Sie sind auf Seite 1von 29

NDMVPSs KBTCOE, Nashik

Dept. of Computer Engineering


Use of Hybrid Fuzzy Neural Network
for Advanced Steganography
Guided By: Prof. V. S. Tidake
Group No.

1, Amit Aher
2, Akshay Bakliwal
3, Pandurang Sadgir
4, Suraj Zankar

Abstract
In today's world steganography is very important due to
the confidential communication between computer users
over the internet. Steganography is known as the study
of invisible i.e. hidden communication that usually
carried
out
by
hiding
the
existence
of
the
communication. Generally data embedding is achieved in
communication, image, text, voice or any multimedia
content
for
military
communication,
copyright,
authentication and many other government sectors and
general
purpose.
In
Steganography,
secret
communication is done by embedding a message into
cover media a.k.a. carrier and generate a stego-image
i.e. output image which consist a hidden secret message.
The quality degradation of resultant stego-image is
handled by an efficient pixel adjustment method with the
use of fuzzy neural networks.
2

Use of Hybrid Fuzzy Neural Network for Advanced Steganography


KBT COE, Nashik

Literature Survey

Use of Hybrid Fuzzy Neural Network for Advanced Steganography


KBT COE, Nashik

Problem Definition
To

implement an Hybrid Fuzzy


Neural
Network
(HFNN)
for
advanced
steganography
for
embedding secret data using Least
Significant Bit (LSB) algorithm to
maintain
confidentiality
of
information
sharing
over
the
network.

Use of Hybrid Fuzzy Neural Network for Advanced Steganography


KBT COE, Nashik

Purpose and Scope


Purpose
Steganography though is still a fairly new

idea, there are constant advancement in


the computer field
Likely there will soon be more efficient and

more advanced technique for steganography


In the future it is hoped that the technique of

steganography will advanced such that it


become much useful
5

Use of Hybrid Fuzzy Neural Network for Advanced


Steganography
KBT COE, Nashik

Purpose and Scope


Scope
To implement secure steganography system
To keep the integrity of message of Stego-

Image using Hybrid Fuzzy Neural Network for


maintaining integrity of information.

Use of Hybrid Fuzzy Neural Network for Advanced


Steganography
KBT COE, Nashik

Minimum H/W and S/W


requirements
H/W
Intel Pentium processor 1.5GHz or above
RAM at least 1GB
Hard disk at least 1GB

S/W
Any web browser that supports JavaScript
Any mobile phone having browser

Use of Hybrid Fuzzy Neural Network for Advanced


Steganography
KBT COE, Nashik

System architecture

Chi Square
Probability
&
Euclidian
norm

1. DATA EMBEDDING PHASE (Sender)


Bits that hold
Secret Data

Cover
Image
LSB
Algorithm
For
Embedding
AES

Stego
Key
Features
of StegoImage

Free Bits

HFNN : Hybrid Fuzzy Neural Network


BPL : Back Propagation Learning

H
F
N
N

StegoImage

Secret Data
Encryption

Modified
Free Bits

Use of Hybrid Fuzzy Neural Network for Advanced


Steganography
KBT COE, Nashik

Features
of
Cover
Image

New
Features
of Stego

BPL

No

Is
Matche
d
?
Yes

Assemble bits that hold


Secret Data with
Modified free bIts

Improved
StegoImage

System architecture

(contd.)

1. DATA EMBEDDING PHASE (Sender)

Cover image and Secret data is

provided
as
input
Embedding phase

Use of Hybrid Fuzzy Neural Network for Advanced


Steganography
KBT COE, Nashik

to

LSB

System architecture

(contd.)

1. DATA EMBEDDING PHASE (Sender)

LSB (Least Significant Bit) substitution


The process of adjusting the least significant

bit pixels of the carrier image.


Simple approach for embedding message into

the image.
The Least Significant Bit insertion varies

according to number of bits in an image.

10

Use of Hybrid Fuzzy Neural Network for Advanced


Steganography
KBT COE, Nashik

LSB ALGORITHM STEPS


If the LSB of the pixel value of cover image C(i,j) is equal

to the message bit m of secret massage to be


embedded, C(i,j) remain unchanged; if not, set the LSB
of C(i, j) to m.
The message embedding procedure is given below
S(i,j) = C(i,j) - 1, if LSB(C(i,j)) = 1 and m = 0
S(i.j) = C(i,j), if LSB(C(i,j)) = m
S(i,j) = C(i,j) + 1, if LSB(C(i,j)) = 0 and m = 1
Where, LSB(C(i, j)) stands for the LSB of cover image

11

C(i,j)
m is the next message bit to be embedded.

project
titleis the stego image .
S(i,j)
KBT COE, Nashik

LSB Encryption
Input: Cover Image, Secret message
Output: Stego Image
PROCEDURE:
Step 1: Extract pixels of cover message.
Step 2: Extract the characters of the text file. Convert that

ASCII characters into binary string.


Step 3: Insert characters of text file in each least significant
bits of cover pixels by replacing it.
Step 4: Repeat step 3 till all the characters has been
embedded.
Step 5: Obtained stego-image

12

project title
KBT COE, Nashik

System architecture

(contd.)

1. DATA EMBEDDING PHASE (Sender)

Cover image and Secret data is provided as

input to LSB Embedding phase


Embedded data is encrypted using AES
Stego-key is used for encryption

13

Use of Hybrid Fuzzy Neural Network for Advanced


Steganography
KBT COE, Nashik

System architecture

(contd.)

1. DATA EMBEDDING PHASE (Sender)


Feature extraction of image
Stego-image is transformed into set of features

called feature extraction.


Chi square probability

is extracted as statistical
feature which is based on expected and observed
frequencies of the image pixels
The formula for chi square probability is given by

Where
o=observed frequency
14

Use of Hybrid
Fuzzy Neural
Network for Advanced
e=expected
frequency
Steganography
KBT COE, Nashik

System architecture

(contd.)

1. DATA EMBEDDING PHASE (Sender)

Euclidian

norm is extracted as visual


feature which is exhibit various forms of
membership functions.

Same features are extracted from cover

image.

15

Use of Hybrid Fuzzy Neural Network for Advanced


Steganography
KBT COE, Nashik

System architecture

(contd.)

1. DATA EMBEDDING PHASE (Sender)


Image features are provided to HFNN
HFNN is used to improve the quality of stego-image
HFNN

gives new features of Stego-image are


compared with Cover-image features if they are
matched then modified free bits and bits that holds
secret data are assembled which forms new
improved stego-image. Otherwise the generated
features are again provided to HFNN with Back
Propagation Learning (BPL) algorithm

16

Use of Hybrid Fuzzy Neural Network for Advanced


Steganography
KBT COE, Nashik

System architecture

(contd.)

1. DATA EMBEDDING PHASE (Sender)


Back Propagation Learning Algorithm
1. Implement data embedding algorithm to embed the
2.
3.
4.

5.

17

secret data and produce the stego-image.


Feature extraction
Creating the buffer containing the free bits.
Use a HFNN with the input layer containing free bits
buffer and the stego-image, the output layer containing
the new free bits buffer and updated stego-image.
Compare the output of NN with cover image. If matched,
the new modified free bit output in buffer is used to form
new stego-image by assembling it with the other bits
where the secret data is hidden. Otherwise goto step 4

Use of Hybrid Fuzzy Neural Network for Advanced


Steganography
KBT COE, Nashik

System architecture

(CONTD.)

2. DATA EXTRACTION PHASE (Receiver)


Improved
Stego-Image

AES
Decryption

Stego-Key

18

Use of Hybrid Fuzzy Neural Network for Advanced


Steganography
KBT COE, Nashik

Secret Data
Cover
Image

System architecture

(CONTD.)

2. DATA EXTRACTION PHASE (Receiver)

Improved Stego-image is decrypted using

AES decryption algorithm


Gives secret data and cover
Secret data is in form of bit stream and later

it is converted into an appropriate data like


image, audio and video
Finally we extracted the secret data
19

Use of Hybrid Fuzzy Neural Network for Advanced


Steganography
KBT COE, Nashik

System architecture

(CONTD.)

2. DATA EXTRACTION PHASE (Receiver)


In LSB retrieval embedded data is extract.
Here encrypted text message is extracted from

Least significant Bits of cover image.


The extraction process is as follows:
Inputs: Stego-image file.
Output: Secret text message.
Procedure:
Step 1: Extract the pixels of the stego image.
Step 2: Extract secret message characters from least

significant bit component of cover message. Follow


Step 2 till up to terminating.
Step 3: Extract secret message.
20

project title
KBT COE, Nashik

System architecture

(CONTD.)

For example, suppose one can hide a message in three pixels of


an image (24-bit colors). Suppose the original 3 pixels are:
(11101010 11101000 11001011)
(01100110 11001010 11101000)
(11001001 00100101 11101001)
A steganographic program could hide the letter J which has a
position 74 into
ASCII character set and have a binary representation
01001010, by altering the channel bits of pixels.
(11101010 11101001 11001010)
(01100110 11001011 11101000)
(11001001 00100100 11101001)
21

project title
KBT COE, Nashik

System architecture

(CONTD.)

2. DATA EXTRACTION PHASE (Receiver)

Cover image and Secret data is provided as

input to LSB Embedding phase


Embedded data is encrypted using AES
Stego-key is used for encryption

22

Use of Hybrid Fuzzy Neural Network for Advanced


Steganography
KBT COE, Nashik

Advantages
Payload capacity
High imperceptibility
Better security
Robustness
High quality stego-image
23

Use of Hybrid Fuzzy Neural Network for Advanced


Steganography
KBT COE, Nashik

Disadvantages
May use for illegitimate purpose like

terrorism
Secret key leakage may occur and it leads

to unauthorized access of data


The confidentiality of information is

maintained by algorithms, and if the


algorithms are known then this technique is
of no use
Use of Hybrid Fuzzy Neural Network for Advanced
24

Steganography
KBT COE, Nashik

Limitations
Supports JPG, BMP, GIF image formats
Requires knowledge of computer programs
No quality improvement blur images
Accept only one input at the time

25

Use of Hybrid Fuzzy Neural Network for Advanced


Steganography
KBT COE, Nashik

Applications
Government sectors
Copyright
Authentication
Trade secrets

26

Use of Hybrid Fuzzy Neural Network for Advanced


Steganography
KBT COE, Nashik

Conclusion
Steganography is fascinating and effective
method of hiding data that has been used
throughout history. This technology is easy to
use and difficult to detect.
This system provides an overview of
steganography cover media types, specific
techniques. Where standard cryptography
and encryption is outlawed, steganography
can be used. Formerly, just an interest of the
government
sectors
is
now
gaining
popularities among the masses.
27

Use of Hybrid Fuzzy Neural Network for Advanced


Steganography
KBT COE, Nashik

References
[1] Saleema A., Dr. T. Amarunnishad, A New Steganography Algorithm Using Hybrid
Fuzzy Neural Networks, International Conference on Emerging Trends in
Engineering, Science and Technology (ICETEST), ScienceDirect, 2016.
[2] Akhtar, N., Johri, P., Khan, S., Enhancing the Security and Quality of LSB Based
Image Steganography, Computational Intelligence and Communication Networks
(CICN), 2013 5th International Conference on, vol., no., pp.385, 390, 27-29 Sept.
2013.
[3] Vipula Madhukar Wajgade , Dr. Suresh Kumar, SGT Institute of Technology And
Management, Gurgaon, Haryana, India, Enhancing Data Security Using Video
Steganography, International Journal of Emerging Technology and Advanced
Engineering, Volume 3, Issue 4, April 2013.
[4] Christy.A.Stanley, Fachinger, J., Pairs of Values and Chi Square Attack,
Department of Mathematics, Iowa State University, Behavior of HTR Fuel Elements
in Aquatic Phases of Repository Host Rock Formations. Nuclear Engineering &
Design 236, 54, May 1, 2005
[5] Westfeld, A., Pfitzmann, A., Attacks on Steganographic Systems, 3rd
International Workshop, Lecture Notes in Computer Science, 2000.
vol. 1768. Springer-Verlag, Berlin, Heidelberg, New York.

28

[6] Cvejic, Nedeljko, Algorithms for audio watermarking and Steganography,


Department of Electrical and Information Engineering, Information Processing
Use of Hybrid Fuzzy Neural Network for Advanced
Laboratory, University of Oulu, Finland 2004.
Steganography
KBT COE, Nashik

References

(Contd.)

[7] Rafael C. Gonzalez, Richard E. Woods, Digital Image Processing, Third Edition.
[8] Petticolas, F.A.P. Anderson, R. J. Kuhn, M. G. , Information Hiding-A survey,
Proceedings of the IEEE, Special Issue on Identification and Protection of Multimedia
content 87, 1062-1078, 1999.
[9] El-Emam, N., Embedded a large amount of information using high secure neural
based steganography algorithm", International Journal of Information and
Communication Engineering 4 (2), 95106, 2008.
[10] S Sivanandam , S Sumathi, Introduction to neural networks using matlab 6.0.
[11] Jinsa K., and Gunavathi K., "Lung cancer classi_cation using neural networks for
CT images", Elsevier, computer methods and programs in biomedicine, pp. 202 209,
2014.
[12] Liu Li, Huo Liqing, Lu Hongru, Zhang Feng, Zheng Chongxun, Shami Pokhrel,
and Zhang Jie1, The Use of Fuzzy Back Propagation Neural Networks for the Early
Diagnosis of Hypoxic Ischemic Encephalopathy in Newborns, Hindawi Publishing
Corporation, Journal of Biomedicine and Biotechnology Volume 2011, Article ID
349490, 5 pages , doi:10.1155/2011/349490, 2011``

29

Use of Hybrid Fuzzy Neural Network for Advanced


Steganography
KBT COE, Nashik

Das könnte Ihnen auch gefallen