DUAN Huai-feng,YANG Yu,LEI Min,et al.Improvement of Image Universal Blind Detection based on Training Set Construction[J].Journal of Chengdu University of Information Technology,2016,(01):70-75.
基于训练集构造的图像通用盲检测算法改进
- Title:
- Improvement of Image Universal Blind Detection based on Training Set Construction
- 文章编号:
- 2096-1618(2016)01-0070-06
- 关键词:
- 通用盲检测; 泛化能力; SPAM; Rich Model; 实用性
- 分类号:
- TP309.2
- 文献标志码:
- A
- 摘要:
- 现有通用盲检测技术普遍存在泛化问题,导致检测器实用性大大下降。根据正交设计原则构建隐写率 失匹配集合,隐写算法失匹配集合和图像源失匹配集合,分别分析检测SPAM 分析算法和Rich Model 分析算法在 隐写率失匹配,隐写算法失匹配和图像源失匹配方面的检测率。并根据测试结果提出通过训练小隐写率图像集, 训练多类隐写算法,图像预分类和改进IQM 分析算法几种方案解决泛化问题,实验结果显示经过改进后隐写分析 算法性能得到明显提升。
- Abstract:
- The practicability of existing universal blind detection reduced greatly due to the generalization problem. Ac- cording to the principle of orthogonal design, this paper builds three sample sets of embedding rates mismatch, embed- ding algorithms mismatch and image sources mismatch between the training sample and the testing sample. The three sets are used to test the detection error rates of SPAM and Rich Model in the case of embedding rates mismatch, embed- ding algorithm mismatch and image source mismatch. This paper proposed several methods to improve the generalization ability of the universal blind detection, including training the sample by small embedding rates, learning various kinds of embeddingalgorithms, pre-classifying the testing sample and improving the IQM algorithm. The results show that the that the performance of the improved algorithm is significantly improved.
参考文献/References:
[1] Dumitrescu S, Wu X L, Wang Z. Detection of
LSB Steganography via Sample Pair Analysis[J].
IEEE Trans. on Signal Processing, 2003, 51
(7): 1995 - 2007.
[2] Zhihua Xia, Xinhui Wang, Xingming Sun, et al.
Steganalysis of least significant bit matching using
multi-order differences[ J]. Security and Commu-
nication Networks, 2014, 7(8): 1283-1291.
[3] Zhihua Xia, Xinhui Wang, Xingming Sun, et al.
Steganalysis of LSB matching using differences be-
tween nonadjacent pixels[ C]. Multimedia Tools
and Applications,2014.
[4] 万宝吉, 张涛, 侯晓丹,等. 基于Boosting 算法
融合的图像隐写分析方法[ J]. 计算机工程,
2013, 39(12): 148-151.
[5] Pevny T, Bas P, Fridrich J. Steganalysis by Sub-
tractivePixel AdjacencyMatrix[ J]. IEEE Transac-
tions onInformation Security and Forensics, 2010,
5(2): 215-224.
[6] Fridrich J, Kodovsky J. Rich Models for Steganal-
ysis of Digital Images[ J]. IEEE Transactions on
Information Forensics and Security, 2012, 7(3).
[7] 李彩霞. 空域自适应EA 隐秘算法及分析算法
的研究(D). 大连:大连理工大学, 2013.
[8] Kodovsky J, Fridrich J, Holub. V. : Ensemble
Classifiers for Steganalysis of Digital Media. IEEE
Trans[J]. Information Forensics Security, 2011.
[9] Kodovsky J, Pevny T, Fridrich J. Modern stega-
nalysis can detect YASS[C]. In Proc. SPIE, E-
lectronic Imaging, Security Forensics of Multime-
diaXII, San Jose, CA, 2010, 7541:201-211.
[10] Breiman L. Random forest[ J]. Machine Learn-
ing, 2001, 45(1):5-32.
[11] Filler T, Pevny T, Bas P. BOSS ( Break Our
Steganography System) 2014 [EB/ OL]. Availa-ble: http: / / agents. fel. cvut. cz/ stegodata/ Boss-
Base-1. 01-cover. tar. bz2,2014.
[12] Sharp T. An Implementation of Key-based Digital
Signal Steganography[ C]. In Proceedings of the
4th International Workshop on Information Hid-
ing, London, UK, 2001: 13-26.
[13] Luo W, Huang F, Huang J. Edge adaptive image
steganographybased on LSB matching revisited
[J]. IEEE Trans. Inform. Forensics Security,
2010, 5(2): 201 – 214.
[14] Pevny T, Filler T, Bas P. Eds. :Using High-di-
mensional Image Models to Perform Highly Unde-
tectable Steganography[ J], in Information Hid-
ing, 12th Int. Workshop, Calgary, Canada,
2010, 6387: 161 – 177.
[15] Filler T, Fridrich J, Judas J. Minimizing Embed-
ding Impact in Steganography Using Trellis-Coded
Quantization [ J]. In Proceedings SPIE, EI,
Media Forensics and Security XII, San Jose,
CA, 2010: 1-14.
[16] Filler T, Judas J, Fridrich J. Minimizing Addi-
tive Distortion in Steganography Using Syndrome-
Trellis Codes[J]. IEEE Transactions on Informa-
tion Forensics and Security,2011, 6(3): 920-
935.
[17] Anxin Wu, Guorui Feng. Payload Mismatch De-
tection of Image Steganalysis Using Ensemble
Linear Discriminant Clustering[ J]. IEEE Trans.
on Signal Processing, 2015: 1-4.
[18] Yu Dong, Tao Zhang, Xiaodan Hou, et al. A
New Steganalysis Paradigm Based on Image Re-
trieval of Similar Image-Inherent Statistical Prop-
erties and Outlier Detection[ J]. Wireless Com-
munications & Signal Processing ( WCSP ),
2015: 1-5.
备注/Memo
收稿日期:2016-02-02 基金项目:国家自然科学基金资助项目(61310306028),浙江省自然 科学基金资助项目(Y15F020053)