提出了一种基于k均值聚类的混合异构图像隐写分析算法. 在训练阶段，根据图像纹理复杂度对图像库
进行聚类，并针对每一类图像训练相应的分类器. 在测试阶段，根据测试图像的纹理复杂度对其进行类别判断，然
后送至相应类别的分类器中进行隐写检测，从而减弱了失配状态对现有隐写分析算法造成的影响. 实验结果表明，
该算法较好地提高了现有隐写分析算法的检测精度.
A new image steganalysis method using k-means clustering is presented. In the training phase,
the input images are classified to several classes using k-means clustering according to texture complexity.
The training process is specialized for each class separately. In the testing phase, the given test image is first
classified to the class it belongs to according to its texture complexity. It is then submitted to the corresponding
steganalysis classifier. The proposed method can reduce mismatch penalty considerably. Experimental results
demonstrate that the method can significantly enhance detection accuracy of existing steganalysis methods.
[1] WESTFELD A, PFITZMANN A. Attacks on steganographic systems [C]//Proceedings of Information Hiding Workshop, Lecture Notes in Computer Science: 1768 (?). Berlin: Springer, 1999: 61-76.
[2] FRIDRICH J, GOLJAN M, DU R. Detecting LSB Steganography in Color and Gray-Scale Images [J]. IEEE Transaction on Multimedia, 2001, 8(4): 22-28.
[3] ZHANG Jun, COX I J, DOЁRRG. Steganalysis for LSB matching in images with high-frequency noise [ C]// Proceedings ofIEEE International Workshop on Multimedia Signal Processing. Chania Crete, Greece: IEEE, 2007: 385-388.
[4] GAOYunkai, LIXiaolong, YANG Bin. Detecting LSB matching by characterizing the amplitude of histogram [C] //Proceedings of IEEE International Conference on Acoustics Speech and Signal Processing. Piscataway, NJ: IEEE, 2009: 1505-1508.
[5] ZHENG Ergong, PING Xijian, ZHANG Tao. Local linear transform and new features of histogram characteristic functions for steganalysis of least significant bit matching steganography [J]. KSII Transactions on Internet and Information System, 2011, 5(4): 840-855.
[6] HARMSEN J, PEARLMAN W A. Steganalysis of additive noise modelable information hiding [C] // Proceedings of the SPIE, Security, Steganography, and Watermarking of Multimedia Contents VI: 5020. San Jose, CA: SPIE, 2003: 131-142.
[7] KER A D. Steganalysis of LSB matching in grayscale images [J]. IEEE Signal Processing Letters, 2005, 12(6): 441-444.
[8] FARID H, LYUS W. Detecting hidden messages using higher-order statistics and support vector machines [C] //Proceedings of 5th International Workshop on Information Hiding, Lecture Notes in Computer Science: 2578.Berlin Heidelberg: Springer, 2002:340-354.
[9] GOLJAN M, FRIDRICH J, HOLOTYAK T. New blind steganalysis and its implications [C]// Proceedings of SPIE, Security, Steganography, and Watermarking of Multimedia Contents: 6072. San Jose, CA: SPIE 2006:1-13.
[10] XUAN Guorong, SHI Yunqing, GAO Jianjiong. Steganalysis based on multiple features formed by statistical moments of wavelet characteristic functions [C] //Proceedings of 7th Information Hiding Workshop, Lecture Notes in Computer Science 3727. Berlin: Springer-Verlag, 2005: 262-277.
[11] SHI Yunqing, XUAN Guorong, GAO Jianjiong. Effective steganalysis based on statistical moments of wavelet characteristic function [C] //Proceedings of International Conference on Information Technology: Coding and Computing: 1. Las Vegas, Nevada, IEEE, 2005: 786-773.
[12] WANG Ying, MOULIN P. Optimized feature extraction for learning-based image steganalysis [J]. IEEE Transaction on Information Forensics and Security, 2007, 2(1): 31-45.
[13] LI Bin, HUANGJiwu, SHI Yunqing. Textural features based universal steganalysis[ C]//Edward J. Security, Forensics, Steganography and Watermarking of Multimedia Contents X: 6819, Bellingham, Wash: SPIE, 2008:1201-1212.
[14] PEVNÝ T, BAS P, FRIDRICH J. Steganalysis by subtractive pixel adjacency matrix [J]. IEEE Transactions on Information Forensics and Security, 2010, 5(2): 215-224.
[15] YANG C H, WENG C Y, WANG S J, SUN HM. Adaptive data hiding in edge areas of images with spatial LSB domain systems [J]. IEEE Transactions on Information Forensics and Security, 2008, 3(3): 488-497.
[16] LUO Weiqi, HUANG Fangjun, HUANGJiwu. Edge adaptive image steganography based on LSB matching revisited [J]. IEEE Transaction on Information Forensics and Security, 2010, 5(2): 201-214.
[17] AMIRKHANI Hossein, RAHMATI Mohammad. New framework for using image contents in blind steganalysis systems [J]. Journal of Electronic Imaging, 2011, 20(1): 1-14.
[18] 汪太月，李志明. 一种广义高斯分布的参数快速估计方法 [J]. 工程地球物理学报，2006, 3(3): 172-176.
WANG Taiyue, LI Zhiming. A fast parameter estimation of generalized gaussian distribution [J] Chinese Journal of Engineering Geophysics, 2006, 3(3): 172-176. (in Chinese)
CANNY J. A computational approach to edge detection [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1986, 8(6): 679-698.
[19] CAI Kaiwei, LI Xiaolong, ZENG Tieyong. Reliable histogram features for detecting LSB matching [C] // Proceedings of IEEE International Conference on Image Processing. Piscataway NJ: IEEE, 2010:1761-1764.
[20] TAN Shunquan, LI Bin. Targeted steganalysis of edge adaptive image steganography based on LSB matching revisited using B-spline fitting [J]. IEEE Signal Processing Letters, 2012, 19(6): 336-339.
