You are on page 1of 5

NEW STEGNOGRAPHY Second method is steganography, where the

secret message is hidden in another message,


TECHNIQUE USING DCT image, audio or video.. The word
steganography is derived from the Greek
QUANTIZATION WITH RSA words stegos meaning cover and
ENCRYPTION ALGORITHM grafia meaning writing [5] defining it as
covered writing.

Dr. Shiv Kumar1 Shekhar Chander2,


1
Senior Professor, ACEIT Jaipur, 2 M.Tech Scholars,
ACEIT Jaipur

ABSTRACT

Steganography is the art and science of


hiding information within information. This
thesis introduces steganography and focus on
the Discrete Cosine Transform (DCT)
technique for hiding messages in an image.
The proposed method use DC Coefficient
method to hide the secret message. This
work is unique from all the existing
techniques that are using AC Coefficient of
DCT Table or LSB bit to hide message, this
thesis enhanced the security by using public
key RSA encryption algorithm, the SNR,
quality and capacity to hide data is higher in
proposed method

1. Introduction to Stegnography

Digital communication is essential part in


everyone life. Often it is desired that
communication should be secret Two
techniques are available to accomplish this
goal the Cryptography [1] and the
Steganography[2,3,4].

In Cryptography the sender encrypts the


message with a key and the encrypted
message is sent to receiver over an insecure
communication channel. On the other side,
the receiver decrypts the received encrypted
message with a key and extracts the original
message. The key for sender and receiver can
be same or different.
2. The Proposed Hiding System 7. Quantize DCT coefficients by multiplying
the same quantization tables that are
A more complex way of hiding a secret used in a compression stage to obtain
message inside an image comes with DCT coefficients.
the use and modifications of discrete
cosine transformations. Discrete cosine 8. Inversing DCT is applied in this step in
transformations (DCT) are used by the each block.
JPEG compression algorithm to
transform successive 8 x 8 pixel blocks 9. Reconstruct the image by combining all
of the image, into 64 DCT coefficients. the blocks.
In this work one coefficient is used
from each block (8 x 8) to hold the bit 10. Transform the image formula from
called DC coefficient in position (0,0),
by preparing the value of pixels in the 3. Message Hiding Stage
block until the DC coefficient becomes
odd or even dependent on the bit In this step, the text bits are embedded in the
which is wanted to be hidden. This cover-image. After inputting the text in the
work includes the following steps: system add # at end of text message, convert
each letter in the text to a decimal number
1. First step begin by Loading a colour image
and encrypt it using RSA encryption
(bitmap format 24 bits), and break up
the colours into the red, green, and algorithm and convert each one to a binary
blue. form. Some steps are implemented to embed
the text bits. These steps are:
2. Convert the image formula from RGB to
YCbCr. 1. From each block (8*8) one DCT
coefficient is chosen to hold the bit. This
3. Separate the image components Y, Cb, Cr coefficient is in position (0, 0). Compute the
into blocks; each one of 64 pixels quantize DC coefficient (C1).
(8*8).
C1 = round (block (0, 0) /16 (1.1)
4. Transform each block (8*8) pixels into
spatial frequency domain using the 2. If (C1) is an odd number and the bit ('1'),
forward DCT. or (C1) is an even number and bit ('0'), no
change happens in the original block pixels.
5. Then embed the stream of bits (encrypted Bring a new block and a new bit to continue
message) into the cover image and in
or work and hide them.
each block embed one bit in the DC
element. 3. If (C1) is an odd number and the bit ('0'),
or (C1) is an even number and bit ('1'), there
6. After that quantize the blocks with
quantization coefficients. The DCT must be a change in the original block
coefficients are divided by their coefficients until (C1) value satisfies the
corresponding quantization coefficients relationship in point (2). This can be done by
(quantization table) and which then applying following steps.
rounded to the nearest integer (floor
value i.e. 0.5 is treated as 0). (a) Now compute the new Quantize DC
coefficient without rounding (C2).

C2 = (block (0,0) /16 (1.2)


(b) Find out the different (T) between 1. Load the stego image (bitmap 24 bits).
C1 and C2. To determine the number of This image contains the secret message.
pixels (NP) that must be changed by
subtracting one to the original value through 2. Convert the image formula from RGB to
comparing (NP) with table 1. It shows the YCbCr.
number of pixels that must be changed, 3. Separate image components into blocks,
subtract or add depended on the number of each one consists of (8 x8) pixels.
difference.
4. Transform each block (8 x 8) pixels to
If (C1>C2) T=C1-C2 (1.3) spatial frequency domain via the forward
If (C2>C1) T=C2-C1 (1.4) DCT. This step is executed on the Y
components only.
NP=0.5-T (1.5)
5. The fifth step includes:
TABLE 1 THE AMOUNT ADDED OR
SUBTRACTED FROM EACH BLOCK Extract the cipher text bits and
convert each group (12 bits) to a decimal
Difference Amount number and decrypt it by using a key.
0.000-0.063 8
Convert each number after
0.064-0.125 16 decryption to corresponding letter until
0.126-0.188 24 extracting all plain text. This step will be
0.189-0.250 32 described in detail.
0.251-0.313 40
6. Print the secret message.
0.314-0.375 48
0.376-0.437 56 5.Extracting cipher Bits
0.438-0.500 64
In this stage, the bits are extracted from the
stego-image. After converting the stego-
image from JPEG to BMP, convert the image
(c) After subtracting or adding from the
from RGB to YCbCr, separate the image
original coefficients block then employs the
components into blocks, each one consists
DCT in the same block again.
of (8 x 8) pixels and transform each block (8
4. Repeat Steps one, two and three continue x 8) pixels to spatial frequency domain via
with each block until hiding all the bits. the forward DCT. Some steps are
implemented to extract the bits:
4. Message Extracting Stage

This stage includes three parts, first extract


bits from a stego image and convert each 1. From each block the same DCT
group of bits (8 bits) to a decimal number, coefficients are used in embedding stage
second extract key from image and third choice to extract the bit. This coefficient is in
decrypt the decimal number to find the position (0, 0) in each block.
message. The extracting message stage
includes the following steps:
2. Divide the value in position (0, 0) by 16
and round the result called (C1) and inspects
the result.

C1 = round (block (0, 0) /16)

(a). If the C1 value is equals to odd


numbers, this means one is hidden in it.

(b). If the C1 value is equals to even


numbers, this means zero is hidden in it. Fig. 2. Fighter Jet Image after Stegnography

3. Convert each 12 bits to decimal numbers. TABLE 2 RESULTS OF THE EMBEDDED


DIFFERENTIAL LENGTH TEXT IN FIGHTER JET
4. Decrypt the decimal number by using the IMAGE
key. Convert the result to a corresponding
letter. Number of SNR By SNR By
Character in Using Using
5. Steps one, two, three and four continue Text file LSB Proposed
until finding the (#) character that refers to Technique DCT
Technique
the end of the text.
5 25.2036 34.933
.6. Test and Results 10 25.2021 34.9221
15 25.2012 34.9132
The testing of this example is performed by 20 25.2003 34.9028
taking the secret text of varying length and 25 25.1994 34.8938
the cover-image (Fighter Jet Image) 50 25.195 34.8501
(256*173). Figure 1 shows the cover image 100 25.1864 34.7674
before encryption and hiding. Figure 2 shows 200 25.1674 34.6024
the stego-image after extracting and 400 25.2012 34.1457
decryption. Table 2 shows the SNR for each 800 25.073 32.4249
Technique (LSB or Proposed DCT) with
varying text length for Fighter Jet Image. Case 1 SNR using LSB and DCT for 5
character text file is 25.2036 and 34.933
respectively

Case 2 SNR using LSB and DCT for 800


character text file is 25.073 and 32.4249

Fig. 1. FighterJet Image Before Stegnography


respectively The SNR value for the color Fighter Jet
image is between (34.9- 32).

The SNR value decreases with the increase


in length of the secret message.

8. Future Scope

1. The method of embedding is DCT


transform insertion, in the future another
embedding method should be employed like
wavelet or Sparta spectrum methods.

2. The Encryption method could be replaced


by other public keys ciphering algorithm.

9. REFERENCES

[1] W. Stallings, Cryptography and Network


Security: Principles and Practice, third ed.,
Pearson Education, New Jersey, 2003.

[2] N. F. Johnson and S. Katzenbeisser, .A


survey of steganographic techniques,
Information Hiding, pp.43-78. Artech House,
Norwood, MA, Journal,2000.

[3] Lou, D. C. and Liu, J. L.. Steganography


Method for Secure Communications.
Elsevier Science on Computers& Security,
21, 5:pp. 449-460, 2002.

[4] J. Fridrich and M. Goljan,Practical


steganalysis of digital images-state of the
Fig .3. Graph Showing Result SNR versus No of
character in text file for LSB and Proposed DCT for
art, Proc. SPIE Photonics West, San Jose,
FighterJet Image California, Vol. 4675, pp. 1-13, Jan. 2002.

7. Conclusions [5] Moerland, T., Steganography and


Steganalysis, Leiden Institute of Advanced
This work is implemented using DCT based Computing Science,www.liacs.nl/home/
image steganography algorithms Also a tmoerl/privtech.pdf
cryptography algorithm is used, which make
it difficult to extract information by the
intruder. A number of conclusions derived
from this study

1. Using cover images and a secret message


with different lengths are that

You might also like