基于H.264的漂移深度控制视频隐写算法  

A Drift-Depth Control Scheme for Video Steganography in H. 264 Stream

作　　者：尤星[1,2] 谢东峰 李宝清[2] 袁晓兵[2] 

机构地区：[1]中国科学院大学,北京100049 [2]上海微系统与信息技术研究所,上海200050

出　　处：《电子学报》2014年第7期1284-1290,共7页Acta Electronica Sinica

基　　金：国防重点实验室基金(No.9140C180101120C18076)

摘　　要：在H.264/AVC视频压缩域隐写技术中,失真漂移问题极大地影响了信息嵌入的容量和稳定性.本文结合工程实际和实验,发现并指出了漂移传播的实质性规律,提出了相应的改进策略.首先,提出了漂移分级的控制策略,依据不同视频质量及隐蔽性要求自适应地调整失真漂移的传播范围,改进了视频隐写的隐蔽性和安全性;进而通过对漂移影响分层分析,挖掘出了更多的隐写位置,增加了隐写容量及选择隐写方案的灵活性.形成了一种新的扩容视频隐写算法——漂移深度控制算法.实验结果和分析表明,该算法能有效地减小嵌入偏移,隐写容量、视频质量及安全性都有较大提升.Distortion drift is a crucial factor for the embedding capacity and stability of H. 264/AVC video in compression domain. In this work, considering practical application, we have exploited the substantive regularities that result in distortion drift. Accordingly, some improvements are made to adapt the influence of distortion drift. Firstly, a drift-depth strategy is proposed to im- prove the embedding invisibility and security, which can control the distortion range according to requirements of video quality. Sec- ondly, through the analysis of the influence of different drift spreading range, more appropriate steganography positions are discov- ered to increase the embedding capacity and flexibility. Finally, a drift-depth control scheme is presented. Both analytical and simula- tion results show that our proposed scheme can improve embedding capacity and security, and meanwhile minimize distortion drift.

关 键 词：视频隐写 预测漂移 漂移深度控制 失真补偿 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]