scispace - formally typeset

Book ChapterDOI

A Multi-Algorithm, High Reliability Steganalyzer Based on Services Oriented Architecture

01 Jan 2013-pp 531-542

TL;DR: This chapter presents a reliable Steganalyzer system with distributed services oriented architecture which allows easy incorporation of new algorithms to support different media types in the existing system.

AbstractSteganalysis deals with detecting the presence of hidden information in different types of media such as images and audio files. Such detection is very challenging because of the variety of algorithms that might be used in embedding secret information in a media type. This chapter presents a reliable Steganalyzer system with distributed services oriented architecture which allows easy incorporation of new algorithms to support different media types in the existing system. Moreover, the distributed architecture presented in this chapter allows concurrent processing which speeds up the system. High system reliability in distinguishing between the cover object and the stego object is achieved by employing multiple steganalysis algorithms, and further by employing efficient feature classifiers based on neural networks. The system developed is versatile with capabilities of detecting stego objects in JPEG images as well as WAV audio files.

Topics: Steganalysis (58%), Steganography (50%), JPEG (50%)

Summary (1 min read)

Jump to:  and [Summary]

Summary

  • But further is capable of providing improved accuracy in stego detection through the use of multiple algorithms running in parallel.the authors.
  • The proposed system integrates different steganalysis techniques in a reliable Steganalyzer with distributed and Services Oriented Architecture (SOA).
  • The distributed architecture not only allows for concurrent processing to speed up the system, but also provides higher reliability than reported in the existing literature.
  • The extendable nature of the SOA implementation allows for easy addition of new Steganalysis algorithms to the system in terms of services.

Did you find this useful? Give us your feedback

...read more

Content maybe subject to copyright    Report

A Multi-Algorithm, High Reliability Steganalyzer
Based on Services Oriented Architecture
Eman Abdelfattah, Ausif Mahmood
Department of Computer Science and Engineering
University of Bridgeport, Bridgeport, CT
It is the art of discovering
invisible communication
Abstract
Results
In this prospectus we are proposing to
develop a unified Steganalyzer that can not
only work with different media types such as
images and audio, but further is capable of
providing improved accuracy in stego detection
through the use of multiple algorithms running
in parallel. Our proposed system integrates
different steganalysis techniques in a reliable
Steganalyzer with distributed and Services
Oriented Architecture (SOA). The distributed
architecture not only allows for concurrent
processing to speed up the system, but also
provides higher reliability than reported in the
existing literature. The extendable nature of the
SOA implementation allows for easy addition
of new Steganalysis algorithms to the system in
terms of services.
The universal steganalysis technique
proposed in this prospectus involves two
processes; feature extraction and feature
classification. Three methods are used for
feature extraction; Mel-Cepstrum and Markov
(for audio), and Intra-blocks for (JPEG
images). The feature classification process is
implemented using neural network classifier.
The unified steganalyzer is tested for JPEG
images and WAV audio files. The accuracy of
classification ranges from 96.8% to 99.8%
depending on the object type and the feature
extraction method. In particular, an
enhancement of Mel-Cepstrum technique is
proposed that achieves an accuracy of 99.8%.
This is significantly better than detection
accuracy of 89.9% to 98.6% [Liu 2011] where
even a much larger training dataset was used
than ours.
Classification confusion matrix for 1000 wav files
Extension
Service
Mel-Cepstrum Service
Complexity
Service
File
JPEG
< 0.08
>= 0.08
WAV
Decision
cover/stego
Extract 169
features
5-layers NN
5-layers NN
4-layers NN
Extract 58
features
Markov Service
Intra-Blocks Service
Extract 169
features
Overview of the Services Oriented Architecture Steganalyzer
Feature extraction:
A set of distinguishing statistics is
obtained from the object
Feature classification:
Training
Testing
Extension Service
Complexity Service
Mel-Cepstrum Service
Markov Service
Intra-Blocks Service
Markov transition probability of the
second order derivative (Stego wav)
Markov transition probability of the
second order derivative (Cover wave)
Feature
extraction
Classification
Feature
extraction
Training
Test files
cover/stego
Training files
cover/stego
Trained
network
Decision
cover/stego
Build using
Deployment
tool
Feature
Extraction
code
Trained
network
.dll
Intra-Blocks Service: 98.9% accuracy
Markov Service: 96.8% accuracy
Mel-Cepstrum Service: 99.8% accuracy
Conclusion
Steganalysis
Clean image Stego image
Universal Steganalysis
Services
Markov Services
Implemented a multi-algorithm, high
reliability steganalyzer based on Services
Oriented Architecture:
The implemented steganalyzer is scalable,
allowing the addition of new services to integrate
other steganalysis algorithms
The steganalyzer supports concurrent processing
which enhances its speed
Currently supports JPEG and WAV types
Universal (blind) steganalysis techniques are
implemented
A multi-stage neural network classifier is
designed for each technique that achieves
reliable results
# 10 : __ __ __
Citations
More filters

Journal ArticleDOI
TL;DR: This dissertation is proposing to develop a unified Steganalyzer that can not only work with different media types such as images and audio, but further is capable of providing improved accuracy in stego detection through the use of multiple algorithms.
Abstract: Network security has received increased attention in the last decades. Encryption has laid itself as the traditional method to transmit information in secrecy. Although strong encryption is a very secure approach for transmitting information, it can be easily identified that transmitted information is encrypted. Once the information is identified as encrypted, an intruder can block the encrypted transmission. In contrast, Steganography is a viable option to hide information in transmission without being identified. It provides a blanket that hides encrypted information. Thus, it becomes essential to develop mechanisms that reveal if the communicated information has any embedded data. Steganalysis is the art of detecting invisible communication and is a very challenging field due to different types of media and embedding techniques involved. Existing research in Steganalysis has focused on developing individual stego detection algorithms for a particular media type or for a particular embedding technique. In this dissertation we are proposing to develop a unified Steganalyzer that can not only work with different media types such as images and audio, but further is capable of providing improved accuracy in stego detection through the use of multiple algorithms. Our proposed system integrates different steganalysis techniques in a reliable Steganalyzer by using a Services Oriented Architecture (SOA). The SOA architecture not only allows for concurrent processing to speed up the system, but also provides higher reliability than those reported in the existing literature because multiple stego detection algorithms are incorporated simultaneously. Furthermore, the extendable nature of the SOA implementation allows for easy addition of new Steganalysis algorithms to the system in terms of services. The universal steganalysis technique proposed in this dissertation involves two processes; feature extraction and feature classification. An improved 2D Mel-Cepstrum implementation is used for wav files feature extraction. Intra-blocks technique is used for jpeg images feature extraction. The feature classification process is implemented using three different classifiers; neural network classifier, Support Vector Machines classifier, and AdaBoost classifier. The unified steganalyzer is tested for jpeg images and wav audio files. The accuracy of classification ranges from 90.0% to 99.9% depending on the object type and the feature extraction method. In particular, an enhancement of 2D Mel-Cepstrum implementation is introduced that achieves an accuracy of 99.9%. This is significantly better result than the average detection accuracy of 89.9% to 96.7% reported by Liu [1]. Finally, an extensible classifier is introduced that allows adding detection of new embedding techniques to the currently supported embedding techniques, so that the framework will maintain its reliability even if new embedding techniques are introduced.

2 citations


Cites background or methods from "A Multi-Algorithm, High Reliability..."

  • ...Furthermore, experimental results of improved 2-D Mel-Cepstrum outperform the experimental results of Markov technique [37, 40]....

    [...]

  • ...We have introduced a feature extraction technique based on Mel-frequency cepstrum in conjunction with second order derivative in [40]....

    [...]

  • ...The details of the architectural components are given in [40]....

    [...]


References
More filters

Book
01 Jan 1993

1,919 citations


BookDOI
01 Dec 2000
TL;DR: Intelligent Control Systems explores recent advances in the field from both the theoretical and the practical viewpoints and integrates intelligent control design methodologies to give designers a set of flexible, robust controllers and providestudents with a tool for solving the examples and exercises within the book.
Abstract: From the Publisher: In recent years, intelligent control has emerged as one of the most active and fruitful areas of research and development. Until now, however, there has been no comprehensive text that explores the subject with focus on the design and analysis of biological and industrial applications. Intelligent Control Systems Using Soft Computing Methodologies does all that and more. Beginning with an overview of intelligent control methodologies, the contributors present the fundamentals of neural networks, supervised and unsupervised learning, and recurrent networks. They address various implementation issues, then explore design and verification of neural networks for a variety of applications, including medicine, biology, digital signal processing, object recognition, computer networking, desalination technology, and oil refinery and chemical processes.The focus then shifts to fuzzy logic, with a review of the fundamental and theoretical aspects, discussion of implementation issues, and examples of applications, including control of autonomous underwater vehicles, navigation of space vehicles, image processing, robotics, and energy management systems. The book concludes with the integration of genetic algorithms into the paradigm of soft computing methodologies, including several more industrial examples, implementation issues, and open problems and open problems related to intelligent control technology.Suited as both a textbook and a reference, Intelligent Control Systems explores recent advances in the field from both the theoretical and the practical viewpoints. It also integrates intelligent control design methodologies to give designers a set of flexible, robust controllers and providestudents with a tool for solving the examples and exercises within the book.

243 citations


Book ChapterDOI
19 Sep 2005
TL;DR: A graph-theoretic approach to steganography based on the idea of exchanging rather than overwriting pixels is suggested and an algorithm based on this approach with support for several types of image and audio files is implemented.
Abstract: We suggest a graph-theoretic approach to steganography based on the idea of exchanging rather than overwriting pixels. We construct a graph from the cover data and the secret message. Pixels that need to be modified are represented as vertices and possible partners of an exchange are connected by edges. An embedding is constructed by solving the combinatorial problem of calculating a maximum cardinality matching. The secret message is then embedded by exchanging those samples given by the matched edges. This embedding preserves first-order statistics. Additionally, the visual changes can be minimized by introducing edge weights. We have implemented an algorithm based on this approach with support for several types of image and audio files and we have conducted computational studies to evaluate the performance of the algorithm.

142 citations


Proceedings ArticleDOI
01 Mar 2007
Abstract: Steganography and steganalysis in VoIP applications are important research topics as speech data is an appropriate cover to hide messages or comprehensive documents. In our paper we introduce a Mel-cepstrum based analysis known from speaker and speech recognition to perform a detection of embedded hidden messages. In particular we combine known and established audio steganalysis features with the features derived from Melcepstrum based analysis for an investigation on the improvement of the detection performance. Our main focus considers the application environment of VoIP-steganography scenarios. The evaluation of the enhanced feature space is performed for classical steganographic as well as for watermarking algorithms. With this strategy we show how general forensic approaches can detect information hiding techniques in the field of hidden communication as well as for DRM applications. For the later the detection of the presence of a potential watermark in a specific feature space can lead to new attacks or to a better design of the watermarking pattern. Following that the usefulness of Mel-cepstrum domain based features for detection is discussed in detail.

72 citations


Proceedings ArticleDOI
19 Oct 2009
TL;DR: This study shows that, in comparison with a recently proposed signal stream based Mel-cepstrum steganalysis, the proposed method prominently improves the detection performance, which is not only related to information-hiding ratio but also signal complexity.
Abstract: In this paper, we present a novel stream data mining for audio steganalysis, based on second order derivative of audio streams. We extract Mel-cepstrum coefficients and Markov transition features on the second order derivative, a support vector machine is applied to the features for discovery of the existence of covert message in digital audios. We also explore the relation between signal complexity and detection performance on digital audios, which has not been studied previously. Our study shows that, in comparison with a recently proposed signal stream based Mel-cepstrum steganalysis, our method prominently improves the detection performance, which is not only related to information-hiding ratio but also signal complexity. Generally speaking, signal stream based Mel-cepstrum audio steganalysis performs well in steganalysis of audios with low signal complexity; it does not work so well on audios with high signal complexity. Our stream mining approach for audio steganalysis gains significant advantage in each category of signal complexity - especially in audios with high signal complexity, and thus improves the state of the art in audio steganalysis.

36 citations