Efficient And Improved Video Steganography using DCT and Neural Network

IJSRD - International Journal for Scientific Research & Development| Vol. 3, Issue 10, 2015 | ISSN (online): 2321-0613
All rights reserved by www.ijsrd.com 432
Efficient and Improved Video Steganography using DCT and Neural
Network
Fozia R.Khan1
Prof. Sujata Anandwani2
1
P.G. Student 2
Assistant Professor
1,2
Department of Computer Science & Engineering
1,2
Noble Group of Institution, Junagadh, Gujarat, India
Abstract— As per the demand of modern communication it
is important to establish secret communication which is
obtain by seganography .Video Steganography is the
technique of hiding some covert message inside a video.
The addition of this information to the video is not
recognizable through the human eye as modify of a pixel
color is negligible. In the proposed method Discrete Cosine
Transform (DCT) and neural network is used. Input image is
divided into blocks and is processed to generate quantization
matrix of cover and stego images by using Discrete Cosine
Transform (DCT).And using neural network performance of
this method can be further improved. The neural network is
trained and on the basis of training and segmentation done,
neural network provide efficient positions where data can be
merge. The performance and efficiency is measured by
PSNR and MSE value.
Key words: Video Steganography, least significant bit,
discrete cosine transform, neural network
I. INTRODUCTION
Steganography is a technique that enables party to transmit
data or message to another without the communication
being perceptible to others. The message is embedded in
cover media in a manner that only the sender and intended
receiver have knowledge of the existence of the message,
and the method to retrieve it. Steganography involves hiding
the contents inside a file and not scrambling the data, so it is
structurally unmodified and intact. Thus, Steganography has
an advantage over cryptography as it involves both
encryption and obscurity. Image, text, audio can be the
cover media. Data in the form of text, audio and video can
be embedded in the carrier. The most commonly used
carrier is image. To transmit much higher amount of secret
data, a video can be used instead [1].
Fig. 1: Segnography[1]
A. Video Steganography
Video Steganography is the art of hiding information in
ways that avert the revealing of hiding messages in videos.
Actually message like text, image, audio, video and etc. It is
focused on spatial and transform domain. Spatial domain
algorithm directly embedded information in the cover image
with no visual changes with good quality. The result of
algorithms has the advantage in Steganography capacity.
Transform domain algorithm is embedding the secret
information in the transform space. This kind of algorithms
has the advantage of good stability, but the disadvantage of
small capacity [2].
B. Techniques Used
1) Least Significant Bit
Least significant bit (LSB) insertion is the most widely used
technique for image embedding. This method became very
popular due to its easy implementation. It embeds data in a
cover image by replacing the least significant bits (LSB) of
cover image with most significant bits (MSB) of message
image.
The least significant bit (in other words, the 8th bit) of some
or all of the bytes inside an image is changed to a bit of the
secret message.
Pixel: (01101001 11010100 11010001)
(11001000 01011100 11101001)
(00100111 11001001 11101001)
From the above grid the LSB of each byte represents the
red, green, blue color, suppose to embed numeric value „15‟
(00001111), the matrix will be modified as,
(01101000 11010100 11010000)
(11001000 01011101 11101001)
(00100111 11001001 11101001)
The above matrix shows that it needs only 3 bits to be
modified to embed numeric value „15‟ successfully. Since
the resulting changes are too small, it is difficult for the
human eye to recognize the changes [3].
2) Discrete Cosine Transform
The DCT transforms a signal from an image representation
into a frequency representation, by grouping the pixels into
8 × 8 pixel blocks. These mathematical transforms convert
the pixels in such a way as to give the effect of “spreading”
the location of the pixel values over part of the image. The
redundant bits selected to embed the hidden data are taken
from the least-significant bits of the quantized DCT
coefficients. Thus the smoothening of the pixel alteration is
virtually impossible for human visual detection [3].
Fig. 2: Discrete Cosines Transform [4]

Efficient and Improved Video Steganography using DCT and Neural Network
(IJSRD/Vol. 3/Issue 10/2015/087)
The general equation for a 1D (N data items) DCT
is defined by the following equation:
The general equation for a 2D (N by M image)
DCT is defined by the following equation:
Here, the input image is of size N X M. c(i, j) is the
intensity of the pixel in row i and column j; C(u,v) is the
DCT coefficient in row u and column v of the DCT matrix.
Signal energy lies at low frequency in image; it appears in
the upper left corner of the DCT. Compression can be
achieved since the lower right values represent higher
frequencies, and generally small enough to be neglected
with little visible distortion. DCT is used in steganography
as- Image is broken into 8×8 blocks of pixels. Working from
left to right, top to bottom, the DCT is applied to each block.
Each block is compressed through quantization table to
scale the DCT coefficients and message is embedded in
DCT coefficients [4].
C. Neural Network
Neural network mimics some features of a real nervous
system that contains a collection of basic computing units
called neurons. These are the basic signalling units of the
nervous system. Each neuron is a discrete cell whose several
processes arise from its cell body. These neurons were
represented as models of biological networks into
conceptual components for circuits that could perform
computational tasks. The basic model of the neuron is
founded upon the functionality of a biological neuron[3].
1) Feed Forward Neural Network
A feed forward network provides input to the next layer
with no closed chain of dependence among neural states
through a set of connection strengths or weights. Feed-
forward neural network allow signals to travel in one way
only; from input to output. There is no feedback (loops) i.e.
the output of any layer does not affect that same layer [3].
2) Back Propagation Neural Network
It requires a dataset of the desired output for many inputs,
making up the training set. The term is an abbreviation for
"Backward propagation of errors" [3].
3) X-Or Propagation Network
In XOR Neural Network:
If the total unit is positive switch ON the unit if the total
input is negative switch off unit. So activation function will
be 1 if the total unit is positive and activation function will
be 0 if the total unit is negative. The network consists of
three-layer. There are two input units; first input unit is in
first layer, second unit which is in the hidden layer and last
one is output unit. The connection weights are shows on
links and every unit is shown inside the unit as threshold.
The hidden unit is concerned, the hidden unit is no different
from both of the input units, and simply provides another
input [5].
Fig. 3: Neural Network [5]
In its most general form, a neural network can be viewed as
comprising the following eight components:
a) Neurons
Neurons can be of three types: input, hidden and output.
Input neurons receive the external stimuli presented to the
network. Hidden neurons compute intermediate functions
and their states are not accessible to the external
environment. Outputs from the network are generated as
signals of output neurons [8].
b) Activation state vector
The activation level of individual neurons indicated by
this vector in neural network.
c) Signal function
The output signal of the Neuron based on its activation is
called a signal function is generates by this function.
Functions may differ from neuron to neuron within the
network; although most networks are field homogeneous i.e.
all neurons within a field or layer have the same signal
function.
d) Pattern of connectivity
This shows the inter-neuron connection architecture or the
graph of the network.
e) Activity aggregation rule
This aggregated the activity at a particular neuron.
f) Activation rule
The new activation level of a neuron based on its current
activation and its external inputs is determined by this
function.
g) Learning rule
With the aim of improving the network performance.The
learning rule provides a means of modifying correction
strengths based on both the external stimuli and the network
performance.
4) Advantages Include:
1. Adaptive learning: An ability to learn how to do tasks
based on the data given for training or initial experience.
2. Self-Organization: An ANN can create its own
organisation or representation of the information it receives
during learning time.
3. Real Time Operation: ANN computations may be carried
out in parallel, and special hardware devices are being
designed and manufactured which take advantage of this
capability.
4. Fault Tolerance via Redundant Information Coding:
Partial destruction of a network leads to the corresponding
degradation of performance. However, some network
capabilities may be retained even with major network
damage [7].
D. Computational Parameters Used
1) PSNR
Peak signal-to-noise ratio, often condensed PSNR, is an
engineering term for the ratio between the maximum
possible power of a signal and the power of undignified
noise that affects the reliability of its representation.
The PSNR of the steganography result defined as follows:
Where MAX= Maximum Possible Pixel Value of
an image. Higher the value of the PSNR, better the
performance of the steganography algorithm. High PSNR
value indicates high security because it indicates minimum
difference between the original and stego values. So no one
can suspect the hidden information.

Efficient and Improved Video Steganography using DCT and Neural Network
(IJSRD/Vol. 3/Issue 10/2015/087)
2) MSE
The mean squared error measures the average of the square
of the error. The error is the amount by which the estimator
differs from the quantity to be estimated.
The mean-squared error (MSE) is given by:
Where m,n- represents the size of image[6].
II. PROPOSED SYSTEM
The proposed method utilizes the discrete cosine transform
and neural network. To generate the quantization matrix the
color image i.e. used as cover image is processed by using
the discrete cosine transform and then neural network
provide efficient positions to merge data by training the
neural network and on the basis of training and
segmentation done.
1) Select video on which steganography to be perform.
2) Select hidden message.
3) Fragment video sequence into frames.
4) Apply DCT to the selected image.
5) For the purpose of the image compression generate
quantization matrix After DCT application.
6) Then values will be saved in table of quantization.
7) Now pass the secret text and the quantized image to the
Neural network .
8) If image size is optimum then select image to be stego
image.
9) Reform the video.
Fig. 1: Block diagram of Proposed System
III. ADVANTAGES OF PROPOSED SYSTEM
The Advantages of the proposed stego machine are a very
usable and good looking wizard based GUI (Graphical User
Interface) for the system Ability to operate the system with
no prior training and consultation of any help files
– Ability to conceal and reveal the exact hidden data
from video file without disturbing the running
application or new application
– Ability to encrypt and decrypt the data with the images
– With this system, an image, after hiding the data, will
not degrade in quality.
IV. CONCLUSION
Neural networks have the capability to develop implication
from complicated or loose data; it may be used to extract
patterns. Neural networks can identify trends that are too
assorted to be noticed by either humans or other computer
techniques. A well trained artificial neural network can be
thought of as an "expert" in the classification of information
this has been specific to evaluate. This type of neural
network may then used to provide projections given new
circumstances of attention and answer "what if"
Questionery. The performance of Steganography technique
is enhancing by the use of artificial neural network. This
makes our application flexible and can be extended to any
field of interest. Incorporated with the other fields like
Artificial intelligence, unclear logic neural networks have a
huge potential to perform. Neural networks have been
applied in solving a wide variety of tribulations. It is an
emerging and fast growing field and there is a huge scope
for research and development.
REFERENCES
[1] Mrudul Dixit, Nikita Bhide ,Sanika Khankhoje
,Rajashwini Ukarande “ Video Steganography” 2015
IEEE .
[2] K.Parvathi Divya , K.Mahesh “Various Techniques in
Video Steganography -A Review” International Journal
of Computer & Organization Trends – Volume 5 –
February 2014.
[3] Lovepreet Kaur , Geetanjali Babbar “Improved
Protection in Image Steganography using Neural
Network and Discrete Cosine Transform” International
Journal of Application or Innovation in Engineering &
Volume 3, Issue 11, November 2014
[4] Dr. Ekta Walia , Payal Jain, Navdeep “An Analysis of
LSB & DCT based Steganography” Global Journal of
Computer Science and Technology Vol. 10 Issue 1 (Ver
1.0), April 2010.
[5] Richa Khare, Rachana Mishra, Indrabhan Arya, “Video
Steganography by LSB Technique using Neural
Network” Sixth International Conference on
Computational Intelligence and Communication
Networks, 2014 IEEE
[6] Heena Goyal, Preeti Bansal,” VIDEO
STEGANOGRAPHY USING NEURAL NETWORK
AND GENETIC ALGORITHM”,IJETIC Volume 1,
Issue 9, September 2015.
[7] Seema Rani1, Nitika Kapoor2, Harish Kundra3 “An
Overview: Enhanced Steganography Using Rule Base
Neural Network” International Journal of Computer
Science and Communication Engineering Volume 3
issue 1(February 2014 issue)
[8] Shalini Choubey, Dr.Ashish Bansal,” Video
Steganography Using Neural Network Methods”
International Journal of Research in Advent
Technology, Vol.2, No.2, February 2014.

Efficient And Improved Video Steganography using DCT and Neural Network

More Related Content

What's hot (17)

Similar to Efficient And Improved Video Steganography using DCT and Neural Network (20)

More from IJSRD (20)

Recently uploaded (20)

Efficient And Improved Video Steganography using DCT and Neural Network