A Study of the Two-Way Effects of Cover Source Mismatch and Texture Complexity in Steganalysis
Cover source mismatch (CSM) occurs when a detection classifier for steganalysis trained on objects from one cover source is tested on another source. However, it is very hard to find the same sources as suspicious images in real-world applications. Therefore, the CSM is one of the biggest stumbling blocks to hinder current classifier based steganalysis methods from becoming practical. On the other hand, the texture complexity (of digital images) also plays an important role in affecting the detection accuracy of steganalysis. Previous work seldom conduct research on the interaction between the two factors of the CSM and the texture complexity. This paper studies the interaction between the two factors, aiming to improve the steganalysis accuracy. We propose a effective method to measure the texture complexity via image filtering, and use the two-way analysis of variance to study the interaction between the two factors. The experimental results have shown that the interaction between the two factors affects the detection accuracy significantly. We also design a method to improve the detection accuracy of steganalysis by utilizing the interaction of the two factors.
KeywordsCover source mismatch Texture complexity Analysis of variance Steganalysis
This work was supported in part by the National Natural Science Foundation of China (No. 61272540, No. U1536204), the National Key Technology R&D Program (No. 2014BAH41B00, No. 2015AA016004), and in part by the Natural Science Foundation of Anhui province (No. 1508085MF115, No. 1608085MF142).
- 2.Cancelli, G., Doërr, G., Barni, M., Cox, I.J.: A comparative study of\(\pm \)steganalyzers. In: 2008 IEEE 10th Workshop on Multimedia Signal Processing, pp. 791–796. IEEE (2008)Google Scholar
- 3.Goljan, M., Fridrich, J., Holotyak, T.: New blind steganalysis and its implications. In: Electronic Imaging, pp. 607201–607201. International Society for Optics and Photonics (2006)Google Scholar
- 4.Huang, F., Zhong, Y., Huang, J.: Improved algorithm of edge adaptive image steganography based on LSB matching revisited algorithm. In: Shi, Y.Q., Kim, H.-J., Pérez-González, F. (eds.) IWDW 2013. LNCS, vol. 8389, pp. 19–31. Springer, Heidelberg (2014). doi: 10.1007/978-3-662-43886-2_2 Google Scholar
- 5.Iversen, G.R., Gergen, M.: Statistics: the Conceptual Approach. Springer Science & Business Media, New York (2012)Google Scholar
- 6.Ker, A.D., Bas, P., Böhme, R., Cogranne, R., Craver, S., Filler, T., Fridrich, J., Pevnỳ, T.: Moving steganography and steganalysis from the laboratory into the real world. In: Proceedings of the First ACM Workshop on Information Hiding and Multimedia Security, pp. 45–58. ACM (2013)Google Scholar
- 7.Ker, A.D., Pevnỳ, T.: A mishmash of methods for mitigating the model mismatch mess. In: IS&T/SPIE Electronic Imaging, pp. 90280I–90280I. International Society for Optics and Photonics (2014)Google Scholar
- 8.Kodovskỳ, J., Sedighi, V., Fridrich, J.: Study of cover source mismatch in steganalysis and ways to mitigate its impact. In: IS&T/SPIE Electronic Imaging, pp. 90280J–90280J. International Society for Optics and Photonics (2014)Google Scholar
- 10.Qian, Y., Dong, J., Wang, W., Tan, T.: Deep learning for steganalysis via convolutional neural networks. In: IS&T/SPIE Electronic Imaging, pp. 94090J–94090J. International Society for Optics and Photonics (2015)Google Scholar
- 11.Rice, J.: Mathematical Statistics and Data Analysis. Nelson Education (2006)Google Scholar
- 13.Walpole, R.E., Myer, R., Myers, S.I., Keying, E.Y.: Essentials of Probabilty & Statistics for Engineers & Scientists. Pearson Higher Ed., Upper Saddle River (2012)Google Scholar