基于隐写噪声深度提取的JPEG图像隐写分析

doi:10.19665/j.issn1001-2400.2023.04.016

西安电子科技大学学报 ›› 2023, Vol. 50 ›› Issue (4): 157-169.doi: 10.19665/j.issn1001-2400.2023.04.016

基于隐写噪声深度提取的JPEG图像隐写分析

范文同^1,^2,³(),李震宇^1,^2,³(),张涛⁴(),罗向阳^1,^2,³()

1.中国人民解放军战略支援部队信息工程大学网络空间安全学院,河南郑州 450001
2.数学工程与先进计算国家重点实验室,河南郑州 450001
3.河南省网络空间态势感知重点实验室,河南郑州 450001
4.常熟理工学院计算机科学与工程学院,江苏常熟 215500

收稿日期:2023-01-15 出版日期:2023-08-20 发布日期:2023-10-17
通讯作者: 李震宇
作者简介:范文同(1998—),男,中国人民解放军战略支援部队信息工程大学硕士研究生,E-mail:fftxxgc@sina.com;|张涛(1977—),男,教授,E-mail:tzhang@cslg.edu.cn;|罗向阳(1978—),男,教授,E-mail:luoxy_ieu@sina.com
基金资助:
国家自然科学基金(U1804263);国家自然科学基金(62072057);国家自然科学基金(62172435);国家自然科学基金(62002387);中原科技创新领军人才项目(214200510019)

JPEG image steganalysis based on deep extraction of stego noise

FAN Wentong^1,^2,³(),LI Zhenyu^1,^2,³(),ZHANG Tao⁴(),LUO Xiangyang^1,^2,³()

1. PLA Strategic Support Force Information Engineering University,Zhengzhou 450001,China
2. State Key Laboratory of Mathematical Engineering and Advanced Computing,Zhengzhou 450001,China
3. Key Laboratory of Cyberspace Situation Awareness of Henan Province,Zhengzhou 450001,China
4. School of Computer Science and Engineering,Changshu Institute of Technology,Changshu 215500,China

Received:2023-01-15 Online:2023-08-20 Published:2023-10-17
Contact: Zhenyu LI

摘要/Abstract

摘要：

当前基于深度学习的隐写分析方法检测效果受限于其获取的隐写噪声的精确度。为了获取更加准确的隐写噪声,提高隐写分析的准确率,提出了一种基于隐写噪声深度提取的JPEG图像隐写分析方法。首先,设计了隐写噪声深度提取网络,通过有监督的学习方式使网络能够准确地提取载秘图像中包含的隐写噪声;而后,利用设计的模型评价指标选择最优的隐写噪声提取网络;最后,根据隐写噪声的特点设计分类网络,实现载秘图像的检测,并将分类网络与隐写噪声深度提取网络融合得到最终的检测网络。实验在两个大规模的公开数据集(BOSSBase和BOWS2)上进行,针对两种自适应JPEG图像隐写方法(J-UNIWARD和UED-JC)在多个不同嵌入率和图像质量因子条件下构建的载秘图像进行检测。实验结果表明,该方法的检测准确率较性能第二的方法分别提高了约2.22%和0.85%。文中方法通过提取更加准确的隐写噪声,减少了图像内容对隐写分析带来的影响,相比于典型的基于深度学习的JPEG图像隐写分析方法,取得了更好的检测效果。

关键词: JPEG图像隐写分析, 隐写噪声, 卷积神经网络, 深度学习

Abstract:

The performance of steganalysis is limited by the quality of the stego noise obtained by current deep learning-based methods.In order to obtain more accurate stego noise and improve the accuracy of steganalysis,a new method is proposed based on deep extraction of stego noise for JPEG image steganalysis.First,a stego noise deep extraction network is formulated to precisely extract the stego noise from stego images with the supervised trained network.Then,a model evaluation index is proposed to select the most effective network for stego noise extraction.Finally,according to the characteristics of stego noise,a classification network is designed to detect the stego images,which is then combined with the stego noise extraction network to obtain the final detection network.In the steganalysis experiment,two large-scale publicly available datasets(BOSSBase and BOWS2)are used to construct the stego images by two adaptive JPEG image steganography methods (J-UNIWARD and UED-JC) under several embedding rates and quality factors.Experimental results show that the detection accuracy of the method proposed in this article has been improved by up to 2.22% and 0.85%,respectively compared to the second-best performing method.By extracting more accurate stego noise and reducing the impact of image content on steganalysis,the proposed method achieves a better detection performance compared to typical deep learning-based JPEG steganalysis methods.

Key words: JPEG image steganalysis, stego noise, convolutional neural network, deep learning

中图分类号:

TP391.4

范文同,李震宇,张涛,罗向阳. 基于隐写噪声深度提取的JPEG图像隐写分析[J]. 西安电子科技大学学报, 2023, 50(4): 157-169.

FAN Wentong,LI Zhenyu,ZHANG Tao,LUO Xiangyang. JPEG image steganalysis based on deep extraction of stego noise[J]. Journal of Xidian University, 2023, 50(4): 157-169.

图/表 11

图1

图2

图3

表1

表2

图4

图5

表3

图6

图7

表4

参考文献 37

[1]	XIANG S J, LUO X R. Reversible Data Hiding in Encrypted Image Based on Homomorphic Public Key Cryptosystem[J]. Journal of Software, 2016, 27(6):1592-1601.
[2]	MANDAL P C, MUKHERJEE I, PAUL G, et al. Digital Image Steganography:A Literature Survey[J]. Information Sciences, 2022, 609:1451-1488. doi: 10.1016/j.ins.2022.07.120
[3]	NISSAR A, MIR A H. Classification of Steganalysis Techniques:A Study[J]. Digital Signal Processing, 2010, 20(6):1758-1770. doi: 10.1016/j.dsp.2010.02.003
[4]	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.
	ZHAI Liming, JIA Ju, REN Weixiang, et al. Recent Advances in Deep Learning for Image Steganography and Steganalysis[J]. Journal of Cyber Security, 2018, 3(6):2-12.
[5]	FRIDRICH J, KODOVSKY J. Rich Models for Steganalysis of Digital Images[J]. IEEE Transactions on Information Forensics and Security, 2012, 7(3):868-882. doi: 10.1109/TIFS.2012.2190402
[6]	HOLUB V, FRIDRICH J. Low-Complexity Features for JPEG Steganalysis Using Undecimated DCT[J]. IEEE Transactions on Information Forensics and Security, 2014, 10(2):219-228.
[7]	HOLUB V, FRIDRICH J. Phase-Aware Projection Model for Steganalysis of JPEG Images[C]// Media Watermarking,Security,and Forensics 2015. San Francisco: SPIE, 2015:259-269.
[8]	SONG X, LIU F, YANG C, et al. Steganalysis of Adaptive JPEG Steganography Using 2D Gabor Filters[C]// Proceedings of the 3rd ACM Workshop on Information Hiding and Multimedia Security. New York: ACM, 2015:15-23.
[9]	SU W, NI J, HU X, et al. Towards Improving the Security of Image Steganography via Minimizing the Spatial Embedding Impact[J]. Digital Signal Processing, 2022, 131:103758. doi: 10.1016/j.dsp.2022.103758
[10]	邱应强, 蔡灿辉, 曾焕强, 等. 可自纠错的联合式加密图像可逆数据隐藏算法[J]. 西安电子科技大学学报, 2021, 48(1):107-116.
	QIU Yingqiang, CAI Canhui, ZENG Huanqiang, et al. Joint Reversible Data Hiding in Encrypted Images with the Self-Correcting Ability[J]. Journal of Xidian University, 2021, 48(1):107-116.
[11]	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.
	CHEN Junfu, FU Zhangjie, ZHANG Weiming, et al. Review of Image Steganalysis Based on Deep Learning[J]. Journal of Software, 2021, 32(2):551-578.
[12]	KHAN A, SOHAIL A, ZAHOORA U, et al. A Survey of the Recent Architectures of Deep Convolutional Neural Networks[J]. Artificial Intelligence Review, 2020, 53:5455-5516. doi: 10.1007/s10462-020-09825-6
[13]	王晓丹, 李京泰, 宋亚飞. DDAC:面向卷积神经网络图像隐写分析模型的特征提取方法[J]. 通信学报, 2022, 43(5):68-81. doi: 10.11959/j.issn.1000-436x.2022089
	WANG Xiaodan, LI Jingtai, SONG Yafei. DDAC:A Feature Extraction Method for Model of Image Steganalysis Based on Convolutional Neural Network[J]. Journal on Communications, 2022, 43(5):68-81. doi: 10.11959/j.issn.1000-436x.2022089
[14]	TAN S, LI B. Stacked Convolutional Auto-Encoders for Steganalysis of Digital Images[C]// Signal and Information Processing Association Annual Summit and Conference (APSIPA).Piscataway:IEEE, 2014:1-4.
[15]	QIAN Y, DONG J, WANG W, et al. Deep Learning for Steganalysis via Convolutional Neural Networks[C]// Media Watermarking,Security,and Forensics 2015. San Francisco: SPIE, 2015:171-180.
[16]	XU G, WU H Z, SHI Y Q. Structural Design of Convolutional Neural Networks for Steganalysis[J]. IEEE Signal Processing Letters, 2016, 23(5):708-712. doi: 10.1109/LSP.2016.2548421
[17]	YE J, NI J, YI Y. Deep Learning Hierarchical Representations for Image Steganalysis[J]. IEEE Transactions on Information Forensics and Security, 2017, 12(11):2545-2557. doi: 10.1109/TIFS.2017.2710946
[18]	ZHANG R, ZHU F, LIU J, et al. Depth-Wise Separable Convolutions and Multi-Level Pooling for an Efficient Spatial CNN-Based Steganalysis[J]. IEEE Transactions on Information Forensics and Security, 2019, 15:1138-1150. doi: 10.1109/TIFS.10206
[19]	TSANG C F, FRIDRICH J. Steganalyzing Images of Arbitrary Size with CNNs[J]. Electronic Imaging, 2018, 2018(7):121-1-121-8.
[20]	YOU W, ZHANG H, ZHAO X. A Siamese CNN for Image Steganalysis[J]. IEEE Transactions on Information Forensics and Security, 2020, 16:291-306. doi: 10.1109/TIFS.10206
[21]	ZENG J, TAN S, LI B, et al. Pre-Training via Fitting Deep Neural Network to Rich-Model Features Extraction Procedure and Its Effect on Deep Learning for Steganalysis[J]. Electronic Imaging, 2017, 2017(7):44-49.
[22]	ZENG J, TAN S, LI B, et al. Large-Scale JPEG Image Steganalysis Using Hybrid Deep-Learning Framework[J]. IEEE Transactions on Information Forensics and Security, 2017, 13(5):1200-1214. doi: 10.1109/TIFS.2017.2779446
[23]	XU G. Deep Convolutional Neural Network to Detect J-UNIWARD[C]// Proceedings of the 5th ACM Workshop on Information Hiding and Multimedia Security. New York: ACM, 2017:67-73.
[24]	BOROUMAND M, CHEN M, FRIDRICH J. Deep Residual Network for Steganalysis of Digital Images[J]. IEEE Transactions on Information Forensics and Security, 2018, 14(5):1181-1193. doi: 10.1109/TIFS.2018.2871749
[25]	SU A, ZHAO X, HE X. Arbitrary-Sized JPEG Steganalysis Based on Fully Convolutional Network[C]// International Workshop on Digital Watermarking.Heidelberg:Springer, 2021:197-211.
[26]	LIU Q, NI J, JIAN M. Effective JPEG Steganalysis Using Non-Linear Pre-Processing and Residual Channel-Spatial Attention[C]// 2022 IEEE International Conference on Multimedia and Expo (ICME).Piscataway:IEEE, 2022:1-6.
[27]	ZHUO S, JIN Z, ZOU W, et al. RIDNet:Recursive Information Distillation Network for Color Image Denoising[C]// Proceedings of the IEEE/CVF International Conference on Computer Vision Workshops.Piscataway:IEEE, 2019:3896-3903.
[28]	EBRAHIMI M S, ABADI H K. Study of Residual Networks for Image Recognition[C]// Intelligent Computing:Proceedings of the 2021 Computing Conference.Heidelberg:Springer, 2021:754-763.
[29]	SETIADI D R I M. PSNR vs SSIM:Imperceptibility Quality Assessment for Image Steganography[J]. Multimedia Tools and Applications, 2021, 80(6):8423-8444. doi: 10.1007/s11042-020-10035-z
[30]	LIU Z, LIN Y, CAO Y, et al. Swin Transformer:Hierarchical Vision Transformer Using Shifted Windows[C]// Proceedings of the IEEE/CVF International Conference on Computer Vision.Piscataway:IEEE, 2021:10012-10022.
[31]	BAS P, FILLER T, PEVNY T. “Break Our Steganographic System”:The Ins and Outs of Organizing BOSS[C]// International Workshop on Information Hiding.Heidelberg:Springer, 2011:59-70.
[32]	BAS P, FURON T. BOWS-2 Contest(Break Our Watermarking System)(2021)[OB/OL].[2021-12-12]. http://bows2.eclille.fr/.
[33]	HOLUB V, FRIDRICH J. Digital Image SteganographyUsing Universal Distortion[C]// Proceedings of the first ACM Workshop on Information Hiding and Multimedia Security. New York: ACM, 2013:59-68.
[34]	GUO L, NI J, SHI Y Q. Uniform Embedding for Efficient JPEG Steganography[J]. IEEE Transactions on Information Forensics and Security, 2014, 9(5):814-825. doi: 10.1109/TIFS.2014.2312817
[35]	ZHANG J, CHEN K, QIN C, et al. Distribution-Preserving-Based Automatic Data Augmentation for Deep Image Steganalysis[J]. IEEE Transactions on Multimedia, 2021, 24:4538-4550. doi: 10.1109/TMM.2021.3119994
[36]	ZHANG K, ZUO W, CHEN Y, et al. Beyond a Gaussian Denoiser:Residual Learning of Deep CNN for Image Denoising[J]. IEEE Transactions on Image Processing, 2017, 26(7):3142-3155. doi: 10.1109/TIP.83
[37]	ZAMIR S W, ARORA A, KHAN S, et al. Learning Enriched Features for Real Image Restoration and Enhancement[C]// European Conference on Computer Vision.Heidelberg:Springer, 2020:492-511.

	Charbonnier损失	L2损失	L1损失
p_sc	46.86	47.02	46.90
p_ss	53.82	53.74	55.25
p_cc	57.03	56.33	58.83
p_cc -p_ss	3.21	2.59	3.58
P	150.40	121.80	167.90
准确率/%	96.23	94.97	96.74

学习率	p_sc	p_ss	p_cc	P
1.0×10^-3	无穷大	无穷大	无穷大
5.0×10^-4	44.56	73.79	无穷大
2.5×10^-4	46.90	55.25	58.83	1.679
1.0×10^-4	47.13	54.98	56.35	0.646
1.0×10^-5	47.54	54.33	54.97	0.304

隐写算法	嵌入率/ bpnzap	质量因子为75					质量因子为85
隐写算法	嵌入率/ bpnzap	SNdesNet	EWNet	SRNet	CSANet	J-XuNet	SNdesNet	EWNet	SRNet	CSANet	J-XuNet
J-	0.5	0.091 6	0.062 5	0.106 2	0.088 2	0.196 6	0.169 5	0.173 0	0.181 4	0.181 6	0.274 8
UNIWARD	0.4	0.116 6	0.099 5	0.131 3	0.114 7	0.248 4	0.231 1	0.246 5	0.243 5	0.235 3	0.330 9
	0.3	0.198 3	0.174 5	0.225 0	0.196 9	0.296 8	0.280 5	0.301 0	0.304 2	0.302 5	0.397 7
	0.2	0.261 9	0.248 5	0.297 4	0.254 3	0.386 5	0.339 3	0.361 5	0.377 1	0.360 4	0.465 4
	0.1	0.390 1	0.377 5	0.406 4	0.388 3	0.454 4	0.428 5	0.423 0	0.445 3	0.450 3	0.487 6
UED-JC	0.5	0.032 6	0.003 5	0.046 1	0.014 2	0.077 5	0.078 5	0.087 0	0.091 9	0.116 4	0.187 6
	0.4	0.066 5	0.024 5	0.084 1	0.037 5	0.114 0	0.114 6	0.118 5	0.132 3	0.141 7	0.228 4
	0.3	0.087 4	0.039 5	0.131 5	0.051 4	0.157 3	0.168 5	0.160 5	0.204 6	0.166 3	0.284 5
	0.2	0.120 2	0.093 5	0.172 6	0.089 2	0.206 1	0.197 2	0.185 5	0.234 4	0.224 7	0.346 4
	0.1	0.204 6	0.196 5	0.226 1	0.204 1	0.262 4	0.321 9	0.304 5	0.335 1	0.311 4	0.417 3

基于隐写噪声深度提取的JPEG图像隐写分析

JPEG image steganalysis based on deep extraction of stego noise

RichHTML

PDF (PC)

赞

可视化

摘要/Abstract

引用本文

使用本文

图/表 11

参考文献 37

相关文章 15

Metrics

本文评价

推荐阅读 10

[1]	王玉画,高胜,朱建明,黄晨. 自适应差分隐私的高效深度学习方案[J]. 西安电子科技大学学报, 2023, 50(4): 54-64.
[2]	王娟,刘子杉,武明虎,陈关海,郭力权. 融合超分辨率重建技术的多尺度目标检测算法[J]. 西安电子科技大学学报, 2023, 50(3): 122-131.
[3]	谢雯,滑文强,焦李成,王若男. 采用深度学习的极化SAR地物分类方法综述[J]. 西安电子科技大学学报, 2023, 50(3): 151-170.
[4]	周硕,周一青,张冲,邢旺. ResNet使能的OTFS联合信道估计和信号检测[J]. 西安电子科技大学学报, 2023, 50(3): 19-30.
[5]	王柯俨,成吉聪,黄诗芮,蔡坤伦,王威然,李云松. 融合上下文感知注意力的低光图像去雾网络[J]. 西安电子科技大学学报, 2023, 50(2): 23-32.
[6]	张强, 杨欣朋, 赵世祥, 卫栋栋, 韩臻. 注意力机制的SAR图像车辆目标检测网络[J]. 西安电子科技大学学报, 2023, 50(1): 36-47.
[7]	刘晓雯, 郭继昌, 郑司达. 采用渐进式网络的弱监督显著性目标检测算法[J]. 西安电子科技大学学报, 2023, 50(1): 48-57.
[8]	刘博翀, 蔡怀宇, 杨诗远, 李灏天, 汪毅, 陈晓冬. 一种用于自动驾驶场景的轻量级语义分割网络[J]. 西安电子科技大学学报, 2023, 50(1): 118-128.
[9]	张泽欢, 刘强, 国狄非. 面向大规模零样本图像识别的高效算法框架[J]. 西安电子科技大学学报, 2022, 49(6): 103-110.
[10]	李娇娇, 刘志强, 宋锐, 李云松. 一种改进Unet网络的遥感影像分割算法[J]. 西安电子科技大学学报, 2022, 49(6): 67-75.
[11]	张兆宇,田春娜,周恒,田西兰. 联合在线分类的双注意力RGBT孪生网络跟踪[J]. 西安电子科技大学学报, 2022, 49(6): 76-85.
[12]	王侃, 王孟洋, 刘鑫, 田国强, 李川, 刘伟. 融合自注意力机制与CNN-BiGRU的事件检测[J]. 西安电子科技大学学报, 2022, 49(5): 181-188.
[13]	齐佩汉,李冰,谢爱平,高向兰. 欠采样跳频通信信号深度学习重构方法[J]. 西安电子科技大学学报, 2022, 49(4): 1-7.
[14]	马仑,刘鑫,赵斌,王瑞平,廖桂生,张亚静. 利用多头-连体神经网络实现障碍行为识别[J]. 西安电子科技大学学报, 2022, 49(4): 100-108.
[15]	井佩光,李亚鑫,苏育挺. 一种多模态特征编码的短视频多标签分类方法[J]. 西安电子科技大学学报, 2022, 49(4): 109-117.