Video watermarking temporal synchronization on motion vector
Kezheng Lin, Wei Yang, Liu Pie · 2008
Temporal synchronization is one of the challenges for blind watermark detection. If the detector¿s input is watermarked but synchronization fails, then the embedded watermark will not be detected. In this paper, a novel video watermarking algorithm is proposed based on data statistics classification of motion vectors. Through analyzing data statistics of motion vectors¿ magnitude, we classify the motion vectors into several groups, and then according to a certain rule, we embedded one bit of watermark sequence into motion vectors in one group. 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. It avoids the sensitivity to watermark synchronization comparing to traditional algorithm. The experimental results demonstrate that this scheme is robust to common kinds of video attack, including frame dropping, frame inserting etc. Moreover, the scheme has the property of perceptual transparency and random detection.