Research On Digital Watermarking Of Region Of Interest Based On H.264

With the rapid development of the Internet, multimedia information access has gradually been very advantage and shortcut. There is an increasing demand for multimedia information such as video, it is more and more widely used in the practical application. The digital video technologies developing very rapidly, the protection of its copyright should also attract a high degree of attention. As an effective means to protect the security and copyright of digital video, digital watermarking has been favored by many scholars in the world. Video is more complex and special than image on the expression of information, there are a lot of special requirements such as blind extraction, random detection and so on. So the technology of video watermarking is severely restricted. During the transmission of video information, there would indeed be a number of problems such as compression encoding, the video frame scrambling, the video frame loss etc. It is impossible to find a kind of algorithm to resist the above attacks. Therefore, this thesis makes a study on this issue.1. I studied video watermarking technology and carried on it with standard H.264 studying thoroughly. The previous studies are focused on the MPEG encoding standard, however,H.264 standard has more excellent performance. In this context, research on watermarking technology based on H.264 standard codec has very important practical significance. The Paper studies on the anti compression algorithm on the basis of H.264 standard. Video watermarking can be divided into three types: a video watermarking scheme of raw, embedding the watermark during the video compression, embedding the watermark after the video compression. Through research and analysis, this paper studies the first type, that is embedded in the original video watermarking.2.In the paper, a spread spectrum watermarking scheme based on anti H.264 compressed video sequence and a video watermarking algorithm based on region of interest are proposed. A watermarking algorithm based on spread spectrum is through watermarking, and add the index information to each watermark segment, finally embedded in the DCT domain coefficients of the video. Spread spectrum watermarking scheme is more used in image watermarking, and it is equally applicable to video watermarking technology. In the paper, firstly a watermarking should be dealt with, then combined with the watermarking algorithm in frequency domain, in order to lay foundation for the next step. And it also can improve the robustness of watermarking. The feasibility of MATLAB test has been verified through simulation and experiment.3.Subsequently, a video watermarking algorithm based on the combination of spread spectrum and region of interest is introduced in the paper. The watermark should be embedded in the area of interest, because the area is the most important information in the entire video, that’s to improve robustness of watermarking. At the same time, the watermarking scheme has better invisibility. The paper introduced how to extract regions of interest in video frames as a watermark embedding area by interaction, the advantages of this method is more intuitive and stochastic. Then we embedded watermarking in the DCT domain coefficients of the video. In order to make the video sequence more orderly, combining this method with the scheme of spread spectrum watermarking, it can also improve the efficiency and robustness of the watermark extraction. Experiments show that the proposed method has advantages in frame deletion, frame reorganization and H.264 compression.