Research On Digital Watermarking Technology Of Scalable Streaming Media

In recent years, with rapid development of multimedia technology and the Internet, people can easily spread, copy, store and process images, audio, video and text or other multimedia information, but how to solve the copyright issue has been the key problem of multimedia applications. Digital watermarking technology is to embed the secret information into digital content and give greater protection of digital copyright, which has become a research focus of information security. This thesis focus on video watermarking algorithm and theory, with the latest scalable coding H.264/SVC standards and technologies. Main work is as follows:First, motion vector-based video coding standard for H.264/SVC watermarking algorithm is proposed. Consider the principles of human visual model and SVC scalability, select motion vector of the time domain Layer0,alsoP-frame which magnitude is larger to embedded watermark. Compare the absolute value of horizontal and vertical motion vector and decide which one to modify.It's easy to extract watermarking blind. Experiments show that the algorithm is unnoticeable, and has good robustness with scalability of SVC. Second,a video watermarking scheme based on DCT domain is proposed . After pre-process, watermark bits will be embedded into the DCT coefficients which are selected by keys and the principles of human visual modeltheory also. The watermarks are embedded while the SVC coding, and the detection is blind. Moreover, making use of SVC multi-layer structure can larger the capacity and strengthen robustness. The result shows that the scheme can reach good robustness and is unnoticeable, also can resist some noise and attacks.Finally, this paper summarizes the covered topics and results. Unresolved issues are pointed out, and the future research directions are provided.