Fisher Information Determines Capacity of ε-Secure Steganography

  • Tomáš Filler
  • Jessica Fridrich
Part of the Lecture Notes in Computer Science book series (LNCS, volume 5806)

Abstract

Most practical stegosystems for digital media work by applying a mutually independent embedding operation to each element of the cover. For such stegosystems, the Fisher information w.r.t. the change rate is a perfect security descriptor equivalent to KL divergence between cover and stego images. Under the assumption of Markov covers, we derive a closed-form expression for the Fisher information and show how it can be used for comparing stegosystems and optimizing their performance. In particular, using an analytic cover model fit to experimental data obtained from a large number of natural images, we prove that the ±1 embedding operation is asymptotically optimal among all mutually independent embedding operations that modify cover elements by at most 1.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Wang, Y., Moulin, P.: Perfectly secure steganography: Capacity, error exponents, and code constructions. IEEE Transactions on Information Theory, Special Issue on Security (June 2008)Google Scholar
  2. 2.
    Comesana, P., Pérez-Gonzáles, F.: On the capacity of stegosystems. In: Dittmann, J., Fridrich, J. (eds.) Proceedings of the 9th ACM Multimedia & Security Workshop, Dallas, TX, September 20-21, pp. 3–14 (2007)Google Scholar
  3. 3.
    Harmsen, J.J., Pearlman, W.A.: Capacity of steganographic channels. Submitted to IEEE Transactions on Information Theory (2008), http://arxiv.org/abs/0810.4171
  4. 4.
    Filler, T., Fridrich, J., Ker, A.D.: The square root law of steganographic capacity for Markov covers. In: Delp, E.J., Wong, P.W., Memon, N., Dittmann, J. (eds.) Proceedings SPIE, Electronic Imaging, Security and Forensics of Multimedia XI, San Jose, CA, January 18-21 (2009)Google Scholar
  5. 5.
    Ker, A.D., Pevný, T., Kodovský, J., Fridrich, J.: The square root law of steganographic capacity. In: Ker, A., Dittmann, J., Fridrich, J. (eds.) Proceedings of the 10th ACM Multimedia & Security Workshop, Oxford, UK, September 22-23, pp. 107–116 (2008)Google Scholar
  6. 6.
    Ker, A.D.: A capacity result for batch steganography. IEEE Signal Processing Letters 14(8), 525–528 (2007)CrossRefGoogle Scholar
  7. 7.
    Filler, T., Fridrich, J.: Complete characterization of perfectly secure stego-systems with mutually independent embedding operation. In: Proceedings IEEE, International Conference on Acoustics, Speech, and Signal Processing, Taipei, Taiwan, April 19-24 (2009)Google Scholar
  8. 8.
    Ker, A.D.: Estimating steganographic Fisher information in real images. In: Katzenbeisser, S., Sadeghi, A.-R. (eds.) IH 2009. LNCS, vol. 5806, pp. 73–88. Springer, Heidelberg (2009)Google Scholar
  9. 9.
    Kodovský, J., Fridrich, J., Pevný, T.: Statistically undetectable JPEG steganography: Dead ends, challenges, and opportunities. In: Dittmann, J., Fridrich, J. (eds.) Proceedings of the 9th ACM Multimedia & Security Workshop, Dallas, TX, September 20-21, pp. 3–14 (2007)Google Scholar
  10. 10.
    Doob, J.L.: Stochastic processes, 1st edn. Wiley, New York (1953)MATHGoogle Scholar
  11. 11.
    Sullivan, K., Madhow, U., Manjunath, B., Chandrasekaran, S.: Steganalysis for Markov cover data with applications to images. IEEE Transactions on Information Forensics and Security 1(2), 275–287 (2006)CrossRefGoogle Scholar
  12. 12.
    Sidorov, M.: Hidden Markov models and steganalysis. In: Dittmann, J., Fridrich, J. (eds.) Proceedings of the 6th ACM Multimedia & Security Workshop, Magdeburg, Germany, September 20-21, pp. 63–67 (2004)Google Scholar
  13. 13.
    Cachin, C.: An information-theoretic model for steganography. In: Aucsmith, D. (ed.) IH 1998. LNCS, vol. 1525, pp. 306–318. Springer, Heidelberg (1998)CrossRefGoogle Scholar
  14. 14.
    Ker, A.D.: The ultimate steganalysis benchmark? In: Dittmann, J., Fridrich, J. (eds.) Proceedings of the 9th ACM Multimedia & Security Workshop, Dallas, TX, September 20-21, pp. 141–148 (2007)Google Scholar
  15. 15.
    Kay, S.M.: Fundamentals of Statistical Signal Processing, Detection Theory, vol. II. Prentice-Hall, Englewood Cliffs (1998)Google Scholar
  16. 16.
    Filler, T.: Important properties of normalized KL-divergence under HMC model. Technical report, DDE Lab, SUNY Binghamton (2008), http://dde.binghamton.edu/filler/kl-divergence-hmc.pdf
  17. 17.
    Fridrich, J., Goljan, M.: Digital image steganography using stochastic modulation. In: Delp, E.J., Wong, P.W. (eds.) Proceedings SPIE, Electronic Imaging, Security, Steganography, and Watermarking of Multimedia Contents V, Santa Clara, CA, January 21-24, vol. 5020, pp. 191–202 (2003)Google Scholar
  18. 18.
    Goljan, M., Fridrich, J., Holotyak, T.: New blind steganalysis and its implications. In: Delp, E.J., Wong, P.W. (eds.) Proceedings SPIE, Electronic Imaging, Security, Steganography, and Watermarking of Multimedia Contents VIII, San Jose, CA, January 16-19, vol. 6072, pp. 1–13 (2006)Google Scholar
  19. 19.
    Filler, T.: Fisher information determines capacity of ε-secure steganography - proofs. Technical report, SUNY Binghamton (2009), http://dde.binghamton.edu/filler/pdf/Fill09ihwproofs.pdf

Copyright information

© Springer-Verlag Berlin Heidelberg 2009

Authors and Affiliations

  • Tomáš Filler
    • 1
  • Jessica Fridrich
    • 1
  1. 1.Department of ECESUNY BinghamtonUSA

Personalised recommendations