Based On Mpeg-2 Video Watermarking Technology

As a new and effective approach to solve the problem of copyright protection, digital watermarking has emerged to be an active and flourishing research topic in the field of information security recently. Among the plethora of digital watermarking techniques, researchers have paid more attention to video watermarking technique because of its high research value and potential commercial profits. However, currently available video watermarking technique is far beyond mature yet compared with the still image watermarking. This thesis focuses on video watermarking algorithms based on MPEG- 1/2 and experiments are performed to test the effectiveness of the proposed algorithms. The main research work and contributions of this thesis are summarized as follows:(1) This thesis systematically summarizes the concepts, the technique requirements, the typical applications and the research status of video watermarking techniques and classifies attacks and benchmarks of the video watermarking techniques as well as focuses the research on all kinds of proposed video watermarking algorithms.(2) Based on the HVS (Human Visual System) model, this thesis proposed a novel video watermarking algorithm using motion vector for MPEG-2 video stream, which by adjusting the horizontal or vertical component value of motion vector to embed the watermark information. Experimental results show that the embedded watermarks are imperceptible and robust against the attack of compression.(3) The author researches the watermarking systems for video authentication deeply and proposes a novel watermarking scheme for video authentication to protect the region of interest. Experimental results show that it can detect a modification place on the video object very well. The security of algorithm relies on the key and it can fulfill blind detection.