Steganalysis of LSB Matching in WAV Audio

DOI

10.2991/iccsee.2013.258How to use a DOI?

Keywords

LSB matching, steganalysis, local correlation, wav-elet de-noising , support vector machine (SVM)

Abstract

To expose the existence of hidden message, this paper presents a steganalysis method for LSB (Least signify-cant bit) matching steganography in wav audio based on multi-order Markov feature. Noise sequences are extracted respectively based on two methods of local correlation and 5/3 wavelet de-noising, and then the correlation of noise sequence is quantified as multi-order Markov feature vectors. Based on the extracted feature vectors, a steganography detector based on support vector machine (SVM) is trained to identify whether the wav audio file carries hidden data. Experimental results show that detection accuracy can reach more than 80% even when the embedding rate is only 10%, which is far better than other method.

Copyright

© 2013, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Jie Zhang
AU  - Feng Du
AU  - Songbin Li
PY  - 2013/03
DA  - 2013/03
TI  - Steganalysis of LSB Matching in WAV Audio
BT  - Proceedings of the 2nd International Conference on Computer Science and Electronics Engineering (ICCSEE 2013)
PB  - Atlantis Press
SP  - 1027
EP  - 1031
SN  - 1951-6851
UR  - https://doi.org/10.2991/iccsee.2013.258
DO  - 10.2991/iccsee.2013.258
ID  - Zhang2013/03
ER  -