
浏览全部资源
扫码关注微信
武警工程大学 电子技术系 网络与信息安全武警部队重点实验室,陕西 西安,710086
[ "孙寿健(1990-),男,山东德州人,硕士研究生,主要从事图像信息隐藏方面的研究。E-mail:ssj1990220@163.com" ]
[ "魏立线(1966-),男,陕西户县人,硕士研究生,教授,主要从事信息安全的研究" ]
[ "刘佳(1982-),男,河南汝州人,博士,讲师,主要从事信息安全的研究" ]
[ "苏光伟(1984-),男,山东淄博人,硕士研究生,主要从事信息安全的研究。" ]
收稿日期:2014-04-02,
修回日期:2014-05-22,
纸质出版日期:2015-04-05
移动端阅览
孙寿健, 魏立线, 刘佳, 苏光伟. 一种基于权重融合的JPEG隐写分析方法[J]. 液晶与显示, 2015,30(2): 326-332
SUN Shou-jian, WEI Li-xian, LIU Jia, SU Guang-wei. A rights-based re-integration method for JPEG steganalysis[J]. , 2015,30(2): 326-332
孙寿健, 魏立线, 刘佳, 苏光伟. 一种基于权重融合的JPEG隐写分析方法[J]. 液晶与显示, 2015,30(2): 326-332 DOI: 10.3788/YJYXS20153002.0326.
SUN Shou-jian, WEI Li-xian, LIU Jia, SU Guang-wei. A rights-based re-integration method for JPEG steganalysis[J]. , 2015,30(2): 326-332 DOI: 10.3788/YJYXS20153002.0326.
针对JPEG图像通用隐写检测中检测效率低、训练时间长的问题
提出一种基于集成分类器的新检测方法。算法以CC-PEV为特征对图像进行描述并作为隐写分析特征;然后
随机构造若干个特征子空间
用bootstrap方法构造图像训练子集
分别进行训练得到数个基分类器;根据基分类器的分类结果赋予基分类器不同的权重
将基分类器的结果按照其权重进行融合得到最终的结果。本文对该算法进行了测试
对它的集成性、检测准确率和训练时间进行分析。实验结果表明
相对于传统的集成方法
本文方法用自举方法构造训练集、随机方法构造子特征空间、赋予基分类器不同权重进行融合能够显著地提高算法准确率。本文方法相对于SVM和传统的集成分类方法
具有更高的检测率
对于特征维数更大的图像检测
具有更好的拓展性和一般适用性。
Aiming at the problem of the high dimension features and complicated training in image universal steganalysis technology
this paper proposes a steganalysis method based on ensemble classifier. The method designs a ensemble classifier which combines the difference of base classifier and weighted fusion. Specifically
the algorithm uses CC-PEV as steganalysis features. Then
the method randomly structures several feature subspace
uses the bootstrap method to construct image training subsets
and train them to get a number of base classifiers. Finally
every base classifier is given weight that different from each other according to the classification result and then the final forecast is got after the combination of classification results with their respective weight. The experimental results show that using ensemble classifier for training and giving base classifiers with different weights fusion could effectively improve the accuracy of the algorithm.
张军, 熊枫, 张丹. 图像隐写分析技术综述 [J]. 计算机工程, 2013, 39(4): 165-168, 172. Zhang J, Xiong F, Zhang D. Overview on image steganalysis technology [J]. Computer Engineering, 2013, 39(4): 165-168, 172. (in Chinese)[2] 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. Wang S Z, Zhang X P, Zhang W M. Recent advances in image-based steganalysis research [J]. Chinese Journal of Computers, 2009, 32(7): 1247-1263. (in Chinese)[3] 韩晓东. JPEG 图像隐写分析技术研究 [D]. 郑州: 解放军信息工程大学, 2009. Han X D. Research on steganlysis for JPEG images [D]. Zhengzhou: PLA Information Engineering University, 2009. (in Chinese)[4] Kodovsky J,Fridrich J, Holub V. Ensemble classifiers for steganalysis of digital media [J]. IEEE Transactions on Information Forensics and Security, 2012, 7(2): 432-444.[5] Kodovský J, Fridrich J. Calibration revisited [C]. Proceedings of the 11th ACM Workshop on Multimedia and Security, 2009: 63-74.[6] 尹华. 面向高位和不平衡数据分类的集成学习研究 [D]. 武汉:武汉大学, 2012. Yin H. Research of ensemble learning for high-dimensional and imbalanced data classification [D]. Wuhan:Wuhan University, 2012. (in Chinese)[7] Kodovsk ý J, Fridrich J.Steganalysis of JPEG images using rich models [C]. International Society for Optics and Photonics, 2012: 83030A-83030A-13.[8] Kodovský J, Fridrich J.Steganalysis in high dimensions: Fusing classifiers built on random subspaces [C]. International Society for Optics and Photonics, 2011: 78800L-78800L-13.[9] 李开达. 隐藏信息检测综合评估及融合技术研究 [D]. 郑州: 解放军信息工程大学, 2012. Li K D.Research on comprehensive evaluation and fusion of steganalysis techniques [D]. Zhengzhou: PLA Information Engineering University, 2012. (in Chinese)[10] Westfeld A.F5—a steganographic algorithm [M]. Springer Berlin Heidelberg, 2001: 289-302.[11] Ullerich C, Westfeld A.Weaknesses of MB2 [M]. Springer Berlin Heidelberg, 2008: 127-142.[12] Fridrich J, Goljan M, Hogea D. Attacking the outguess [C]. Proceedings of the ACM Workshop on Multimedia and Security, 2002.[13] 徐茂. 基于SVM的JPEG图像隐写分析研究[D]. 长沙: 湖南大学, 2012. Xu M.Steganalysis for JPEG Image Based on SVM [D].Changsha: Hunan University, 2012. (in Chinese)
0
浏览量
24
下载量
0
CSCD
关联资源
相关文章
相关作者
相关机构
京公网安备11010802024621