Title of article :
Active Steganalysis of Transform Domain Steganography Based on Sparse Component Analysis
Author/Authors :
Modaghegh، Hamed نويسنده Ferdowsi University of Mashhad Modaghegh, Hamed , Seyedin، Seyed Alireza نويسنده Ferdowsi University of Mashhad Seyedin, Seyed Alireza
Issue Information :
فصلنامه با شماره پیاپی 10 سال 2015
Pages :
10
From page :
75
To page :
84
Abstract :
This paper presents a new active steganalysis method to break the transform domain steganography. Most of steganalysis techniques focus on detecting the presence or absence of a secret message in a cover (passive steganalysis), but in some cases we need to extract or estimate a hidden message (active steganalysis). Despite the importance of estimating the message, little research has been conducted in this area. In this study, a new active steganalysis method based on Spars Component Analysis (SCA) technique is presented. Here, the sparsity property of the cover image and hidden message has been used to extract the hidden message from a stego image. In our method, the transform domain steganography is formulated mathematically as a linear combination of sparse sources. Thus, the active steganalysis can be presented as a SCA problem. The feasibility of the SCA problem solving is confirmed by Linear Programming methods. Then, a fast algorithm is proposed to decrease the computational cost of steganalysis and still maintains the accuracy. The accuracy of the proposed method has been confirmed in different experiments on a variety of transform domain steganography methods. According to these experiments, our method not only reduces the error rate, but also decreases the computational cost compared to the previous active steganalysis methods in the literature.
Journal title :
Journal of Information Systems and Telecommunication
Serial Year :
2015
Journal title :
Journal of Information Systems and Telecommunication
Record number :
2230160
Link To Document :
بازگشت