基于奇异值修改的视频隐写算法  

A New Video Steganographic Algorithm Based on Modifying Singular Value Decomposition

作　　者：段俊逸 张英男[2,3] 郝伟[3] 

机构地区：[1]武警工程大学装备工程学院 [2]武警工程大学电子技术系 [3]武警工程大学网络与信息安全武警部队重点实验室,西安710086

出　　处：《科学技术与工程》2016年第21期289-294,共6页Science Technology and Engineering

基　　金：国家自然科学基金(61379152)、(61403417)资助

摘　　要：针对现有大部分隐写算法基于理想模型作为研究基础,而未考虑现实复杂网络环境对隐写载体传递产生影响的问题,提出一种基于奇异值修改的视频隐写算法。先对载体帧进行分块,然后取出每块中的最大奇异值,并根据秘密信息从而滑动最大奇异值的后几位。实验结果表明,算法具有较好的视觉不可见性,以及较大的嵌入容量,并对视频载体的比特率影响较小,在模拟网络环境中噪声、滤波攻击后位错率值也较小。算法具有较高的实用价值。Focus on most of the existing steganography algorithms research on the ideal model, and without considering the impact on cover video by the real complex network environment, a new video steganography algorithm based on singular value decomposition is proposed. Firstly, it divides the current frame, then take out the maximum singular value of each block, and then embed the information according to the rules. The experimental results indicate that the algorithm has little impact on visibility, and has a larger embedding capacity, and impact little on the bit rate of the cover video, and when the noise and filtering attacks of real network environment were simulated, the SIM values is larger. In summary, this algorithm has high practical value in steganography.

关 键 词：视频 信息隐藏 奇异值分解 鲁棒性 

分 类 号：TP391.11[自动化与计算机技术—计算机应用技术]