A Feature Selection Methodology for Steganalysis

  • Yoan Miche
  • Benoit Roue
  • Amaury Lendasse
  • Patrick Bas
Part of the Lecture Notes in Computer Science book series (LNCS, volume 4105)


This paper presents a methodology to select features before training a classifier based on Support Vector Machines (SVM). In this study 23 features presented in [1] are analysed. A feature ranking is performed using a fast classifier called K-Nearest-Neighbours combined with a forward selection. The result of the feature selection is afterward tested on SVM to select the optimal number of features. This method is tested with the Outguess steganographic software and 14 features are selected while keeping the same classification performances. Results confirm that the selected features are efficient for a wide variety of embedding rates. The same methodology is also applied for Steghide and F5 to see if feature selection is possible on these schemes.


Support Vector Machine Feature Selection Feature Selection Method Stego Image Noise Estimation 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Fridrich, J. (ed.): IH 2004. LNCS, vol. 3200, pp. 67–81. Springer, Heidelberg (2004)CrossRefGoogle Scholar
  2. 2.
    Dumitrescu, S., Wu, X., Wang, Z.: Detection of LSB steganography via sample pair analysis. IEEE transactions on Signal Processing, 1995–2007 (2003)Google Scholar
  3. 3.
    Roue, B., Bas, P., Chassery, J.-M.: Improving lsb steganalysis using marginal and joint probabilistic distributions. In: Multimedia and Security Workshop, Magdeburg (2004)Google Scholar
  4. 4.
    Lyu, S., Farid, H.: Detecting hidden messages using higher-order statistics and support vector machines. In: Petitcolas, F.A.P. (ed.) IH 2002. LNCS, vol. 2578, pp. 340–354. Springer, Heidelberg (2003)CrossRefGoogle Scholar
  5. 5.
    Pevný, T., Fridrich, J.: Towards Multi-class Blind Steganalyzer for JPEG Images. In: Barni, M., Cox, I., Kalker, T., Kim, H.-J. (eds.) IWDW 2005. LNCS, vol. 3710, pp. 39–53. Springer, Heidelberg (2005)CrossRefGoogle Scholar
  6. 6.
    Weston, J., Mukherjee, S., Chapelle, O., Pontil, M., Poggio, T., Vapnik, V.: Feature selection for SVMs. In: Leen, T.K., Dietterich, T.G., Tresp, V. (eds.) NIPS, pp. 668–674. MIT Press, Cambridge (2000)Google Scholar
  7. 7.
    Efron, B., Tibshirani, R.: An Introduction to the Bootstrap. Chapman and Hall, London (1993)MATHGoogle Scholar
  8. 8.
    Zhang, T.: An introduction to support vector machines and other kernel-based learning methods. AI Magazine, 103–104 (2001)Google Scholar
  9. 9.
    Rossi, F., Lendasse, A., François, D., Wertz, V., Verleysen, M.: Mutual information for the selection of relevant variables in spectrometric nonlinear modelling. Chemometrics and Intelligent Laboratory Systems 80, 215–226 (2006)CrossRefGoogle Scholar
  10. 10.
    Provos, N.: Defending against statistical steganalysis. In: USENIX (ed.) Proceedings of the Tenth USENIX Security Symposium, Washington, DC, USA, August 13–17 (2001)Google Scholar
  11. 11.
    Ronneberger, O.: Libsvmtl extensions to libsvm (2004), http://lmb.informatik.uni-freiburg.de/lmbsoft/libsvmtl/

Copyright information

© Springer-Verlag Berlin Heidelberg 2006

Authors and Affiliations

  • Yoan Miche
    • 1
  • Benoit Roue
    • 2
  • Amaury Lendasse
    • 1
  • Patrick Bas
    • 1
    • 2
  1. 1.Laboratory of Computer and Information ScienceHelsinki University of TechnologyHutFinland
  2. 2.Laboratoire des Images et des Signaux de GrenobleSaint Martin d’HèresFrance

Personalised recommendations