A video watermarking scheme resistant to synchronization attacks based on statistics and shot segmentation

One of the challenges of blind watermark detection is synchronization. In this paper, a new video watermarking procedure for resistant synchronization is proposed. We partition watermark into several segments, and embed every segment into different scenes of video sequence. Firstly, motion vectors are grouped according to their magnitude and each group is further partitioned into two subsets. We use element number ratio in two subsets of the same group to denote watermark bit 0 or bit 1. Thus, watermark is associated with motion vectors' statistical characteristics, and those motion vectors carrying the identical watermark bits are uniformly distributed to the whole video sequence, they are spatio-temporal indiscerptibly. It is shown that this kind of watermark is more resilient against temporal synchronization attacks. Experimental results from an implementation of the algorithm are presented.