| With the rapid developing of Internet technology, the delivery of digital video contents, comparing to traditional video materials broadcasted through televisions and cinemas, has become more simple and convenient. Meanwhile, the popularity of video editing software has made the editing and sharing of videos easier. As a result, the copyright protection of video contents has become more challenging. Video watermarking, which is a very effective method of copyright protection by means of embedding copyright information in the digital video, has become a hot research field of information security.Starting from the feature of digital watermarking, we explain the three categories of video watermarking according to the position of algorithm in the video codec standard and conclude the classic algorithms of the three types in this paper. A review of the H.264video codec standard is presented in the next chapter, including the encoder, decoder and key technologies of the standard. After that, we propose a robust video watermarking algorithm in codec combining the H.264video codec standard. We apply the Spread Spectrum and error correct code in the original watermark in pre-processing stage to enhance the robustness of our algorithm, and synchronous head information is added to the watermark to resist frame attacks. We select the sum of DCT coefficients of residuals, middle coefficients of H.264encoder, and the DCT coefficients of predicted block of current block, as the embedding carrier. In the embedding process, we employ a HVS of4*4DCT to control the intensity of embedding to achieve better perceptual quality. The watermark would be embedded in the middle area of DCT coefficients. Finally, we implement our algorithm based on the standard H.264codec JM8.6software. The experiment results show that the algorithm hardly affected the perceptual quality, and robust to video attacks like resolution attack, bitrate attack, frame attack and trans-code. A series of tests were applied in high definition videos, it performs good as well. The algorithm is blind watermark method, and combined with the H.264video codec standard closely, which makes the algorithm with low complexity by using the middle section of the encoder and decoder. |
