JPEG图像被动取证研究进展综述  

Research Progress of JPEG Image Passive Forensics

作　　者：牛亚坤 赵耀[1,2] 李晓龙 NIU Yakun;ZHAO Yao;LI Xiaolong(The Institute of Information Science,Beijing Jiaotong University,Beijing 100044,China;The Beijing Key Laboratory of Advanced Information Science and Network Technology,Beijing 100044,China)

机构地区：[1]北京交通大学信息科学研究所,北京100044 [2]现代信息科学与网络技术北京市重点实验室,北京100044

出　　处：《信号处理》2022年第6期1170-1179,共10页Journal of Signal Processing

基　　金：国家自然科学基金(U1936212,62120106009);中央高校基本科研业务费专项资金资助(2018JBZ001)。

摘　　要：数字图像被动取证是利用图像的固有属性对其真实性认证、拼接区域定位及篡改历史分析的技术。由于大多数图像都需要进行JPEG压缩以便存储和传输,基于JPEG格式的数字图像被动取证越来越受到关注。为了更加全面地对现有取证方法进行梳理与归纳,本文着重对JPEG图像相关的取证技术进行介绍,详细论述了双重JPEG压缩检测、量化矩阵估计和拼接检测与定位方法中的关键技术,分析现有方法存在的问题与面临的困难,并对未来发展方向进行展望。Passive forensics of digital images is a technique that identifies the authenticity,localize the spliced regions and analyzing the history of tampering by using the inherent properties of images.As most images are JPEG compressed for storage and transmission,passive forensics of digital images based on JPEG format has attracted more and more attention.In order to more comprehensively summarize existing methods,this paper emphatically introduces JPEG image related forensics approaches,discusses the key technologies of double JPEG compression detection,quantitative matrix estimation and splicing detection and location in detail,analyzes the main issues and faced difficulties of existing methods,prospects for the future development directions.

关 键 词：JPEG图像被动取证 双重JPEG压缩 量化矩阵 拼接检测与定位 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]