Research On Digital Watermarking Techniques Based On Video Sequences

With the rapid development of the Internet and digital multimedia technology, people canacquire multimedia material more extremely easy, such as the digital audios, images, texts,videos and other multimedia information. Especially the video include images, audio andcharacters has become a very active material of our daily life. While the convenienceaccompanied by many offenses and crimes, everyone can modify the video and other multimediamaterials by using AVS Video Editor and other tools, all of that made people can’t recognize thecopyright of digital multimedia products. Therefore, in order to prevent unauthorized copying oras an evidence of copyright infringement of digital multimedia products, video watermarkingtechnique, as one of the candidates for effective protection means, has been produced anddeveloped. The existing methods of video watermarking could generally be divided into threecategories. One is to embed watermark in the original video directly; the second is based on thecompressed video; the third method is to watermark the compressed video.In this paper, we explore the way to embed watermark into uncompressed video directly.Firstly, the paper make a summary of the development history and the necessity of the videowatermarking researches. Then we introduce and discuss some basic conceptions of watermarksuch as the basic methods of video watermark, the characteristic of video watermark, theclassification of video watermark, the fundamental of video watermark, the attacks of videowatermark, the evaluation of video watermark and so on, which provide a basis for the followingresearch of video watermark in the context.The Spread Spectrum video watermarking scheme as one of the most importantwatermarking schemes are introduced in detail. The Spread Spectrum watermarking scheme is asimple and effective classic algorithms of images watermark. We apply this method in videosand combined it with frequency domain and watermarking pre-processing method to enhance therobustness of the algorithm. This method needs precondition the watermark firstly, then the processed watermark was embedded into the DCT domain of the original video sequences..Afterdiscussing of classic algorithms, A new algorithm of video watermarking based on changed areais studied in this paper. We proposed a method to locate areas whose pixels change greatly fromthree successive frames of uncompressed videos by using three-frame differencing. Because thepart of change areas is the most important information of videos, the watermark which wasembedded on important components is beneficial to improve the robustness of the watermarkalgorithm. Meanwhile, our human eye is less sensitive to the "error" of movement parts ofvideo, so the great change areas is extracted and put in this method to highlight the advantagesof this method. Then we choose the region where changes drastically to embed watermark byusing the Quantization Modulation algorithm in DCT coefficients. The experimental resultsindicate that the proposed scheme has a good performance on frame dropping, frame cropping,frame rating, Gaussian noising, MPEG2compression, etc. And it has a advantage on littledegrading the video quality.In modern society, as the rapid development of digital video watermarking technology, allkinds of video watermarking methods will appeared in endlessly, which will be accompanied bythe problems of digital multimedia technology. So we need keep up the trend of science andtechnology to design more effective watermarking schemes.