基于自适应哈希算法的直播视频篡改检测

深圳大学计算机与软件学院,广东深圳 518060

计算机感知; 自适应哈希算法; 视频篡改; 视频内容保护; 实时视频; H.264编码器

Tamper detection based on adaptive hashing algorithm for live video streaming
Wen Zhenkun, Chen Bin, and Wu Huisi

Wen Zhenkun, Chen Bin, and Wu HuisiCollege of Computer Science and Software Engineering, Shenzhen University, Shenzhen 518060,Guangdong Province, P.R.China

computer perception; adaptive hashing algorithm; video forgery; video content protection; real-time video; H.264 encoder

DOI: 10.3724/SP.J.1249.2017.02165

备注

提出基于视频压缩帧的自适应哈希算法,并将该算法用于实时视频篡改检测.根据帧内数据长度划分非固定步长区间,提取帧的空间结构特征,使用哈希量化形成固定长度的密钥.实验结果表明,该算法能达到实时性要求,表现出较好的鲁棒性和区分性,通过统计视频画面组的密钥相似度,能有效检测视频篡改.

A new adaptive hashing algorithm based on video compression frames for real-time detection of the tampered video streaming is proposed. According to data length of a frame, the algorithm divides the frame into a number of intervals with different step-size and extracts the frame feature in spatial structure. The feature is quantized to generate the secret key. The experimental results show that the algorithm can meet real-time requirement and is robust and discriminative. It can detect accurately a video tampered through counting the similarity of keys in each group of pictures in the video.

·