基于数字化卡登格的生成图像隐写方案

doi:10.3969/j.issn.1671-1122.2021.02.009

信息网络安全 ›› 2021, Vol. 21 ›› Issue (2): 70-77.doi: 10.3969/j.issn.1671-1122.2021.02.009

基于数字化卡登格的生成图像隐写方案

王耀杰¹^,², 杨晓元¹^,², 刘文超¹^,²()

1.武警工程大学网络与信息安全武警部队重点实验室,西安 710086
2.武警工程大学密码工程学院,西安 710086

收稿日期:2020-11-17 出版日期:2021-02-10 发布日期:2021-02-23
通讯作者: 刘文超 E-mail:liuwenchao@163.com
作者简介:王耀杰（1990—）,男,河南,博士研究生,主要研究方向为信息安全、深度学习|杨晓元（1959—）,男,湖南,教授,硕士,主要研究方向为信息安全、信息隐藏技术、密码学|刘文超（1994—）,男,甘肃,讲师,硕士,主要研究方向为信息安全、信息隐藏技术
基金资助:
国家重点研发计划(2017YFB0802000);国家自然科学基金(61379152);国家自然科学基金(61403417)

Generative Image Steganography Based on Digital Cardan Grille

WANG Yaojie¹^,², YANG Xiaoyuan¹^,², LIU Wenchao¹^,²()

1. Key Laboratory for Network and Information Security of Chinese Armed Police Force, Engineering University of PAP, Xi’an 710086, China
2. College of Cryptographic Engineering, Engineering College of PAP, Xi’an 710086, China

Received:2020-11-17 Online:2021-02-10 Published:2021-02-23
Contact: LIU Wenchao E-mail:liuwenchao@163.com

摘要/Abstract

摘要：

针对图像隐写中载体修改会留有修改痕迹的问题,文章提出一种基于数字化卡登格的图像隐写方案。首先自动生成数字化卡登格,作为信息隐写和提取的密钥;其次以空白图像作为载体,根据数字化卡登格的区域填充秘密信息;在整个隐写过程保持秘密消息不变的情况下,再利用深度生成模型补全受损的图像,生成具有自然语义的含密图像。文章方案不受载体类型的限制,还可以扩展到其他数字媒体。实验结果表明,新的图像隐写方案具有良好的可行性,相比其他隐写方法,文中方案在实际通信中安全性较高。

关键词: 图像隐写, 数字化卡登格, 图像补全, 生成对抗网络

Abstract:

In order to solve the problem that the carrier modification in image steganography will leave the modification trace, this paper proposes an image steganography scheme based on digital Cardan Grille. The scheme first generates the digital Cardan Grille automatically as the key of information steganography and extraction. Secondly, the blank image is used as the carrier to fill in the secret information according to the area of digital Cardan Grille. Under the condition that the secret message is kept unchanged in the whole steganography process, the damaged image of the model is generated by depth to generate the stego image with natural semantics. The results of experiments on CelebA dataset and LUSN dataset show that the new steganography scheme is feasible. Compared with other steganographic methods, this scheme is more secure in actual communication.

Key words: image steganography, digital Cardan Grille, semantic completion, GAN

中图分类号:

TP309

王耀杰, 杨晓元, 刘文超. 基于数字化卡登格的生成图像隐写方案[J]. 信息网络安全, 2021, 21(2): 70-77.

WANG Yaojie, YANG Xiaoyuan, LIU Wenchao. Generative Image Steganography Based on Digital Cardan Grille[J]. Netinfo Security, 2021, 21(2): 70-77.

图/表 9

图1

图2

图3

图4

图5

表1

图6

图7

表2

参考文献 24

[1]	ZHANG Ru, DONG Shiqi, LIU Jianyi. Invisible Steganography Via Generative Adversarial Networks[J]. Multimedia Tools and Applications, 2019,18(2): 34-39.
[2]	GOODFELLOW I, POUGET-ABADIE J, MIRZA M. Generative Adversarial Networks[EB/OL]. http://arxiv.org/abs/1406.2661, 2014-06-10.
[3]	LIU Jia, ZHOU Tianping, ZHANG Zhuo, et al. Digital Cardan Grille: A Modern Approach for Information Hiding[EB/OL]. https://arxiv.org/abs/1803.09219.pdf, 2018-03-25.
[4]	SURHONE L M, TENNOE M T, HENSSONOW S F, et al. Card Catalog (cryptology)[M]. Montana USA: Betascript Publishing, 2010.
[5]	VOLKHONSKIY D, NAZAROV I, BORISENKO B, et al. Steganographic Generative Adversarial Networks[EB/OL]. http://arxiv.org/abs/1703.05502.pdf, 2017-03-16.
[6]	WANG Yaojie, NIU Ke, YANG Xiaoyuan. Image Steganography Scheme Based on Generative Adversarial Network[J]. Journal of Computer Applications, 2018,38(10): 177-182.
	王耀杰, 钮可, 杨晓元. 基于生成对抗网络的信息隐藏方案[J]. 计算机应用, 2018,38(10): 177-182.
[7]	KE Yan, LIU Jia, ZHANG Minqing, et al. Steganography Security: Principle and Practice[J]. IEEE Access, 2018,1(99): 15-19.
[8]	FRIDRICH J. Steganography in Digital Media: Principles, Algorithms, and Applications[M]. Cambridge: Cambridge University Press, 2010.
[9]	LIU Mingming, LI Xinchao. Information Hiding Method Based on Cardan Grille[J]. Journal of Engineering University of PAP, 2018. 34(6): 58-62.
	刘明明, 李新超. 基于卡登格子的信息隐藏方法[J]. 武警工程大学学报, 2018,34(6): 58-62.
[10]	ODENA A, OLAH C, SHLENS J. Conditional Image Synjournal with Auxiliary Classifier GANs[J]. Computer Science, 2016,18(1): 14-20.
[11]	CUI Zhongli, WANG Jiazhen. Generalizable Steganography Algorithm Based on DCT Domain[J]. Computer Engineering and Design, 2006,27(13): 2449-2451.
	崔忠立, 王嘉祯. 可泛化的基于DCT域的隐写术算法[J]. 计算机工程与设计, 2006,27(13): 2449-2451.
[12]	NOVOTNY M A. Matrix Products with Applications to Classical Statistical Mechanics[J]. Journal of Biomedical Materials Research, 1978,56(3): 452-458. doi: 10.1002/1097-4636(20010905)56:3<452::aid-jbm1115>3.0.co;2-1 URL pmid: 11372064
[13]	RAGAB A H M, ISMAIL N A, ALLAH O S F. AN Efficient Message Digest Algorithm (MD) for Data Security[C]//IEEE. 2001 Tencon IEEE Region 10th International Conference on Electrical & Electronic Technology, November 8-12, 2001, Seattle, USA. New York: IEEE, 2001: 723-734.
[14]	ZHANG Hongying, PENG Qizong. Overview of Digital Image Restoration Technology[J]. Journal of Image and Graphics, 2007,1(1): 1-10.
	张红英, 彭启琮. 数字图像修复技术综述[J]. 中国图象图形学报, 2007,1(1): 1-10.
[15]	LIU Jia, KE Yan, LEI Yu, et al. The Reincarnation of Grille Cipher: A Generative Approach[EB/OL]. https://arxiv.org/abs/1804.06514.pdf, 2018-04-15.
[16]	YEH R A, CHEN C, LIM T Y, et al. Semantic Image Inpainting with Deep Generative Models[J]. IEEE Computer Society, 2016,16(7): 83-88.
[17]	RADFORD A, METZ L, CHINTALA S. Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks[J]. Computer Science, 2015,34(8): 127-131.
[18]	HOU Xuedong. Image Translation Based on Generative Adversarial Networks[D]. Nanjing: Nanjing University of Posts and Telecommunications, 2018.
	候学东. 基于生成对抗网络的图像翻译[D]. 南京: 南京邮电大学, 2018.
[19]	AMOS B. Image Completion with Deep Learning in TensorFlow[EB/OL]. https://github.com/bamos/dcgan-completion.tensorflow, 2018-09-23.
[20]	MANDGAONKAR V S, KULKARNI C V. No Reference Image Quality Assessment[C]// IEEE. 2014 Annual IEEE India Conference (INDICON), November 8-12, 2014, New York, USA. New York: IEEE, 2015: 1241-1248.
[21]	SURYAWANSHI G R, MALI S N. Study of Effect of DCT Domain Steganography Techniques in Spatial Domain for JPEG Images Steganalysis[J]. International Journal of Computer Applications, 2015,127(6): 16-20.
[22]	ZHEN Yu, CHEN Kefei. Analysis and Improvement on RS Detection Algorithm[J]. Computer Engineering, 2008,34(8): 168-170.
[23]	YAN Huijuan, QIN Jie. Trojan Horse Detection Method Based on Nonlinear SVM Model[J]. Computer Engineering, 2011,11(2): 134-137.
[24]	CHEN Tao, LU Shijian, FAN Jiayuan. S-CNN: Subcategory-aware Convolutional Networks for Object Detection[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017,16(4): 12-17.

软件平台	Tensorflow v0.12
硬件环境	CPU	i7-8250U 3.2 GHz
	RAM	16GB DDR3 1600 MHz
	GPU	NVIDIA 1080

检测算法	检测准确率
检测算法	分组准确率	平均准确率
DCT域隐写分析	0.534	0.556
DCT域隐写分析	0.579	0.556
RS检测	0.510	0.516
RS检测	0.523	0.516
非线性SVM检测	0.573	0.562
非线性SVM检测	0.551	0.562
S-CNN检测	0.672	0.658
S-CNN检测	0.644	0.658

基于数字化卡登格的生成图像隐写方案

Generative Image Steganography Based on Digital Cardan Grille

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

图/表 9

参考文献 24

相关文章 2

编辑推荐

Metrics

本文评价

[1]	彭中联, 万巍, 荆涛, 魏金侠. 基于改进CGANs的入侵检测方法研究[J]. 信息网络安全, 2020, 20(5): 47-56.
[2]	毕新亮, 杨海滨, 杨晓元, 黄思远. 基于StarGAN的生成式图像隐写方案[J]. 信息网络安全, 2020, 20(12): 64-71.