Statistical steganalyis of images using open source software

Bhargavi Kaipa, Stefan Robila

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

In this paper we present a novel steganalytic tool based on statistical pattern recognition. The main aim of our project was to design and implement a system able to classify the images into ones with no hidden message and steganographic images using classic pattern classification techniques such as Bayesian classification and decision trees. Experiments are conducted on a large data set of images to determine the classification algorithm that performs better by comparing classification success and error rates in each case. We have employed Weka, a data-mining tool developed in java for this purpose. We have also developed an application using Weka Java library for loading the data of the Images and classify the images into normal images and steganographic images. This application runs a GUI(Graphical User Interface) that enables the user to choose the classifier and other options required for the classification. Our results are aligned with current state of the art research and have the advantage of using open source software.

Original languageEnglish
Title of host publication2010 Long Island Systems, Applications and Technology Conference, LISAT 10
DOIs
StatePublished - 16 Jul 2010
Event2010 Long Island Systems, Applications and Technology Conference, LISAT 10 - Farmingdale, NY, United States
Duration: 7 May 20107 May 2010

Publication series

Name2010 Long Island Systems, Applications and Technology Conference, LISAT 10

Other

Other2010 Long Island Systems, Applications and Technology Conference, LISAT 10
CountryUnited States
CityFarmingdale, NY
Period7/05/107/05/10

Fingerprint

Pattern recognition
Graphical user interfaces
Decision trees
Data mining
Classifiers
Open source software
Experiments

Keywords

  • Image processing
  • Steganalysis
  • Steganography
  • Wavelets

Cite this

Kaipa, B., & Robila, S. (2010). Statistical steganalyis of images using open source software. In 2010 Long Island Systems, Applications and Technology Conference, LISAT 10 [5478333] (2010 Long Island Systems, Applications and Technology Conference, LISAT 10). https://doi.org/10.1109/LISAT.2010.5478333
Kaipa, Bhargavi ; Robila, Stefan. / Statistical steganalyis of images using open source software. 2010 Long Island Systems, Applications and Technology Conference, LISAT 10. 2010. (2010 Long Island Systems, Applications and Technology Conference, LISAT 10).
@inproceedings{c314be98b8824a80916d64b0bca86e8f,
title = "Statistical steganalyis of images using open source software",
abstract = "In this paper we present a novel steganalytic tool based on statistical pattern recognition. The main aim of our project was to design and implement a system able to classify the images into ones with no hidden message and steganographic images using classic pattern classification techniques such as Bayesian classification and decision trees. Experiments are conducted on a large data set of images to determine the classification algorithm that performs better by comparing classification success and error rates in each case. We have employed Weka, a data-mining tool developed in java for this purpose. We have also developed an application using Weka Java library for loading the data of the Images and classify the images into normal images and steganographic images. This application runs a GUI(Graphical User Interface) that enables the user to choose the classifier and other options required for the classification. Our results are aligned with current state of the art research and have the advantage of using open source software.",
keywords = "Image processing, Steganalysis, Steganography, Wavelets",
author = "Bhargavi Kaipa and Stefan Robila",
year = "2010",
month = "7",
day = "16",
doi = "10.1109/LISAT.2010.5478333",
language = "English",
isbn = "9781424455508",
series = "2010 Long Island Systems, Applications and Technology Conference, LISAT 10",
booktitle = "2010 Long Island Systems, Applications and Technology Conference, LISAT 10",

}

Kaipa, B & Robila, S 2010, Statistical steganalyis of images using open source software. in 2010 Long Island Systems, Applications and Technology Conference, LISAT 10., 5478333, 2010 Long Island Systems, Applications and Technology Conference, LISAT 10, 2010 Long Island Systems, Applications and Technology Conference, LISAT 10, Farmingdale, NY, United States, 7/05/10. https://doi.org/10.1109/LISAT.2010.5478333

Statistical steganalyis of images using open source software. / Kaipa, Bhargavi; Robila, Stefan.

2010 Long Island Systems, Applications and Technology Conference, LISAT 10. 2010. 5478333 (2010 Long Island Systems, Applications and Technology Conference, LISAT 10).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Statistical steganalyis of images using open source software

AU - Kaipa, Bhargavi

AU - Robila, Stefan

PY - 2010/7/16

Y1 - 2010/7/16

N2 - In this paper we present a novel steganalytic tool based on statistical pattern recognition. The main aim of our project was to design and implement a system able to classify the images into ones with no hidden message and steganographic images using classic pattern classification techniques such as Bayesian classification and decision trees. Experiments are conducted on a large data set of images to determine the classification algorithm that performs better by comparing classification success and error rates in each case. We have employed Weka, a data-mining tool developed in java for this purpose. We have also developed an application using Weka Java library for loading the data of the Images and classify the images into normal images and steganographic images. This application runs a GUI(Graphical User Interface) that enables the user to choose the classifier and other options required for the classification. Our results are aligned with current state of the art research and have the advantage of using open source software.

AB - In this paper we present a novel steganalytic tool based on statistical pattern recognition. The main aim of our project was to design and implement a system able to classify the images into ones with no hidden message and steganographic images using classic pattern classification techniques such as Bayesian classification and decision trees. Experiments are conducted on a large data set of images to determine the classification algorithm that performs better by comparing classification success and error rates in each case. We have employed Weka, a data-mining tool developed in java for this purpose. We have also developed an application using Weka Java library for loading the data of the Images and classify the images into normal images and steganographic images. This application runs a GUI(Graphical User Interface) that enables the user to choose the classifier and other options required for the classification. Our results are aligned with current state of the art research and have the advantage of using open source software.

KW - Image processing

KW - Steganalysis

KW - Steganography

KW - Wavelets

UR - http://www.scopus.com/inward/record.url?scp=77954496110&partnerID=8YFLogxK

U2 - 10.1109/LISAT.2010.5478333

DO - 10.1109/LISAT.2010.5478333

M3 - Conference contribution

AN - SCOPUS:77954496110

SN - 9781424455508

T3 - 2010 Long Island Systems, Applications and Technology Conference, LISAT 10

BT - 2010 Long Island Systems, Applications and Technology Conference, LISAT 10

ER -

Kaipa B, Robila S. Statistical steganalyis of images using open source software. In 2010 Long Island Systems, Applications and Technology Conference, LISAT 10. 2010. 5478333. (2010 Long Island Systems, Applications and Technology Conference, LISAT 10). https://doi.org/10.1109/LISAT.2010.5478333