Digital Watermarking Technology Of Anti-Compression Algorithm Based On Spatiotemporal Characteristics

As protection information,digital watermark is embedded in multimedia files by computer algorithm.Due to its security,concealment and robustness,digital watermarking has been the focus of research in recent years.The research of digital watermarking based on video compression algorithm is more and more diversified with the development of science and technology.But the applications of video digital watermarking mainly focus on copyright protection,content authentication,copy control,information confidentiality and so on.This paper is based on the actual problems of surveillance video,such as the illegal transmission of surveillance video,stolen video and so on.The author researches a digital watermarking technology to realize video traceability application.The technology is mainly used in sensitive surveillance video monitoring.Video transmission may be compressed by different compression algorithms and some malicious attacks.Therefore,it is necessary to design an anti-compression digital watermarking technology based on spatiotemporal characteristics.The author focuses on the video using H.264 compression coding algorithm.A CAVLC digital watermark modulation algorithm based on corner selection is proposed.Consider multi-frame continuous embedding digital watermarking,an anti-compression digital watermarking technology based on spatiotemporal features is realized.The main work includes the following aspects:(1)The author studies three common embedding points of H.264 digital watermarking algorithm.Entropy coding(CAVLC),image features(corner points)and motion vectors are selected as embedding points.Among the three embedded positions,CAVLC is the information of time dimension,which is used for space domain redundancy compression in H.264 coding standard.corner points and motion vectors are spatial dimension information,which represent the correlation between adjacent macroblocks in the H.264 coding standard and represent the time domain continuity of video frames.The research of embedding point focuses on,the damage degree of digital watermark after H.264 compression,the damage degree of digital watermark after H.264 recompression,the change degree of video quality with digital watermark.Analyze the advantages and disadvantages of various algorithms.It is found that CAVLC and corner point have better compression resistance and have their own advantages.(2)The author combined two embedded points of CAVLC and entropy coding.The author proposes a CAVLC digital watermark modulation algorithm based on corner selection.Firstly,the author theoretically analyzes the possible influence of embedding digital watermark on video by combining two kinds of embedding points and the rationality of embedding digital watermark.Secondly,an example test is carried out to study the influence of the compression algorithm on the digital watermark embedding algorithm.At the same time,The author proposes to protect digital watermark with error-correcting code,considering the watermark video may be attacked by noise.This paper analyzes the ability of digital watermarking based on error-correcting code protection to recover watermark after being attacked by noise.(3)Implement the algorithm and test it.The results show that the CAVLC digital watermarking modulation algorithm based on corner selection can resist H.264 recompression well.However,the performance of the algorithm against non-H.264 recompression coding is not good enough.After the digital watermark is protected by error-correcting code,the digital watermark has certain anti-noise ability.And the error-correcting performance decreases with the increase of noise.In the end,the paper analyzes the recovery of digital watermark in the possible video frame loss.