上海交通大学学报(自然版) ›› 2019, Vol. 53 ›› Issue (3): 334-340.doi: 10.16183/j.cnki.jsjtu.2019.03.011

• 学报(中文) • 上一篇    下一篇

基于双编码参数模型的同量化参数双压缩检测算法

沈婷,孙锬锋,蒋兴浩   

  1. 上海交通大学 电子信息与电气工程学院; 信息内容分析技术国家工程实验室,上海 200240
  • 出版日期:2019-03-28 发布日期:2019-03-28
  • 通讯作者: 孙锬锋,男,副研究员,电话(Tel.):021-34206657;E-mail:tfsun@sjtu.edu.cn.
  • 作者简介:沈婷(1992-),女,江苏省常熟市人,硕士生,主要研究内容为视频双压缩检测.
  • 基金资助:
    国家自然基金(61572320,61572321)资助项目

Detection of Double Compression with the Same Quantization Parameter Based on Dual Encoding Parameter Model

SHEN Ting,SUN Tanfeng,JIANG Xinghao   

  1. School of Electronic Information and Electrical Engineering; National Engineering Laboratory on Information Content Analysis Techniques, Shanghai Jiao Tong University, Shanghai 200240, China
  • Online:2019-03-28 Published:2019-03-28

摘要: 基于I帧宏块帧内预测模式(IPMBM)和量化离散余弦变换(DCT)系数相结合的双编码参数模型,本文提出了一种H.264相同量化参数双重压缩检测算法,利用单次压缩和双次压缩视频编码参数具有不同收敛性的特点来设计分类特征.首先将输入视频以相同量化参数压缩多次,然后从视频的I帧中提取相应IPMBM和量化DCT系数,统计两者在相邻两次压缩版本之间的变化数量,以此作为分类特征进行提取,最后采用支持向量机(SVM)来判别视频是否经历双重压缩.实验结果表明,该算法可正确有效地检测双重压缩.

关键词: 双压缩, 相同量化参数, 帧内预测模式, 离散余弦变换(DCT)系数

Abstract: A novel method was proposed to detect double H.264 compression with the same quantization parameter, based on the dual encoding parameter model composed of Intra Prediction MacroBlock Modes (IPMBM) and the quantized discrete cosine transform (DCT) coefficients of I frames. The method designed the classfication feature based on the property that coding parameters of single compressed videos and double compressed videos have different convergencies. Firstly, the input video was recompressed repeatedly with the same quantization parameter. Then, the IPMBMs and the quantized DCT coefficients were extracted from I frames. The numbers of altered IPMBMs and quantized DCT cofficients between adjacent recompressions were used to calculate the classification feature. Finally, the extracted feature was fed to the trained support vector machine to detect double H.264 compression. Experimental results show that the proposed method can detect double H.264 compression correctly and efficiently.

Key words: double compression, same quantization parameter, intra prediction modes, discrete cosine transform (DCT) coefficients

中图分类号: