On Text Realization Image Steganography

Dr. Mohammed Nasser Hussain Al-Turfi
International Journal of Engineering (IJE), Volume (6) : Issue (1) : 2012 1
Text Realization Image Steganography
Dr. Mohammed Nasser Hussein Al-Turfi mohammed_alturfi@yahoo.com
Department of Computer and Software Engineering
Al-Mustansiriyah University
Baghdad-Iraq
Abstract
In this paper the steganography strategy is going to be implemented but in a different way
from a different scope since the important data will neither be hidden in an image nor
transferred through the communication channel inside an image, but on the contrary, a well
known image will be used exists on both sides of the channel and a text message contains
important data will be transmitted. With the suitable operations, we can re-mix and re-make
the source image.
MATLAB7 is the program where the algorithm implemented on it, where the algorithm shows
high ability for achieving the task to different type and size of images. Perfect reconstruction
was achieved on the receiving side. But the most interesting is that the algorithm that deals
with secured image transmission transmits no images at all.
Key Words: Steganography, Text Stream, Secured Data
1. INTRODUCTION
The art of hiding information inside an image is called image steganography, where the ability
of hiding information inside this image without noticeable distortion depends on the hiding
algorithm and this decides the degree of successfulness [1]. The degree of successfulness
may be specified by the amount of security and hardness that must be paid to extract features
and important data. [9]
Steganography is a Greek word of two syllables, "stego" means "cover" while "grafia" means
"writing" define it’s a "cover writing" [10], therefore text file is the first impression about the
meaning of the word. Hiding data in a text file is more difficult than hiding data in an image
because it might change the text meaning or format or both [11]. Therefore two ways where
followed either by not changing text meaning or by creating a rubbished, un-understandable,
symbolic file keeping the file format. [12,14]
Early works depends upon using the Least Significant Bits (LSB) of the cover image for
implementing data hiding since its effect is the minimum and the size effect is equi-probabile
to other bits. Then certain operations are implemented on the source data in order to increase
the rate of securing the data and then to be inserted inside the cover image and mostly in the
LSB since the distortion rate will rise if the data inserted inside other bits. [2, 3]
DSP and convolutional techniques are used now a day for implementing image
steganography to increase the rate of security and to reduce the rate of distortion but of
course all these complications consume time, efforts, and processing powers. [2, 4]
Using some special functions that randomize the way of choosing the position was one of the
alternatives where the processed data will be stored in. This will enlarge the plain of
searching inside the cover image and reduces the rate of distortion since there is no linearity
in the process especially if the image is RGB because at least this will triple the efforts. [3][5]
Some use the pixel-value differences in their researches inorder to increase the amount of
embedded information where an Optimal Pixel Adjustment Process (OPAP) is used to
enhance the stego-image quality. The confidential information can be extracted from the
stego-image without the assistance of the original image [4].

Others search for certain places inside the image like a 2*2 block of pixels with high contrast
where message bits can be embedded into these blocks. Mathematical functions like MOD-4
and a coding key is used to increase the amount of embedded data while maintaining the
data fidelity and the process is as easy as possible since the process applied in the data
hiding phase is the same as the one in the data extraction phase[10].
Some researches didn’t apply the process from the computer point of view but from the
communication point of view by applying Spread Spectrum Image Steganography (SSIS).
SSIS conceals a message of substantial length with in digital imagery while maintaining the
original image size and dynamic range. The hidden message can be recovered using the
appropriate keys without any knowledge of the original image [13].
In this paper a new algorithm is proposed, the algorithm chooses the best cover image size
exists on the system data base that matches the size of the source image. The difference
between the cover and the source image is evaluated and transformed to text, then the text
will be transferred as a text message, an E-Mail, or even as an SMS where this make the
change as small as possible, hence reduce the need for a sophisticated channel and is more
secured since the source image is not transferred at any way.
2. IMAGE VERSES TEXT
One of the most important principles that must be obeyed in image steganography is that the
cover image must be larger in size than the source image so that we can hide the small inside
the large where the larger cover the best secured data transmission and less distortion but
this will be in need for larger, faster, and more powerful channel which means more cost and
complex electronic circuitry, this will be insistence if the used images are colored. [4, 6]
Images may fall in different types depending upon their extensions, but in general they
contains pixels, their representation starts from 8-bits of brightness in Black & White images
(Gray Scale) where the brightness level may reaches up to 32-bits especially in true color
image applications like Photoshop's and rendering (as in 3-D Max).[5]
Texts are quite different, they may fall in different type of files but the text itself inside these
files is the same since the way of implementation (writing) is the same where here the ASCII
code is the bible book and the representation is fixed given by 7-bits only and the variety is
given by different re-arrangements but not by different types. [6, 7]
The important property that exists on text but not in images is that texts need no operational
function for decoding while for images, there must be an operational de-coding and de-
compression function depending upon the image extension "jpeg, tiff, gif, …etc" and the way
of de-compression and operation "played with fox viewer or on media player, …etc" where
this means that dealing with text is much easier than dealing with images. [6, 7]
In this paper we are going to use text to represent the difference between the source image
and a pre-notified image stored and indexed in the system data-base exists on both sides of
the channel where no noise will affect the images. The important useful data is delivered in
different type and may be in different way " for example if the application is not urgent or
channels are occupied or not secured then the difference may be sent in a message written
;where different types of watermarking and letter security algorithms may be applied; through
the post office or by the person himself". [8]
3. PROPOSED ALGORITHM
One of the most important features of the proposed algorithm in this paper is that the cover
image may be equal in size to the source image where data security and the hardness for
extracting the important data maintained. Second, the algorithm doesn’t transmit any image at
all which is very important feature because the image stands for millions of words and any
intrusion or system breaking may cause security alert.
Third, the easiness that stand for transporting the vital data from one end to the other by
transmitting through the secured channel, or by mobile equipments like cellular mobiles or

PDA's as an SMS, or by letters either written and entered by hands or written and entered
using OCR's.
Perfect reconstruction that appears at the receiving ends which is the fourth important
property, since the system can stand strong and fight noise well in an easy ways and not in
need for sophisticated equipments which rises the overall system cost.
4. ALGORITHM IMPLEMENTATION
The proposed algorithm may fall into two parts: - Forward (where the process applied at the
transmitting end. The difference between source and cover image is generated), and
Backward (Re-arrange the received text to extract the source image from the received text).
4.1 Forward Phase
The system must pass through the following steps to achieve data hiding where fig(1) shows
the flow chart of the process explained as given:-
a) Image Comparative: - The cover image must be selected from the system data-base of
images in such a way that the selected one must be as near as possible to the size of
the source image so that we can maintain the channel resources. An important note
must be taken in consideration is that it's not necessary they have the same extension
or properties since both images are going to be changed to matrices but it's preferred
that they are from the same color type (Both are RGB or Black & White).
b) Image Transforming: - In this step the algorithm will read the image using
imread(name) function and transform each image into a matrix and each matrix into a
vector where both vectors must be unified in length by adding zeros to the shorter one.
c) Difference Evaluation: -This step is performed by the ordinary mathematical
subtraction. The difference is stored in a difference vector; its length equal to the larger
one. An important tag is added for each operation performed which is very necessary
at the receiving end to hold perfect reconstruction. If the operation output is positive
then the tag of the operation is (0) and (1) if the output is negative.
d) Bit Traverse: - This part of the system is responsible for transforming the difference
vector (which consists of 8-bits of brightness) from decimal to binary values in order to
prepare these values to be changed into letters or texts (consists of 7-bits ASCII code).
e) Re-arranging: - This part of the algorithm is responsible for text generation where the
system will begin to take seven bits at a time and transfer them into a letter (because
the image pixels are represented in 8-bits) by applying the char(number) function. For
example char (65) is "A" while char (97) is "a".
f) File Creation:- The file to be transferred is created and arranged as the following:-
1- The file is created for an open and append operations while the extension of the text
file must be chosen (for example "txt", "doc" …etc).
2- The first part of contents of the file up to the first coma (,) represents the cover image
index number (for example "173," this means the used cover image is numbered 173).
3- From the first to the second coma the data represents the image size (for example the
image dimensions are 80*60*3).
4- From the second to the third coma the stream of (0's) and (1's) represents the status of
the output of the subtraction operations where the number of bits in this stream is equal
to the number of subtractions and is equal to the multiplication of the image dimensions
(80*60*3=14400).
5- From the third coma to the EOF (End Of File) the text represents the difference
between the pixels of the source and the cover images which is transformed to text.
At this point we can notify two things; the first is that the source image is not exchanged
between the ends in any way or any how because the algorithm can extract the source image
using the text file only and depending on the index value given which represents the cover
image index. Second is that the text file and the source image are nearly of the same size
which means no extra size is needed for the cover image as in the traditional steganography.
4.2 Backward Phase
At the receiving end the algorithm is in need for a text message contains the difference
between the source and the cover images as a text, no matter how the letter is received

whether this letter may be an E-Mail, Secured encrypted message through secured channel,
a letter entered by hand delivered by the post mail or even by a letter printed by a printer and
entered using OCR scanner program, the letter must pass through the following steps:-
a) When the file is received the algorithm will start to read the data until it reaches the first
coma (,) where this part represents the index number of the image used from the system
images data base. At this point the stored vector (which represents the specified indexed
image that is transformed from an image into a vector) is located and is ready to be used
for comparison process.
b) The data stored between the 1
st
& 2
nd
coma represents the dimensions of the source
image which is very important for image reassembling.
c) The data stored between the 2
nd
& 3
rd
coma (,) represents a stream of (0's) & (1's) which
specifies the status of the difference operation where the number of bits is equal to the
number of subtraction operations implemented and equal to the number of pixels in it.
d) Starting from the 3rd
coma (,) to the EOF the text will be transferred to start processing
which include:-
1- Reading the text and transfer it in a vector, letter by letter or symbol by symbol.
2- Apply the abs('letter') function to transfer the text in the vector to decimal values.
3- Transfer the decimal values in the text vector to binary values made of 7-bits.
4- Re-arrange the new train of bits into 8-bits values and convert it into new decimal values.
e) Applying the correspondence algebraic operations to re-create the source image from
the cover image where if the status bit is (0) means the subtraction output is positive and
the operation to be implemented now is subtraction to retrieve the original data. While if
the status bit is (1) then the output is negative and the operation to be implemented is
addition to retrieve original data.
f) The new decimal values obtained are arranged in a pre-defined matrix dimension where
the function imwrite(a,filename,fmt) writes the matrix "a" of dimensions similar to the
dimensions of the image to the file specified by "filename" in the format specified by
"fmt". Matrix "a" can be an M-by-N (grayscale image) or M-by-N-by-3 (color image) array.
If the format specified is "tiff" it can also accept an M-by-N-by-4 array containing color
data that uses the CMYK color space.
5. RESULTS
Fig (2) clarifies parts of the file that will appear. The 1
st
line starts with the value 173 which
represents the index code number of the cover image in the system image database that
leads to the vector represents that cover image. In this vector the cover image is stored in a
vector ready for subtraction to reduce the processing time and to increase the system speed.
Next value after the 1st
coma represents the source image dimension which is important to
prepare a matrix with these dimensions to store the results from the subtraction process
between the received values and the stored one in the recieving side and to create the status
vector, its size equal to the multiplication of the matrix dimension (as in our example
80*60*3=.14400) to store status bits.
After the 2
nd
coma the status bits will be stored in the status vector, each bit in a separate cell
to decide whether the operation to be implemented between the received and the stored
values are addition or subtraction.
The difference between the cover and the source image is stored as text after the 3
rd
coma.
Part of the reconstruction process is demonstrated in fig (3). The demonstration shows how
the real values are extracted from the received text applied on 20 characters as an example
for the process flow.
The reconstruction was perfect; distortion rate was 0%, the process very fast because the
algorithm is not complex especially at the receiving end where the cover images on both
sides are transferred into vectors stored on the system data base ready to implement the
subtraction process immediately.
From the brief shown in fig (2) it can be noticed that the status bits have high redundancy
where a compression technique may be applied to reduce size, while a randomization

function may be added to randomize the sequence of data for mixing both status bits with
difference texts to increase file security, on the same way data hiding or coding may be
applied to cover the index, image size, …etc.
FIGURE 1: the flow chart of the forward phase
Start
Read Source Image and Decide the Best
Fit Cover Image in Size and Type
Yes
Are the
vectors
finish
Yes
Transferring the source and the cover images into
Decimal Matrices representation
Evaluate the Difference between the Vectors Using
Subtraction mathematical Function
Implement Bit Traverse to Change the Difference
From Decimal to Binary and Adding Status Bit
Yes
Re-arrange the Difference Vector into 7-bits
Text using ASCII Code Representation
No
No
Is Text Representation
Finished
Saving File Information
Are the
Matrices
ready
Represent both images into vectors
in order to evaluate the difference vector
No
End
File Creation and Preparation
Applying file Extension & File Saving
Is Data
Written to File
Finished
YesNo

Fig(2-a):-Partfromthestartofthetextfile
Fig(2-b):-Partfromthemiddleofthetext
Fig(2-c):-Partfromtheendofthetextfile
FIGURE 2: Part of the contents of the text file from Different places
The first group is the
first number in the
column which is for
example 173 where this
number represents the
index of the image in
the system data base
The second group of
numbers in the text file
represents the source
image dimensions
which is in our
example that
represents the index
image 173 is
80by60by3
The status bits which
appears after the image
size in fig (2-a) and a
part on the start of fig
(2-b) where the number
of these status bits is
14400 produced from
the multiplication of the
image size 80*60*3
where (1) represents –
ve result while (0)
represents +ve results
produced from the
The text appears after
the status bits in fig (2-
b) and in fig (2-c)
represents the text
produced by changing
the absolute value
produced from the
mathematical
subtraction between the
source and indexed
image while the sign of
this operation is
represented by the
status bit therefore the

107
70
48
72
37
78
103
67
50
51
76
76
110
33
77
26
26
1 0 1 1 0 0 0 1 0 1 0 0 1 0 0 0
0 0 0 0
k F 0 H % N g C 2 3 L L n ! M }
Using char('letter') to get the ASCII
Code
Changing from Decimal to binary representation
1101011
1000110
0110000
1001000
0100101
1001110
1100111
1000011
0110010
0110011
1001100
1001100
1110110
0100001
1001101
0011010
0011010
Changing from binary to decimal
11010111
00011001
10000100
10000100
10110011
10110011
11000011
01100100
11001110
01100100
11001110
11001000
01100110
10011010
00110100
00100100
00100111
Re-arrange in 8-bit representation
Re-Arrange using matrix Dimension to
create source image
215
25
132
132
179
179
195
100
206
100
206
200
102
154
52
37
39
Received
Text
ASCI
I
Code
7-bit
represent
8-bit
represent
Decimal
represe
nt
Status
vector
40
25
21
21
26
26
42
53
53
53
53
47
51
0
0
0
39
Stored
Vector
-
+
-
-
+
+
+
-
+
-
+
+
-
+
+
+
+
255
0
153
153
153
153
153
153
153
153
153
153
153
154
52
37
0
Pixels
values
FIGURE 3: A demonstration for the process applied at the receiving end for 20 character

6. REFERENCES
[1] Juneja, M. --- Sandhu, P.S., “Designing of robust image steganography technique
based on LSB insertion and encryption” International Conference on Advances in
Recent Technologies in Communication and Computing Year: 2009 Pages: 302-305
Provider: IEEE Publisher: IEEE DOI: 10.1109/ARTCom.2009.228.
[2] Raja, K.B. --- Chowdary, C.R. --- Venugopal, K.R. --- Patnaik, L.M. “A secure image
steganography using LSB, DCT and compression technique on raw image”
Intelligent Sensing and Information Processing, 2005. ICISIP 2005. Third International
Conference on ISBN: 0780395883 Year: 2005 Pages: 171-176 Provider: IEEE
Publisher: IEEE.
[3] Neeta, D. --- Snehal, K. --- Jacobs, D., “Implementation of LSB steganography and its
evaluation for various bits”, 1st International Conference on Digital Information
Management ISBN: 142440682X Year: 2007 Pages: 173-178 Provider: IEEE
Publisher: IEEE DOI: 10.1109/ICDIM.2007.369349.
[4] Hanling Zhang --- Guangzhi Geng --- Caiqiong Xiong ,“Image steganography using
pixel – value differencing” , Second International Symposium on Electronic Commerce
and Security Year: 2009 Volume: 2 Pages: 109-112 Provider: IEEE Publisher: IEEE
DOI: 10.1109/ISECS.2009.139.
[5] Nasser Hamad Faculty of Information Technology, Arab American University, Palestine,
Hiding Text Information in a Digital Image Based on Entropy Function , The
International Arab Journal of Information Technology, Vol. 7, No. 2, April 2010
[6] Muhalim Mohamed, Mohamed Amin, Subariah Ibrahim, Mazleen Salleh, Mohd Rozi
Katmin. , “Information Hiding Using Steganography”, Department of Computer
System & Communication Faculty of Computer Science and Information system
University Technology Malaysia, 2003.
[7] Sukhpreet Kaur , Department of Computer Science and Engineering Baba Farid
College of Engineering and Technology Bathinda-India & Sumeet Kaur, Department
of Computer Engineering Yadavindra College of Engineering Punjabi ,University
Guru Kashi Campus Talwandi Sabo, A Novel Approach for Hiding Text Using Image
Steganography , (IJCSIS) International Journal of Computer Science and Information
Security ,Vol. 8, No. 7, Punjab, India October2010.
[8] H.B.Kekre, Archana Athawale, and Pallavi N.Halarnkar , Increased Capacity of
Information Hiding in LSB’s Method for Text and Image, International Journal of
Electrical and Electronics Engineering 2:4:2008
[9] K Suresh Babu*, K B Raja*, Kiran Kumar K*, Manjula Devi T H*, Venugopal K R*, L
M Patnaik** “Authentication of Secret Information in Image Steganography” ,
TENCON 2008 - 2008 IEEE Region 10 Conference Year: 2008 Pages: 1-6 Provider:
IEEE Publisher: IEEE DOI: 10.1109/TENCON.2008.4766581
[10] *Department of Computer Science and Engineering University Visvesvaraya College of
Engineering, Bangalore University, Bangalore 560 001
[11] ** Microprocessor Applications Laboratory, Indian Institute of Science, Bangalore
ksureshbabu uvce@rediff.com
[12] K. Pramitha1, Dr. L.Padma Suresh2, K.L.Shunmuganathan3 “ Image Steganography
Using MOD-4 Embedding Algorithm Based on Image Contrast” TENCON 2008 - 2008
IEEE Region 10 Conference Year: 2008 Pages: 1-6 Provider: IEEE Publisher: IEEE
DOI: 10.1109/TENCON.2008.4766581
[13] 1 Department of Electrical and Electronics Engineering, P.G. Scholar, NIU, Kumaracoil.
E-mail:pramitha2007@gmail.com

[14] 2 Department of Electrical and Electronics Engineering, HOD, NIU, Kumaracoil. E-mail:
suresh_lps@yahoo.co.in
[15] 3 Department of Computer Science Engineering, Professor, RMK Engineering
College. E-mail:kls_nathan@yahoo.com
[16] Deshpande Neeta, Kamalapur Snehal Computer Science Dept. K.K.Wagh Institute of
Engineering Education & Research, Nashik India , deshpande_neeta@yahoo.com,
kamalapur_snehal@yahoo.com Daisy Jacobs School of Information Technology
University of Pretoria, Pretoria 002 South Africa daisy.jacobs@up.ac.za,
“Implementation of LSB Steganography and Its Evaluation for Various Bits”, : 1st
International Conference on Digital Information Management ISBN: 142440682X Year:
2007 Pages: 173-178 Provider: IEEE Publisher: IEEE DOI:
10.1109/ICDIM.2007.369349
[17] Hossein Malekmohamadi and Shahrokh Ghaemmaghami Sharif University of
Technology, Tehran, Iran h_malekmohamadi@ee.sharif.ir, ghaemmag@sharif.edu
“Steganalysis of LSB Based Image Steganography Using Spatial and Frequency
Features”, IEEE International Conference on Multimedia and Expo ISSN: 19457871
Year: 2009 Pages: 1744-1747 Provider: IEEE Publisher: IEEE DOI:
10.1109/ICME.2009.5202858
[18] Lisa M. Marvel and Charles T. Retter U.S. Army Research Laboratory Aberdeen
Proving Ground, MD 21005 marvelQar1. Mil Charles G. Boncelet, Jr University of
Delaware Newark, DE 19716 boncelet @ee.udel.edu, “Hiding Information in Images *”
Image Processing, 1998. ICIP 98. Proceedings. 1998 International Conference on
ISBN: 08186882 Year: 1998 Volume: 2 Pages: 396-398 vol.2 Provider: IEEE Publisher:
IEEE Comput. Soc
[19] 14) Mohammad Shirali-Shahreza, Sajad Shirali- Shahreza. d Shirali-
Shahreza”Steganography in TeX Documents”, 3rd International Conference on
Intelligent System and Knowledge Engineering Year: 2008 Volume: 1 Pages: 1363-
1366 Provider: IEEE Publisher: IEEE DOI: 10.1109/ISKE.2008.4731144
[20] 1.Computer Science Department Sharif University of Technology Tehran, IRAN
[21] 2.Computer Engineering Department Sharif University of Technology Tehran, IRAN
shirali@cs.sharif.edu, shirali@ce.sharif.edu

On Text Realization Image Steganography

More Related Content

What's hot (14)

Viewers also liked (8)

Similar to On Text Realization Image Steganography (20)

Recently uploaded (20)

On Text Realization Image Steganography