Research On Digital Video Watermarking Technology Based On MPEG-4

The new generation of video codec standard MPEG-4provides coding methodwith higher efficiency and better video-compressed quality. Meanwhile, with thewidely use of Internet, the technology of digital watermarking used as copyrightprotection becomes more and more important. However, some special properties ofMPEG-4make many video watermarking methods used before invalidated, so it isnecessary to find new video watermarking method to adapt the new usage.Based on a great many papers and documents, by analyzing common features ofexisting video watermarking method and the properties of MPEG-4, especially forsome common technology and algorithm used in the MPEG-4standards, this papercome up with two video watermarking methods based on the coding process ofMPEG-4. Water information is embedded by modifying the DCT coefficients afterquantity in I-frame of video.The first adaptively watermarking method is based on the Watson perceptualmodel. This model is a usage of Human Visual System. The brightest and more detailsarea are chosen in8x8block to embed water information in this algorithm, so theposition chose can be adaptive for the video. Meanwhile, some results of the Watsonmodel are used to control the feature of this algorithm.Energy argument is used to select area need to be embedded water information inthe second watermarking method. Three area of a Y-Block in every micro block arechosen their energy arguments are calculated, according these arguments, a area ischose to embed water information, and then the DCT coefficients in this area aremodified according special rules.Finally, the performances of these algorithms proposed in this paper are analyzed.The experiment results show that these algorithms are robust and imperceptible tosome extent.