Watermarking Method Research Based On Videos

With the development of the Internet technology and digital technology, multimedia information, such as image, audio, video and even text information, has become the main source of obtaining information for people. The individuals can download and transmit digital multimedia products via the Internet. It comes up with severe challenges for the copyright protection issue of digital products. Because of the progressive improvement with the science and technology, digital video products have been extended widely in the Internet which makes the secure transmission of digital video to become the hot issue of copyright protection. Watermarking technology is the most method of implementing effect protection for digital products.In this paper, taking video sequences as the main research object, first we introduce the theoretical frame of digital watermarking, categories, performance requirement, application fields and existing problems briefly. And then we proposed a digital watermarking based on video sequences which are proposed on the base of researching existing image human visual system model. Besides, it utilizes the abundant redundant information adequately and the characters of human visual system to determine the embedding area for watermarking information and then selects DC coefficient which is derived from coefficients after carrying out discrete cosine transform for video sequences to implement the embedding and the extracting of watermarking information. In this paper, it demonstrates the proposed method has good invisibility and robustness against common attacks such as Gaussian filter, Gaussian noise, JPEG compression, Salt and Pepper Noise, dropping frames and swapping frames by using different video testing sequences. This method achieves the function of copyright protection.Finally, we make a conclusion for the main work of the paper and present several ideas of development direction for future video watermarking.