarking algorithm based on the combination of the block's DCT coefficients and motion vector. This algorithm embeds watermarking into I frame's intermediate frequency coefficients which are quantity and P frame's motion vector according to some regulations. This can embed watermarking adaptively by human vision characters to guarantee the synchronization between embedding and blind detecting, and improve the robustness of watermarking. Experimental results show that this algorithm has obvious low complexity, high watermark payload and the strong robustness against compression, re- compression, antinoise and mean frames.Finally, according to the stability of AC coefficients, this thesis proposed another system about face protection based on the digital watermarking. In the algorithm, fix on face correctly in the video capture system by face detecting technology, and make face area detected as color watermarking embeded into the original image, then encrypt facial area. At last, use image capture system based on DM642 to test the practicability and validity of this algorithm. Paper also do many attack experiments which show the algorithm have favorable security and robustness, and have wide application foreground.