图像USM锐化取证与反取证技术综述  

Overview of Image USM Sharpening Forensics and Anti-forensics Techniques

作　　者：赵洁 宋爽 武斌 ZHAO Jie;SONG Shuang;WU Bin(School of Computer and Information Engineering,Tianjin Chengjian University,Tianjin 300384,China)

机构地区：[1]天津城建大学计算机与信息工程学院,天津300384

出　　处：《广西师范大学学报（自然科学版）》2024年第3期1-16,共16页Journal of Guangxi Normal University:Natural Science Edition

基　　金：天津市科技计划项目重大科技专项(14ZCZDGX00868);天津市重点研发计划科技支撑重点项目(19YFZCGX00130);天津市企业科技特派员项目(19JCTPJC47200)。

摘　　要：反锐化掩膜(unsharp masking,USM)锐化是图像编辑中改善图像视觉清晰度的一种常用操作。然而,若将这一操作应用到篡改司法证据或者公共事件图片等场景,则会造成严重危害。近年来,图像取证领域的研究者已经提出许多USM锐化取证与反取证的方法,本文对该领域文献进行系统性综述。首先阐述USM锐化操作的基本原理,然后从算法实现原理角度对现有的USM锐化取证与反取证方法进行分类总结,接着介绍相关文献常用的数据集,并对现有方法的性能进行对比评价,最后分析现有方法面临的挑战并对未来的研究方向进行展望,主要包括以下几个方面:使用新的数据集,深入取证与反取证博弈的研究,深入研究锐化反取证,注重取证方案的执行效率,提高反取证的可靠性。Unsharp masking(USM)sharpening is a common operation to improve image visual clarity in image edition,However,it can cause serious harm if applied to scenarios involving the tampering of forensic evidence or public event photographs.In recent years,many methods for USM sharpening and anti-sharpening in digital image forensics have been proposed by researchers.So,a systematic review of the literature in this field is provided by this paper.The fundamental principles of the USM sharpening operation are first explained in this paper,followed by the categorization and summary of existing USM sharpening and anti-sharpening methods based on algorithm implementation principles.Additionally,commonly used datasets in related literature are introduced,and performance evaluations of existing methods are conducted for comparative purposes.Finally,the challenges faced by existing methods are analyzed and future research directions are discussed,mainly including the use of updated datasets,in-depth research on the game of forensics and anti-forensics,in-depth research on sharpening anti-forensics,the issue on the efficiency of forensics scheme execution,and the improvement of the reliability of anti-forensics.

关 键 词：反锐化掩膜 USM锐化 锐化取证 锐化反取证 数字图像取证 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]