A Video Watermarking Scheme Based On Motion Vectors

The rapid development of internet and multimedia has brought the increasing popularity of digital video, with problems such as data security, copyright protection and authenticity certification becoming hot topics. Video watermarking technology, putting the copyright information into digital video through a special method, is an effective way for copyright protection.This paper proposes a new video watermarking scheme based on motion vector, combining the video copyright protection requirements with features of MPEG-2 coding standards. Arnold Transform and spread spectrum scrambling are used in the pretreatment of watermark so that the generated watermarking sequence is capable of resisting normal video attacks. Select motion vectors with higher amplitude in B frame, and then embed the watermark by change the parity relation of the motion vectors. We adopt data repeating technology in the embedding algorithm. Watermark is embedded into each GOP(Group of Pictures) of the video to ensure accurate extraction of watermark when the video is attacked. Determine the value of each position of watermarking sequence by majority judgment according to the watermarks extracted from different GOP. The results indicate that the algorithm with low complexity has excellent characters on invisibility and robustness. It can not only be compatible with MPEG-2 but also satisfy the requirements of real-time and blind extraction.