摘 要:隐写术的不断发展使得隐写分析术面临的挑战越来越大。近年来,有不少学者围绕深度学习网络强大的图像特征表达学习能力进行隐写分析术研究,从而避开人工设计分类特征,减少人的参与度,用数据和算法驱动机器去实现数字图像是否含密的判定。本文将从数字图像的全局和局部统计分布特性这两个方面出发,梳理传统的和基于深度学习的隐写与隐写分析术在空域和 JPEG 域上的研究进展,并对数字图像隐写和隐写分析术未来发展方向做简要讨论。
关键词:深度学习;隐写术;隐写分析术;统计分布特性;数字图像
DOI:10.19850/j.cnki.2096-4706.2021.13.017
基金项目:贵州省教育厅青年科技人才 成长项目(黔教合 KY 字[2017]226);教育部第二批国家级新工科研究与实践项目(E-D XKJC20200524);2020 年大学生创新创业训 练计划项目(202011731005)
中图分类号:TP391.4;TP183 文献标识码:A 文章编号:2096-4706(2021)13-0068-05
Study Overview of Digital Image Steganography and Steganalysis Based on Deep Learning
TAN Yanping, LUO Yong, ZHANG Jun
(Guizhou University of Commerce, Guiyang 550014, China)
Abstract: With the continuous development of steganography, steganalysis is facing more and more challenges. In recent years, many scholars have carried out steganalysis research around the powerful image feature expression learning ability of deep learning network, so as to avoid manually designing classification features, reduce people’s participation, and drive the machine with data and algorithm to determine whether the digital image contains secret. Starting from the global and local statistical distribution characteristics of digital images, this paper combs the research progress of traditional and deep learning based steganography and steganalysisg in spatial domain and JPEG domain, and briefly discusses the future development direction of digital image steganography and steganalysis.
Keywords: deep learning; steganography; steganalysis; statistical distribution characteristics; digital image
参考文献:
[1] 王朔中,张新鹏,张卫明 . 以数字图像为载体的隐写分析 研究进展 [J]. 计算机学报,2009,32(7):1247-1263.
[2] JOHNSON N F,JAJODIA S. Exploring Steganography: Seeing the Unseen [J].Computer,1998,31(2):26-34.
[3] PETITCOLAS AP F,ANDERSON J R,KUHN G M. Information hiding-a survey [J].Proceedings of the IEEE,1999,87(7): 1062-1078.
[4] 周翠红,秦姣华,左伟明,等 . 空域图像隐藏信息检测技 术研究 [J]. 湖南城市学院学报(自然科学版),2010,19(2): 谭艳萍 , 等 : 基于深度学习的数字图像隐写和隐写分析术研究概述 第 13 期 现代信息科技 72 2021.7 55-60.
[5] TRIVEDI S,CHANDRAMOULI R. Secret key estimation in sequential steganography [J]. IEEE Transactions on Signal Processing, 2005,53(2):746-757.
[6] TRIVEDI S,CHANDRAMOULI R. Active steganalysis of sequential steganography [C]//Conference on Security and Watermarking of Multimedia Contents V Jan 21-24.Santa Clara CA:Proceedings of SPIE - The International Society for Optical,2003.
[7] 茹炳晟.JPEG图像隐写信息盲检测系统的设计与开发 [D]. 上海:上海交通大学,2008.
[8] 翟黎明,嘉炬,任魏翔,等 . 深度学习在图像隐写术与隐 写分析领域中的研究进展 [J]. 信息安全学报,2018,3(6):2-12.
[9] YANG C H,WENG C Y,WANG S J,et al. Adaptive Data Hiding in Edge Areas ofImages With Spatial LSB Domain Systems [J]. IEEE Transactions on Information Forensics and Security,2008,3(3): 488-497.
[10] HUANG W,ZHAO Y,NI R R. Block-Based Adaptive Image Steganography Using LSB Matching Revisited [J]. Journal of Electronic Science and Technology,2011(4):291-296.
[11] LIU G J,LIU W W,DAI Y W,et al. Adaptive steganography based on block complexity and matrix embedding [J]. Multimedia Systems,2014,20(2):227-238.
[12] HUANG F J,HUANG J W,SHI Y Q. New channel selection rule for JPEG steganography [J].IEEE Transactions on Information Forensics and Security,2012,7(4):1181-1191.
[13] WANG C,NI J Q. An efficient JPEG steganographic scheme based on the block entropy of DCT coefficients[C]//2012 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).Kyoto:IEEE,2012.
[14] ZHANG Y W,ZHANG W M,CHEN K J,et al. Adversarial Examples Against Deep Neural Network based Steganalysis[C]//Proceedings of the 6th ACM Workshop on Information Hiding and Multimedia Security. ACM,2018:67-72.
[15] TANG W X,LI B,TAN S Q,et al. CNN Based Adversarial Embedding with Minimum Alteration for Image Steganography [J].IEEE Transactions on Information Forensics and Security,2018.
[16] MA S,GUAN Q X,ZHAO X F,et al. Weakening the Detecting Capability of CNN-based Steganalysis [EB/OL].(2018-03-29). https://www.researchgate.net/publication/324104295_Weakening_the_ Detecting_Capability_of_CNN-based_Steganalysis/fulltext/5abdb18daca 27222c755994b/324104295_Weakening_the_Detecting_Capability_of_ CNN-based_Steganalysis.pdf?origin=publication_detail.
[17] Yang J,Liu K,Kang X,et al. Spatial image steganography based on generative adversarial network [EB/OL].(2018-04-21). https://arxiv.org/abs/1804.07939v1.
[18] 付章杰,李恩露,程旭,等 . 基于深度学习的图像隐写 研究进展 [J]. 计算机研究与发展,2021,58(3):548-568.
[19] 陈君夫,付章杰,张卫明,等 . 基于深度学习的图像隐 写分析综述 [J]. 软件学报,2021,32(2):551-578.
[20] 黄思远,张敏情,柯彦,等 . 基于显著性检测的图像隐 写分析方法 [J]. 计算机应用,2021,41(2):441-448.
[21] Tang W,Li H,Luo W,et al. Adaptive steganalysis against WOW embedding algorithm[C]//Proceedings of the 2nd ACM workshop on Information hiding and multimedia security.2014:91-96.
[22] 李德维,任魏翔,王丽娜,等 . 基于噪声感知残差网络 的 JPEG 隐写分析方法 [J]. 计算机应用研究,2021,38(10): 3148-3152+3165.
作者简介:谭艳萍(1990—),女,汉族,湖南邵阳人,讲师, 硕士研究生,研究方向:数字图像隐写与隐写分析;罗永(1991—), 男,汉族,贵州毕节人,讲师,硕士研究生,研究方向:物联网技 术应用;张俊(1987—),男,汉族,河南信阳人,副教授,博士, 研究方向:物联网技术应用。