Research Of Digital Video Watermarking Techniques Based On MPEG

The present computer technology makes people more and more easy to create and use digital media works, what's more, the development of Internet promotes the wide-spread of digital media works, one excellent works would be spread all over the world almost one night. However, the copyright of the digital media works is seriously damaged, because one can easily access, duplicate and spread the works at random, as a result the interests of the owner are heavily damaged.To effectively protect the copyright of works, many methods are produced, digital watermarking is one of those methods. As one way of information hiding, digital watermarking can embed secret information (watermark) into original data (carrier data) without having effects on the using of the carrier data, in addition, the watermark can not be removed, only special software can extract it. At present, there have many results in the research about digital watermark, there even have usable products of image digital watermark.In this paper, some techniques about digital watermarking and the standard of MPEG are presented, then the characters of video digital watermark and the current technical methods are discussed. At last, a scheme of digital watermark embedding based on spread spectrum about MPEG-2 is produced. By this way, embedding and extracting watermark in compressed video domain are carried out. When embedding, the localizations of the watermarks can be adjusted adaptively in accordance with the detail information of the video image. The watermark detection and extraction process does not require the original video. Because the video sequence has more data than the watermark in common, we can get a more invisible and robust watermark by decreasing the number of watermark embedding. The experimental results demonstrate that the watermark has the considerable perceptual invisibility and robustness, and the operation complexity is comparatively low in this scheme.