Conceptual Approach of Hybrid DWT and SVD Based Digital Image Watermarking For Fingerprint Security

International Journal on Recent and Innovation Trends in Computing and Communication ISSN: 2321-8169
Volume: 5 Issue: 6 1067 1072

_______________________________________________________________________________________________
Conceptual Approach of Hybrid DWT and SVD based Digital Image

Watermarking for Fingerprint Security
Zareen Khan Prof. Tirupati Goskula
Research scholar (M.Tech), Dept. of ECE Assistant Professor, Dept. of ECE
Anjuman College of Engineering and Technology Anjuman College of Engineering and Technology
Nagpur, India Nagpur, India
zareenkhan.zk91@gmail.com thiru_aar@rediffmail.com
Abstract Protection of biometric data is gaining interest and digital watermarking techniques are used to protect the biometric data from either
accidental or intentional attacks. The information used for identification or verification of a fingerprint, mainly lies in its minutiae. Digital image
watermarking is been a useful solution to numerous problems like for information security, for copyright and for network security. So a new
watermarking algorithm is proposed based on Singular Value Decomposition (SVD) and discrete wavelet transform (DWT). The algorithm
converted the minutiae into binary watermark, increasing embedded information capacity. The algorithm can satisfy the transparence and
robustness of the watermarking system very well and the useful information can be extracted accurately even if the fingerprint is severely
degraded.
Keywords- Digital watermarking, DWT ,SVD, Arnold transform, PSNR, NC, Copyright protection.
__________________________________________________*****_________________________________________________
I. INTRODUCTION
The reason for watermarking in the frequency domain is its
Biometric based authentication systems have inherent
characteristics of the Human Visual System (HVS) ..and are
advantage over traditional personal identification techniques.
better captured by the spectral coefficients.
Mainly biometrics, fingerprints are more famous in the
In this paper, we introduce Adaptive DWT and SVD based
authentication area, as they are unique to each person and are
watermarking method for fingerprint images and this method
widely used in identification and verification of personal
can be used in steganography-based application to embed
individuality. However, they are influenced by accidental and
minutiae data in fingerprint images. The use of a biometric
intentional attacks. Thus, a defensive scheme is needed which
data (fingerprint image) to hide an other one (fingerprint
will preserve fidelity and prevent modifications. Digital
minutiae) increases the level of security because the
watermarking provides strong solution for it. Digital Image
unauthorized person who obtains the fingerprint image
Watermarking is a technique which provide solution for
watermarked is likely to treat the fingerprint image instead the
copyright, image validity and other issues. Watermarking
hidden minutiae data. The objective of this watermarking
deals with decomposing original image called cover image
scheme is to ensure secure transmission of minutiae details
using some wavelet transforms and embedding watermark into
even over a noisy communication channel or over one subject
one of the sub band (LL,LH,HL,HH) the obtained image is
to intentional tampering.
called watermarked image. This image have transmitted
This is the introductory phenomena which is section 1.
through channel where various noise affect watermarked
In this paper, In section 2 we described about the Fingerprint
image. At receivers side embedded watermark has extracted
minutiae, In section 3 DWT methods and SVD are introduced,
from watermarked image.
In section 4 digital watermark embedding and extraction
There are many methods to embed the watermark. There are
algorithm are presented in detail.
two classes classified as spatial-domain and transform-domain
watermarks. The spatial domain is so simple that the
watermark can be damaged easily, but the transform-domain II. FINGERPRINT MINUTIAE
algorithm can be resist intensity attack, watermark information A fingerprint is formed from an impression of the thumb
cant be damaged easily. pattern of ridges on a finger. . Each individual has its own
The most commonly used transforms are the Discrete Cosine fingerprints with the permanent uniqueness. So fingerprints
Transform (DCT), Discrete Fourier Transform (DFT) and have been used for identification and forensic investigation for
Discrete Wavelet Transform (DWT). a long time. A fingerprint is composed of many ridges and
bifurcation.
1067
IJRITCC | June 2017, Available @ http://www.ijritcc.org
_______________________________________________________________________________________
Volume: 5 Issue: 6 1067 1072
_______________________________________________________________________________________________
However, shown by intensive research on fingerprint and robustness, the approximation sub-image LL3 coefficients
recognition, fingerprints are not distinguished by their ridges are chosen to embed watermark.
and furrows, but by Minutiae, which are some abnormal points We can achieve the transform of the separable wavelet as in
on the ridges as shown in Fig (a) . Minutiae points are most Fig.2.
generally the locations of ending or bifurcation of ridges in a
fingerprint
Fig.1. Three level wavelet decomposition

(a)
B. SINGULAR VALUE DECOMPOSITION (SVD)

The Theory of singular value decomposition (SVD) was
established for real square matrices in the 1870s by Beltrami
and Jordan, for complex matrices by Autonne in 1902 and has
been extended to rectangular matrices by Eckart and Young in
the 1939. Recently, singular value decomposition has been
(b) used in image processing applications, including image
compression image hiding and noise reduction
If a m*n image is represented as a real matrix A , it
Fig.1.(a) A typical fingerprint image showing (b) Minutiae .
can be decomposed as-
(Valley is also referred as Furrow, Termination is also called
Ending and Bifurcation is also called Branch)
. A= U S VT
III. WATERMARK EMBEDDING

It is called a singular value decomposition of A . Where U
A. DISCRETE WAVELET TRANSFORM (DWT) is a m*m unitary matrix, S is a m*n matrix with nonnegative
Wavelet transform is a time-frequency domain numbers on the diagonal and zeros on the off diagonal, and V T
combined analysis method. It has multi-resolution analysis denotes the conjugate transpose of V, an n*n unitary matrix.
features. Each level of the wavelet decomposition has four sub- The nonnegative components of S represent the luminance
images with same size. Let the LLk stands for the pixel value of the image. Changing them slightly does not
approximation sub image and LHk, HLk ,and HHk stand for the affect the image quality and they also dont change much after
horizontal , vertical and diagonal direction high-frequency attacks, watermarking algorithms make use of these two
detail sub image respectively. Where the variable k = 1,2,3,...(k properties.
N) is the scale or the level of the wavelet decomposition.
DWT is much preferred because it provides both simultaneous
spatial localization and frequency spread of the watermark s1 0 0
within the host image. The basic idea of discrete wavelet
transform in image process is to multi-differentiated the image
into sub image of different spatial domain and independent S= 0 s2 0
frequencies.
After wavelet decomposition, many signal processing, such as 0 0 sN
compression and filter are likely to change the high frequency
wavelet coefficients. If the watermark sequence is embedded
into this part, its information may be lost in the processing in
sequence, which will reduce the robustness of the watermark .
In order to ensure the watermark has a better imperceptibility
1068
_______________________________________________________________________________________
Volume: 5 Issue: 6 1067 1072
_______________________________________________________________________________________________
Singular value decomposition is a linear algebra technique
used to solve many mathematical problems. The theoretical
START
background of SVD technique in image processing
applications to be noticed is-
a) The SVs (Singular Values) of an image has very good
Read host cover fingerprint
stability, which means that when a small value is
added to an image, this does not affect the quality
with great variation.
b) SVD is able to efficiently represent the intrinsic Fingerprint Pre-Processing Implement 3-level
algebraic properties of an image, where singular (Enhancement, Wavelet transform( haar
Binarization,Thinning) wavelet)
values correspond to the brightness of the image and
singular vectors reflect geometry characteristics of
the image. Select approximation
Minutiae extraction (x,y)
c) An image matrix has many small singular values LL3 sub-band image(A)
compared with the first singular value. Even ignoring
these small singular values in the reconstruction of
the image does not affect the quality of the Conversion of minutiae Implement SVD on A
reconstructed image. points to binary i.e; A= A=USVT
watermark(W)
C. ARNOLD TRANSFORM
Arnold transform is commonly known as cat face
transform, it randomizes the original organization of pixels in Arnold Transform
an original image. However, if iterated enough times, the
original image reappears. Arnold transform is the position shift
of one to-one point. Arnold transformation on digital image of Obtain watermark image coefficient
size NN defined by Eq. 1 which is a one-to-one Aw as S + W = UW SW VWT
T
transformation. AW = USW V
(1) Apply reverse wavelet tarnsform
Where (i, j) is the location coordinates of the original image

pixels and (i, j) is the location coordinates of image pixels that Change double precision real no. to unsigned 8-bit
after transform When all the coordinates are transformed, the integer
image we obtain is scrambled images.
Watermarked Image
IV. PROPOSED WATERMARKING ALGORITHM
The proposed watermarking scheme has been shown in Figure

below. The cover image is the fingerprint while the watermark Fig.2. Proposed flow-diagram architecture
is a binary image equivalent to the minutiae of the cover
fingerprint. The scheme has been divided into two sections-
A. Watermark (Minutiae) Embedding.
B. Watermark (Minutiae) Extraction.
1069
_______________________________________________________________________________________
Volume: 5 Issue: 6 1067 1072
_______________________________________________________________________________________________
A. Watermark (Minutiae) Embedding scheme If crossing Number is 1, 2 and 3 or greater than 3 then
minutiae points are classified as Termination points, Normal
ridges and Bifurcation respectively. The only minutiae of
3-level 2D dwt interest are ridge endings and bifurcation (corresponding to
SVD
decomposition CN=1 and CN=3 respectively).From the extracted minutiae
points x and y co-ordinate, (orientation of ridges ) and type of
Compare
Svs of minutiae i.e. termination or bifurcation are obtained. Type of
watermark minutiae can already be set as 0 for ending and 1 for
with bifurcation, making it a binary format . The remaining three
Fingerprint Arnold original columns of minutiae can be converted into binary form
Pre- transform image representation and a binary watermark is generated by
Processing
concatenating the eight individual bit planes and binary
IDWT watermark (W) from minutiae is created.
Conversion of
Step 4-Perform Arnold transform for watermark image W.
Minutiae minutiae to Watermarked
extraction binary Fingerprint
watermark Step 5- We obtain the watermarked image coefficients matrix
AW through the following three steps:
1. A = USVT
Fig.3. BLOCK DIAGRAM OF WATERMARK (Minutiae) 2. S + W = UW SW VWT
EMBEDDING
3. AW = USW VT
The steps of embedding watermarks can be described as
follows- Step 6- Apply reverse wavelet transform for original image,
and then changing the double-precision real number to
Step1- The fingerprint image is decomposed into its 3-level 2- unsigned 8-bit integer. Thus, obtain the watermarked image in
dimensional DWT coefficients. Out of the all sub-bands, only which watermark are embedded.
LL3 approximation sub-band is selected (denoted as A ).
B. Watermark (Minutiae) Extraction scheme
Step 2- Fingerprint Preprocessing
A real fingerprint might have discontinuities that might lead
to spurious minutiae. Therefore, minutiae extraction is Watermarked 3-level 2D dwt
SVD
preceded by fingerprint preprocessing, which involves Fingerprint decomposition
normalization, ridge orientation and frequency estimation. (LL3 sub-band)
Finally, the ridge orientation and frequency estimation values
are used for filtering the fingerprint using Gabor wavelet.
Gabor filtering enhances the ridges oriented in the direction of Extracted
the local orientation, and decreases anything oriented watermark Arnold
(binary Inverse arnold transform
differently. Hence, the filter increases the contrast between the
equivalent transform
foreground ridges and the background, whilst effectively to minutiae)
reducing noise. The filtered output is then binarized and
thinned to one-pixel width.
Fig.4. BLOCK DIAGRAM OF WATERMARK (Minutiae)
Step 3- Minutiae Extraction EXTRACTION
Minutiae points such as end points and bifurcation points
are identified by calculating Crossing number (CN). CN value
We can extract the watermark by the reverse
is defined as half the sum of the differences between pairs of
adjacent pixels in the eight neighbourhoods.
Calculation of watermark embedding:
CN=0.5Pi_Pi+1 |, P1= P9
Step 1- Perform a 3-level wavelet transform using haar wavelet
where Pi is the pixel value in the neighborhood of a pixel P.
for watermarked image, and obtain low-frequency wavelet
coefficient LL3 (denotes as A*).
1070
_______________________________________________________________________________________
Volume: 5 Issue: 6 1067 1072
_______________________________________________________________________________________________
Step 2- Apply SVD to the A*, such that A* =U * S1 *VT*, and
obtain U *,S1* and VT*.
Step 3- Now by using values of U w, VW and S1 *, obtain D*

according D*= UW S1 *V T, in the end we can obtain the
watermark which is embedded according to W* = (D* S)
/.
Step 4- Finally changing the double-precision real number to

unsigned 8-bit integer for watermark image, and perform
IMAGE 2 [ GUI ]
inverse Arnold transform for watermark image.
Step 5- Minutiae points are then regenerated by stacking bit

planes and converting them back to decimal system.
V. RESULTS
IMAGE 3 [NOISE ATTACK]
IMAGE 4 [NOISE ATTACK]
IMAGE 1 (DWT output)
IMAGE 5 [ NOISE ATTACK]
1071
_______________________________________________________________________________________
Volume: 5 Issue: 6 1067 1072
_______________________________________________________________________________________________
VI. CONCLUSION REFERENCES
[1] Asma Furqan ,Munish Kumar ,"Study and Analysis of Robust
DWT-SVD based Digital Image Watermarking technique
using MATLAB ", 2015 IEEE International Conference on
Computational Intelligence and Communication Technology.
[2] Onur Jane ,Hakki Gokhan Ik , Ersin Elbasi, "A Secure and
Robust Watermarking Algorithm Based on the Combination
of DWT,SVD and LU Decomposition with Arnold's Cat Map
Approach ", 2013 IEEE International Conference on
Electrical and Electronics Engineering (ELECO).
[3] Shuchi Sirmour, Archana Tiwari , "Hybrid DWT-SVD based
Digital Watermarking Algorithm for Copyright Protection" ,
March 2014 International Journal of P2P Network Trends and
Technology (IJPTT).
[4] Prabhishek Singh, R S Chadha, "A Survey of Digital
Watermarking Techniques, Applications and Attacks",
International Journal of Engineering and Innovative
Technology (IJEIT), Volume 2, Issue 9, ISSN- 2277-3754,
In this paper, we proposed a watermarking algorithm for March 2013.
fingerprint images based on hybrid DWT and SVD domain. [5] Akshay Kumar Gupta, and Mehul S Raval, "A robust and
The purpose of the proposed watermarking scheme is to extract secure watermarking scheme based on Singular values.
minutiae from even tampered fingerprint. . Due to spatio- replacement,"in Indian Academy of Sciences, Volume-37,
frequency resolution of DWT, it offers more degrees of Part 4, August -2012.
[6] Rajlaxmi Chouhan, Pritee Khanna, Robust Minutiae
freedom as compared with DCT.The main objective of this
Watermarking in Wavelet Domain for Fingerprint Security
scheme is to provide security and robustness to fingerprint
in World Academy of Science, Engineering and Technology-
images. The watermark (binary equivalent of minutiae) is 60 2011.
embedded into LL3 sub-band because watermark detection at [7] Miss Jalpa , Mr.Prayag Patel, A brief survey on digital image
lower resolution is computationally effective. Also in order to watermarking techniques in International Journal for
ensure watermark has better imperceptibility.Applying the Technological Research in Engineering Volume 1, Issue 7,
Arnold transform to the watermark makes the results even March-2014 in ISSN.
better. [8] Praful Saxena, Shanon Garg and Arpita Srivastava, DWT-
SVD SemiBlind Image Watermarking Using High Frequency
Band," 2nd International Conference on Computer Science
VII. FUTURE WORK and Information Technology (ICCSIT2012), Singapore, April
As a future work, the implemented algorithm can be 28-29, 2012.
[9] Dr. M. Mohammad Sathik, S. S. Sujatha, Authentication of
improved using full band DWT-DCT-SVD and further can be
Digital Images by using a Semi-Fragile Watermarking
extended to color images and video processing.
Technique, International Journal of Advanced Research in
TABLE I. LEVELS OF NOISE ATTACKS ON FINGERPRINT IMAGE
Computer Science and Software Engineering, Volume 2.
[10] Emir Ganic, Ahmett M.Eskicioglu, Robust DWT-SVD
Effect of Noise Attacks on Watermark Extraction domain Image Watermarking-Embedding date in all
Sr. no. Cross-
Compression
Frequencies, CitesteerX,MM&SEC04, September 20-21,
Attacks Correlation 2004, Magdeburg, Germany.
levels
value
[11] Praful Saxena, Shanon Garg and Arpita Srivastava, DWT-
Barcoded
0 No attack 0.9864 SVD Semi Blind Image Watermarking using High Frequency
Image
Gaussian 0.9353
Barcoded band, 2nd International Conference science and Information
1 Image technology(ICCSIT2012), Singapore, April 28 to 29, 2012.
Barcoded
2
Salt and pepper 0.94672
Image [12] Zhou X, Dusn X., A Semi-fragile Watermark scheme for
Barcoded Image Authentication, IEEE International Conference on
3 Cropping O.94122
Image Multimedia modeling, pp.374 to 377, 2004.
Barcoded
4 Rotation 0.8739
Image
Barcoded
5 Blurring 0.95563
Image
Barcoded
6 Median filter 0.95315
Image
1072
_______________________________________________________________________________________

Conceptual Approach of Hybrid DWT and SVD Based Digital Image Watermarking For Fingerprint Security

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Conceptual Approach of Hybrid DWT and SVD Based Digital Image Watermarking For Fingerprint Security

Uploaded by

Copyright:

Available Formats

International Journal on Recent and Innovation Trends in Computing and Communication ISSN: 2321-8169

Volume: 5 Issue: 6 1067 1072

Conceptual Approach of Hybrid DWT and SVD based Digital Image

Fig.1. Three level wavelet decomposition

B. SINGULAR VALUE DECOMPOSITION (SVD)

III. WATERMARK EMBEDDING

(1) Apply reverse wavelet tarnsform

Where (i, j) is the location coordinates of the original image

The proposed watermarking scheme has been shown in Figure

A. Watermark (Minutiae) Embedding.

B. Watermark (Minutiae) Extraction.

Step 3- Now by using values of U w, VW and S1 , obtain D

Step 4- Finally changing the double-precision real number to

Step 5- Minutiae points are then regenerated by stacking bit

IMAGE 3 [NOISE ATTACK]

IMAGE 4 [NOISE ATTACK]

IMAGE 1 (DWT output)

IMAGE 5 [ NOISE ATTACK]

You might also like