STEGNOGRAPHY AND STEGANALYSIS

 

INTRODUCTION

n STEGANOGRAPHY

   The art and science of hiding information in a cover document such as digital images in a way that conceals the existence of hidden data.

STEGANALYSIS

n  The art and science of detecting hidden data, determining the length of the message, and extracting the data.

n  Why is it important?

nPrevent Terrorist Attacks

nCatch people engaging in illegal activities

nDiscourage Piracy

STEGANOGRAPHY Vs CRYPTOGRAPHY

n CRYPTOGRAPHY

    The science of using mathematics to encrypt and decrypt data. It enables you to send sensitive information across insecure networks (such as the Internet).

 

n STEGANOGRAPHY

    Hiding a secret message within a larger object in such a way that others can not discern the presence of the hidden message.

OBJECTIVE OF THE TRAINING

n Performance of the spread spectrum steganography for a set of cover images and determine the best image out of them which will be detected with best efficiency.

 

WORKING PRINCIPLES

n The system uses Artificial Neural Network (ANN) as classifier.

n The system uses probabilistic Neural Network for classification.

n After getting the cover images the image is converted into set of Markov chain, data divergence will be predicted.

n  Data divergence of stego image is taken by Markov chain.

n The efficiency of system is predicted by comparing the cover image and the stego image.

STEGANOGRAPHIC SYSTEM

GRAPHICAL VERSION OF STEGANOGRAPHIC SYSTEM

THE RGB COLOR MODEL

DESIGN AND ALGORITHM

TRANSMITTED STEGANOGRAPHIC DATA

BLOCK DIAGRAM FOR DIVERGENCE CALCULATION

BLOCK DIAGRAM OF THE SYSTEM

ROLE IN TRAINING

n Learnt .Net 2.0

n Learnt MATLAB 7.0

n Developed simple steganography model

n Performed steganalysis test

 

FEATURES OF .NET & MATLAB

n .Net

nFast image processing capability

nBoth byte and bit level data handling

nObject oriented aspect

n MATLAB

nReady function for DCT

nCo occurrence matrix (graycooc)

 

INTERPRETATION

CONCLUSION

n  The Markov chain model gives the better performance for the steganographer as well as steganlyst. This benchmark provides the platform for steganalyst to choose the best algorithm for the measure of security. This model is general, flexible and used to evaluate dependencies.

 

n  The only limitation of this work is the model does not support when the level of dependencies are increased. To evaluate these inter pixel dependencies of pixels the order of Markov has to be increased. Hence second order Markov model is needed to overcome this drawback. In future it is required to investigate expansion of the proposed work to increase the level of dependencies of pixels.