A fpga implementation of data hiding using lsb matching method

IJRET: International Journal of Research in Engineering and Technology eISSN: 2319-1163 | pISSN: 2321-7308
_______________________________________________________________________________________
Volume: 04 Issue: 03 | Mar-2015, Available @ http://www.ijret.org 194
A FPGA IMPLEMENTATION OF DATA HIDING USING LSB
MATCHING METHOD
S. Raveendra Reddy1
, Sakthivel S M2
1
M. Tech (VLSI Design), Department of Sense, VIT University, Chennai, India
2
Asst. Professor (Sr.G), Department of Sense, VIT University, Chennai, India
Abstract
Data Hiding is one of the familiar methodologies used nowadays to authenticate and resolve the copyrights of the digital data.
This paper discusses the verilog based FPGA implementation of data hiding in grayscale image using the concept of the LSB
Matching technique. In this verilog based data hiding process the secret digital signature is hidden in the host image then
analyzed with the PSNR value and Payload capacity.
Keywords: Data hiding, Least-Significant-bit matching method, Reversible, Gray Scale Image, Real Time
LSB Embedding, FPGA- Implementation etc..
--------------------------------------------------------------------***----------------------------------------------------------------------
1. INTRODCUTION
The originality of the digital data is altered during the
process of transmission and sharing over internet [1]. Hence
there exists a problem in sending the data in a imperceptible
manner [2]. Data hiding is one of the methods to hide the
secret data in the host content and transmit in an effective
manner [3]. The digital data can be an image, audio and
video etc [4,5]. The data hiding process is said to be
reversible if the original data can be reconstructed just by
reversing the algorithm. Here the host is considered as a gray
scale image and watermark as a sequence of binary data bits.
Using LSB Matching methodology the secret data bits are
hidden inside an image as secret signature.
All the data hiding algorithms till date are of software
implementation only, a few are in hardware. If the data
hiding operation is carried out by the Field Programmable
Gate Array (FPGA) Chip, the process is quite fast and area,
power and timing optimization is possible due to the
presence of custom circuitry [6-8]. The FPGA based data
hiding is a real time implementation whereas the software
based approaches are doing the data hiding in offline mode
of processing [6-8].
The proposed data hiding approach is the FPGA
implementation of LSB Matching by J.Mielikainen[9] in real
time using Verilog HDL language. Generally the data hiding
process has to be efficient if it has a high payload capacity
with good degree of imperceptibility. In this work the data
hiding process is evaluated using the values of PSNR (i.e. to
measure the Imperceptibility) and hiding capacity (i.e. no. of.
bits hidden).
2. LSB MATCHING
This Section explains the process of data hiding in a
grayscale image using the concept of LSB Matching
technique. In the normal LSB embedding method simply we
replace the LSB bits of the host content directly whereas the
LSB Matching based data embedding involves the checking
of the four conditions and finally the secret data signature is
hidden in the host content.
In this LSB matching based data hiding process the two
pixel pairs are selected for the secret data embedding and
checked for the four conditions stated in the four different
cases as below. The same process is repeated for the entire
pairs of the pixels in the host image are processed. In case of
the general LSB embedding technique the embedding is
carried out by direct pixel replacement here it is processed
by four checking process and then embedding. Hence this
methodology will preserve the host image quality when
compared to ordinary LSB embedding technique. The entire
LSB matching based data hiding process is represented
using a flow chart form in the Fig-1.
Input pi,j , pi,j+1 , s1 , s2
p’i,j =pi,j
p’i,j+1 =pi,j+1
p’i,j =pi,j
p’i,j+1 =pi,j+1 +1
p’i,j =pi,j -1
p’i,j+1 =pi,j+1
p’i,j =pi,j +1
p’i,j+1 =pi,j+1
LSB(pi,j) =s1
F(pi,j , pi,j+1) =s2 F(pi,j -1, pi,j+1) =s2
Yes
NoYes
No
NoYes
Fig.1. Strategy of Data Hiding
In the data hiding process two pixels and are
selected first and the four case stated below are checked for
the digital signature hiding. Here the secret data bits are
considered as the and . Similarly and are the

_______________________________________________________________________________________
two modified pixels after the data hiding process. The four
cases for the embedding action is given below as follows.
Case-1: When and , the pixel
pair and does not need modification.
does not change, and .
, and does not change.
, and does not change.
For all the cases consecutive pixel are taken
as and and the functional LSB bits of are
calculated.
In the Case-1 the criteria is checked in such a way that if the
LSB bit of is same as the secret bit and the
calculatedfunctional LSB bit F(.) bit using the equation -1 is
same as the secret bit S2then the
pixels and doesn’t need any modification
The Case-2 analysis holds the situation if the LSB bit
of is same as the secret bit S1 and the F(.) is not equal to
S2 then pixel is replaced by simply.
Whereas the pixel is maintained as such.
The Case-3 is applicable to data hiding if the LSB of is
not equal to S1 and F(.) is equal to S2 then the
pixel undergoes a modification as
whereas the is not disturbed.
For the Case-4 analysis if the LSB of is not equal to
S1and the F(.) is not equal to S2 then the pixel is only
replaced with .
The entire process of data hiding is explained here with an
example sample pair of pixels. Consider the pixel pairs
, and confidential information
as an example. When LSB (49) =1 and LSB (49) =1,
then it is substituted in Eq. (1) to give
F(49,44)=LSB(68)=0 and F(49,44)=0. The case-1 condition
are met, and this results in modified pixels as and
. In the same way the four cases are iteratively
verified for data hiding row wise and column wise of the
image until all the pixels pairs are processed. By this LSB
matching concept with the secret bits and modification in
the pairs of the pixel the data hiding is carried out
effectively. The same can be used for the any type of
grayscale images. As compared to ordinary LSB embedding
the data hiding is happened based upon the matching
process we can able to preserve the host image quality as
such in this process.
Fig.2. Input image before Data Hiding
Fig.3. Output image after Data Hiding
Fig.4. Image hexadecimal values

_______________________________________________________________________________________
3. RESULTS AND DISCUSSIONS
This section gives the simulation results of the LSB
matching process of data hiding using Matlab and the Xilinx
for the FPGA implementation. During the Matlab simulation
of data embedding we have taken an image size of 160 X
160 in Fig-2. The entire strategy is modeled in Matlab and
the secret signature is embedded. Here the signature is
considered as a random pattern. The input image before and
after data hiding are shown in Fig- 2 & 3 respectively. From
the Matlab simulation results it is found that the embedding
capacity for this kind of LSB matching based embedding is
of 12800bits with a high PSNR ratio of 52dB.
For the Verilog simulation of the LSB based data hiding the
image pixel intensity values are converted into hexa decimal
values as in Fig-4. Then the converted Hexa decimal values
are stored in a RAM of size as the host image size. Then the
data embedding strategy is modeled as a top level entity
using Verilog HDL. Then the same entity is simulated using
Xilinx ISE 14.3 with inputs of image hex values and secret
signature from the RAMs. Here the secret signature is also
stored in a RAM. The simulation results for the data hiding
process are shown in the Fig-5. Finally the output images
are also stored in the output Ram as hexa decimal values.
For the reconstruction of the image the image hexa decimal
values are wrote into the text file and watermarked image is
converted using the Matlab. From the simulation of the
verilog results, a comparison has been made with Matlab
outputs and found that both are same in nature. (i.e. there is
no deviation for Matlab and FPGA simulation results).
After verifying the simulated results the top level entity is
synthesized using the Xilinx ISE 14.3 using the device
family Artix 7 family with a speed grade of -3. In the Artix
family the device XC7A100t is used with a package of
CSG324. The device utilization summary for the data
hiding strategy after the synthesis is shown in the Table-II.
The table-I shows the embedding capacity and PSNR ratio
value for the data embedding process. From the synthesis
results it is found that there is no critical path in the design.
The RTL technology level schematic is shown in the Fig-6.
The data hiding process is evaluated by the values of the
embedding capacity and how effectively the secret signature
is embedded. Here the above two metrics are evaluated by
the values of embedding capacity and PSNR ratio
respectively. The embedding capacity is indicated by bits
per pixel (bpp) and the imperceptibly is given by the values
of PSNR in dB.
Table –1 Embedding capacity and PSNR Value
Embedding Capacity (bpp) PSNR (dB)
0.50 bpp 52 dB
Fig.5. Simulation result of Data Hiding process

_______________________________________________________________________________________
Table – 2 Device utilization Summary for Data Hiding
Parameters LSB Method
Number of slice LUTs 23
Number of fully used LUT-FF pairs 0
Number of bonded IOBs 35
Number of BUFG / BUFGCTRLs 1
Fig.6. RTL Schematic of Data Hiding process
4. CONCLUSION
In this paper a FPGA implementation of watermarking in
real time using LSB matching process is done.Our algorithm
does embedding in the host image in an invisible manner.
As it is a FPGA chip it can be incorporated a watermarking
coprocessor inside a digital camera to carry out the
watermarking, in the instant of capturing the picture itself.
REFERENCES
[1] Cox, I.J., Miller, M.L., Bloom, J.A., Fridrich, J.,
Kalker, J . : ‘Digital watermarking and
steganography’ (Morgan Kaufmann
Publishers,2008).
[2] C.T. Li, Digital fragile wateramrking scheme for
authentication of JPEG images, in: Proceedings –
Vision, Image and Signal processing, vol. 151(6),
2004, pp. 460-466.
[3] Neil F. Johnson, Sushil Jajodia, Exploring
steganography: seeing the unseen, Computer
Practices (1998) 26–34.
[4] R.J. Anderson, F.A.P. Petitcolas, On the limits of
steganography, IEEE Journal on Selected Areas in
Communications 16 (1998) 474–481.
[5] F.A.P. Petitcolas, R.J. Anderson, M.G. Kuhn,
Information hiding—a survey, Proceedings of the
IEEE, special issue on protection of multimedia
content 87 (7) (1999) 1062–1078.
[6] Saraju P. Mohanty, N. Ranganathan and Ravi K.
Namballa, “VLSI Implementation of Visible
Watermarking for a Secure Digital Still Camera
Design”, IEEE Proceedings of the 17th International
Conference on VLSI Design (VLSID’04)
[7] S.P. Mohanty, N. Ranganathan, K. Balakrishnan, “A
dual voltage-frequency VLSI chip for image
watermarking in DCT domain”, IEEE Transactions
on Circuits and Systems II (TCAS-II) 53 (5)
(2006),pp. 394–398.
[8] P.Karthigaikumar,K.Baskaran, FPGA and ASIC
implementation of robust invisible binary image
watermarking algorithm using connectivity
preserving criteria, Microelectronics Journal 42
(2011),pp.82-88.

_______________________________________________________________________________________
[9] Mielikainen, LSB matching revisited, IEEE signal
Processing Lett, 13(5)(2006) 285-287.
[10] D.C.Lou, C.L.Chou, H.Y.Wei,H.F.Hung, Active
steganalysis for interpolation error based reversible
data hiding, Pattern Recognit. Lett. 34(9) (2013)
1032-1036.
[11] X.Zhang and S.Wang, Steganography using
multiplebase notational system and human vision
sensitivity,IEEE Signal Process. Lett., vol.12, no. 1,
pp. 67–70, Jan. 2005.
[12] A.M. Alattar, Reversible watermark using the
difference expansion of a generalized integer
transform, IEEE Trans. Image Process.13 (2004)
1147-1156.
[13] A. Ker, Improved detection of LSB steganography in
Grayscale images, inProc. Information Hiding
Workshop, vol. 3200, Springer LNCS, 2004, pp. 97–
115.
[14] T. Sharp,” An implementation of key-based digital
Signal steganography, in Proc. Information Hiding
Workshop, vol. 2137, Springer LNCS, 2001, pp. 13–
26.
[15] Gil-Je Lee, Eun-Jun Yoon, Kee-Young Yoo, A new
lsb based digital water marking scheme with random
mapping function, IEEE Computer Society (2008).
BIOGRAPHY
S. Raveendra Reddy pursuing M.TECH
degree in VLSI Design from the Vellore
Institute of Technology, Chennai, (2013-
2015) and the B.TECH degree in
Electronics and communication
engineering from the R.V.R institute of
Engg & technology, Hyderabad, in 2012.

A fpga implementation of data hiding using lsb matching method

More Related Content

What's hot (17)

Viewers also liked (18)

Similar to A fpga implementation of data hiding using lsb matching method (20)

More from eSAT Journals (20)

Recently uploaded (20)

A fpga implementation of data hiding using lsb matching method