摘要
针对以往视频加密研究成果不能直接适用于H.264的现状和保持语义格式兼容的难点问题,对H.264编码码流中量化步长、预测模式字、运动矢量和残差系数相关的重要信息进行了深入分析,提出了保持语义格式兼容的自适应的选择性加密方法;同时提出了一种将残差块按照其nC值和非零系数个数进行分组,然后在组内产生置乱码表,最后统一置乱的加密方法.实验结果表明:该加密算法在保证视频内容信息安全性的前提下,能保持语义格式兼容和码流大小不变,满足视频的实时传输.
After key information in H. 264 compressed stream, such as quant step, predictive mode, motion vectors and residual coefficients level were analyzed, corresponding format-compliant selective encryption schemes are proposed respectively. A format-compliant residual block spatial permutation scheme is also proposed. Residual blocks were categorized into different groups according to their nC values and numbers of total none-zero coefficients. Random shuffling table was generated in each residual block group respectively and residual blocks were permutated. The algorithm was composed of two parts： selective encryption and residual block spatial permutation. The experimental results show that H. 264 compressed video stream could be encrypted in real-time while the format is kept compliant with the same stream size.
出处
《华中科技大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2007年第8期13-17,共5页
Journal of Huazhong University of Science and Technology(Natural Science Edition)
基金
国家重大基础研究发展计划资助项目(2006CB303104)
湖北省科技攻关计划资助项目(2004AA101C18)
武汉市重点科技攻关计划资助项目(20031003021)
