FVEA-H：一种用于H．264的快速视频加密算法  

FVEA-H: A Fast Video Encryption Algorithm for H.264

作　　者：蒋建国[1] 包先雨[1] 李化雷[1] 宣曼[1] 陈晓蕊[1] 

机构地区：[1]合肥工业大学计算机与信息学院

出　　处：《系统仿真学报》2008年第16期4363-4367,共5页Journal of System Simulation

基　　金：国家自然科学基金(60474035);国家教育部博士点基金(20060359004);安徽省自然科学基金(070412035)。

摘　　要：视频加密技术被广泛地应用于多媒体系统中。由于要求加密速度快,以往的H.264算法通常选择编码效率高的关键数据进行加密,其安全性、压缩比和加密复杂度互为矛盾,难以满足商业应用的需求。为此,提出了一种新的H.264快速视频加密算法(FVEA-H),该算法(1)在基于上下文的自适应变长编码(CAVLC)中加密残差数据中的拖尾系数;(2)在指数哥伦布(Exp-Golomb)编码中扰乱帧内预测模式、帧间分块模式以及运动矢量差(MVD)符号位等。实验结果表明,该算法在不影响压缩比的条件下,可以取得比以往算法更好的加密安全性,而执行时间上却很快,为H.264的实时安全应用打下了基础。Video encryption technology has been widely used in multimedia systems. Previous H.264 algorithms usually encrypt the sensitive data which is important to improve coding efficiency, in order to meet the rapid encryption speed needs. However, security, compression ratio, and encryption complexity are mutual contradiction, thus these algorithms cannot be used in commerce applications. Towards this direction, a new fast video encryption algorithm for H.264 （FVEA-H） was proposed, where the Trailing Ones coefficients of residual data were encrypted based on Context-based Adaptive Variable Length Coding （CAVLC）, and intra prediction modes, inter block-division modes and motion vector difference （MVD） signs were encrypted based on Exp-Golomb Coding. Experimental results show that this algorithm can exhibit higher security than existing ones without having any adverse impact on compression ratio, as well as fast execution efficiency. These properties lay a sound foundation for H.264 secure applications with real-time requirement.

关 键 词：多媒体安全 视频加密 视频加扰 混沌密钥管理 H.264 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]