Highly Secured Bio-Metric Authentication Model with Palm Print Identification

Dr. Raja Murali Prasad. Int. Journal of Engineering Research and Application www.ijera.com
ISSN: 2248-9622, Vol. 6, Issue 4, (Part - 7) April 2016, pp.19-24
www.ijera.com 19|P a g e
Highly Secured Bio-Metric Authentication Model with Palm Print
Identification
Dr. Raja Murali Prasad
Professor, Vardhaman College of Engineering, Hyderabad, Telangana, India
ABSTRACT
For securing personal identifications and highly secure identification problems, biometric technologies will
provide higher security with improved accuracy. This has become an emerging technology in recent years due to
the transaction frauds, security breaches and personal identification etc. The beauty of biometric technology is it
provides a unique code for each person and it can’t be copied or forged by others. To overcome the draw backs
of finger print identification systems, here in this paper we proposed a palm print based personal identification
system, which is a most promising and emerging research area in biometric identification systems due to its
uniqueness, scalability, faster execution speed and large area for extracting the features. It provides higher
security over finger print biometric systems with its rich features like wrinkles, continuous ridges, principal
lines, minutiae points, and singular points. The main aim of proposed palm print identification system is to
implement a system with higher accuracy and increased speed in identifying the palm prints of several users.
Here, in this we presented a highly secured palm print identification system with extraction of region of interest
(ROI) with morphological operation there by applying un-decimated bi-orthogonal wavelet (UDBW) transform
to extract the low level features of registered palm prints to calculate its feature vectors (FV) then after the
comparison is done by measuring the distance between registered palm feature vector and testing palm print
feature vector. Simulation results show that the proposed biometric identification system provides more
accuracy and reliable recognition rate.
Index terms: personal identification, finger print recognition, face recognition, IRIS recognition, ROI, Hybrid
wavelet and recognition rate
I. INTRODUCTION
Identifying our selves is omnipresent in
daily lives at many places such as accessing bank
accounts, cash draw from ATM, computer logging,
entering into a protecting sites and so on. Formally,
one can access their self by physically carrying the
passports, remembering pass words, access cards,
keys; personal identification numbers (PINs) and
secret codes. Regrettably, all the mentioned
identifications can be lost, copied, forgotten or
even stolen. Such loopholes or deficiencies cause
many serious issues to all concerned people. For
example, all over world the hackers often interrupt
computer networks; credit card fraud is
approximated at billion dollars per annum.
Forgotten passwords cost will be very high,
Therefore, we need a solution for all the above
deficiencies in conventional personal identification
techniques which is more reliable, robust and
foolproof personal identification solution that could
verify that physically he/she claims to be. A
biometric is a method that recognizes the identity
of a person or human being automatically by doing
the statistical analysis of biological characteristics.
The measurable characteristics can be physical,
such as finger, eye, face or palm. Common
modalities being used as biometric personal
identification systems are face recognition and
finger print identification. But authentication with
face is still a problem due to its illumination
invariance, occlusion effects and pose variations
where as finger print does not have a good
psychological effect on the user because of it wide
use in investigations of crime. Hence, in future if
any biometric system that should get succeeds have
the attributes like accuracy, easy acquisition,
richness, uniqueness, reliability and all above user
acceptance. Palm print identification system is a
new modality of biometric system which will
overcome all the deficiencies occur with
conventional personal identification systems such
as finger print, face recognition and iris
recognition. It not only has the unique information
but also has far more amount of details such as
principal lines, creases and wrinkles. Moreover, it
has rich features to analyze more effectively and to
improve the security. It has entered into a biometric
family and become most promising personal
identification system with higher security and
improved accuracy due to its easy acquisition,
reliability and high user acceptance. There are
many researchers in the literature, who have
developed palm print based personal identification
systems using edge detection, region of interest
RESEARCH ARTICLE OPEN ACCESS

(ROI), discrete cosine transform (DCT), short time
fourier transform (STFT), principle component
analysis (PCA) and independent component
analysis (ICA). All the above algorithms have
suffered from lack of features extraction and time
complexity. Here, in this we presented a highly
secured palm print identification system with
extraction of region of interest there by applying
hybrid wavelet to extract the low level features of
registered palm prints to calculate its feature
vectors then after the comparison is done by
measuring the distance between registered palm
feature vector and testing palm print feature vector.
II. EXISTING METHODS
In the literature, there are many
researchers who have developed biometric
authentication modules based on various spatial
and transformation domain techniques. D. Huang,
W. Jia, and D. Zhang[1] proposed a novel
algorithm for the automatic classification of low-
resolution palmprints. First the principal lines of
the palm are defined using their position and
thickness. Principal lines are defined and
characterized by their position and thickness. A set
of directional line detectors is devised for principal
line extraction. By using these detectors, the
potential line initials of the principal lines are
extracted and then, based on the extracted potential
line initials, the principal lines are extracted in their
entirety using a recursive process. The local
information about the extracted part of the principal
line is used to decide a ROI and then a suitable line
detector is chosen to extract the next part of the
principal line in this ROI. After extracting the
principal lines, some rules are presented for
palmprint classification. A. Kong and D. Zhang [2]
have presented a novel feature extraction method,
the Competitive Coding Scheme for palmprint
identification. This scheme extracts the orientation
information from the palm lines and stores it in the
Competitive Code. An angular match with an
effective implementation is developed for
comparing Competitive Codes. Total execution
time for verification is about 1s, which is fast
enough for real-time applications. The proposed
coding scheme has been evaluated using a database
with 7,752 palmprint images from 386 different
palms. For verification, the proposed method can
operate at a high genuine acceptance rate of 98.4%
and a low false acceptance rate of 3*10-6. Dai and
Zhou [3] introduces high resolution approach for
palmprint recognition with multiple features
extraction. Features like minutiae, density,
orientation, and principal lines are taken for feature
extraction. For orientation estimation the DFT and
Radon-Transform-Based Orientation Estimation
are used. For minutiae extraction Gabor filter is
used for ridges enhancement according to the local
ridge direction and density. Density map is
calculated by using the composite algorithm, Gabor
filter, Hough transform. And to extract the
principal line features Hough transform is applied.
SVM is used as the fusion method for the
verification system and the proposed heuristic rule
for the identification system. Jiaa, Huanga and
Zhang [4] and [5] have proposed palmprint
verification based on robust line orientation code.
Modified finite Radon transform has been used for
feature extraction, which extracts orientation
feature. For matching of test image with a training
image the line matching technique has been used
which is based on pixel-to-area algorithm. Zhang,
Kong, You and Wong [6] have proposed Online
Palmprint Identification. The proposed system
takes online palmprints, and uses low resolution
images. Low pass filter and boundary tracking
algorithm is used in preprocessing phase. Circular
Gabor filter used for feature extraction and 2-D
Gabor phase coding is used for feature
representation. A normalized hamming distance is
applied for matching. J. You, W. Kong, D. Zhang,
and K. Cheung[7] proposed a dynamic selection
scheme by introducing global texture feature
measurement and the detection of local interesting
points. Our comparative study of palmprint feature
extraction shows that palmprint patterns can be
well described by textures, and the texture energy
measurement possesses a large variance between
different classes while retaining high compactness
within the class. The coarse-level classification by
global texture features is effective and essential to
reduce the number of samples for further
processing at fine level. The guided searching for
the best matching based on interesting points
improves the system efficiency further. W. Li, J.
You, and D. Zhang[8], have proposed an effective
indexing and searching scheme for an image
database to facilitate fast retrieval when the size of
a palmprint database is large. There are three key
issues to be considered: feature extraction,
indexing, and matching. In general, in an image
database, the extracted features are often associated
to the original images as indices. A search for the
best matching is conducted in a layered fashion,
where one feature is first selected to lead the search
by reducing the set of candidates. Then other
features are used to reduce the candidate set
further. Such a process will be repeated until the
final output is determined based on the given
matching criteria. The selection of features plays an
important role for efficient search. An effective
feature selection scheme should exclude the most
impossible candidates, compare easily, require
small size of space for storage. Prasad, Govindan
and Sathidevi[9], have proposed Palmprint

Authentication Using Fusion of Wavelet Based
Representations. Features extracted are Texture
feature and line features. In proposed system pre-
processing includes low pass filtering,
segmentation, location of invariant points, and
alignment and extraction of ROI. OWE used for
feature extraction. The match scores are generated
for texture and line features individually and in
combined modes. Weighted sum rule and product
rule is used for score level matching. Cappelli,
Ferrara, and Maio[9] proposed high resolution
palmprint recognition system which is based on
minutiae extraction. Pre-processing is formed by
segmentation of an image from its background. To
enhance the quality of image, local frequencies and
local orientations are estimated. Local orientation is
estimated using fingerprint orientation extraction
approach and local frequencies are estimated by
counting the number of pixels between two
consecutive peaks of gray level along the direction
normal to local ridge orientation. Minutiae feature
is extracted in feature extraction phase. To extract
the minutiae features contextual filtering with
Gabor filters approach is applied. Minutiae cylinder
code has been used for matching the minutiae
features.A. Gyaourova and A. Ross[10] have
proposed an indexing technique that can either
employ the biometric matcher that is already
present in the biometric system or use another
independent matcher. Index codes are generated for
each modality using the corresponding matcher.
During retrieval, the index code of the probe is
compared against those in the gallery using a
similarity measure to retrieve a list of candidate
identities for biometric matching. The proposed
indexing technique on a chimeric multimodal
database resulted in a reduction of the search space
by an average of 84% at a 100% hit rate. The main
factor for the amount of speedup during
identification was the penetration rate of the
indexing.
To overcome all the drawbacks of above
works developed by many authors, here we
supposed to introduce a highly secured biometric
authentication system with palm print using
UDBW transform and Morphological ROI
extraction.
III. PROPOSED METHOD
Here in this section, we described the
proposed palm print authentication model using
hybrid process and UDBWtransform. Fig shows
that the proposed model for palm print
authentication, in which we had three modules:
a. Registration process
b. Testing and
c. Palm matching
3.1. Registration
In this module input palm image will be
registered by applying region of interest with
morphological operation there by calculate the
distance transform and then extracting the low level
features using 3-level UDBW transform. After
getting the UDBW coefficients, statistical
computation will be done by taking the mean and
variance of the decomposed coefficients. Then all
the statistics will be stored in a vector to make a
train feature vector.
3.1.1. Morphological Operation
Binary images may contain numerous
imperfections. In particular, the binary regions
produced by simple thresholding are distorted by
noise and texture. Morphological image processing
pursues the goals of removing these imperfections
by accounting for the form and structure of the
image.
Fig1. Flow chart of proposed palm print
authentication system
3.1.2. ROI extraction
Region of interest is a selected samples
subset within a dataset distinguished for a
particular purpose.This can be used in many
applications such as medical imaging, the tumor
boundaries may be defined on an MR or CT image
for measuring of its size. The endocardial border
may be defined on an image, perhaps during
different phases of the cardiac cycle, for example
end-systole and end-diastole, for the purpose of
assessing cardiac function. In geographical
information systems (GIS), a ROI can be taken
literally as a polygonal selection from a 2D map.
In computer vision and optical character
recognition, the ROI defines the borders of an
object under consideration.

3.1.3. Distance Transform
The distance transform is an operator
which can only be applied to binary images. It
results in a gray level image which looks like same
as input image, except that the gray level intensities
of points inside foreground regions are changed to
show the distance to the closest boundary from
each point.
Fig2. Example of distance transform with
chessboard metric
3.1.4. UDBW Transform
Un-decimated biorthogonal transform is
well used for multi resolution analysis due to its
multi scaling functionality i.e., two scaling
functions to generate wavelet filter banks for
decomposition and reconstruction separately. It
will give more effective decomposition coefficients
due to its multi scaling property.
In the case of orthogonal, we have one
hierarchy of approximation spaces 𝑉𝑗−1 ⊂ 𝑉𝑗 ⊂
𝑉𝑗+1 and an orthogonal decomposition
𝑉𝑗+1 = 𝑉𝑗 ⊕ 𝑊𝑗 (1)
which leads us to use two filter sequences
𝑕 𝑛 and 𝑔 𝑛 for decomposition and reconstruction.
Hence, we need to construct two different wavelet
functions and two different scaling functions.
Let𝑓𝑘,𝑔 𝑘 ∈ 𝐻.𝑖𝑓 𝑓𝑗 , 𝑔 𝑘 = 𝛿𝑗𝑘 Then we will say that
the two sequences are biorthogonal.
Now, our aim is to build two sets of wavelets
𝜓𝑗 ,𝑘 = 2
𝑗
2 𝜓 2𝑗
𝑥 − 𝑘 (2)
𝜓𝑗,𝑘 = 2
𝑗
2 𝜓2𝑗
𝑥 − 𝑘 (3)
To do so, we need four filters𝑔, 𝑕, 𝑔, 𝑕 i.e.,
two sequences to be act as decomposition
sequences and two sequences as reconstruction
sequences. For example, if 𝑐 𝑛
1
is a data set, it will be
decomposed as follows:
𝑐 𝑛
0
= 𝑕2𝑛−𝑘 𝑐 𝑘
1
𝑘 (4)
𝑑 𝑛
0
= 𝑔2𝑛−𝑘 𝑐 𝑘
1
𝑘 (5)
And the reconstruction is given by
𝑐𝑙
1
= 𝑕2𝑛−𝑙 𝑐 𝑛
0
+ 𝑔2𝑛−𝑙 𝑑 𝑛
0
𝑛 (6)
We can achieve perfect reconstruction by following
some conditions given below:
𝑔 𝑛 = −1 𝑛+1
𝑕−𝑛 ,𝑔 𝑛 = −1 𝑛+1
𝑕 𝑛
𝑕 𝑚 𝑕 𝑛+2𝑘 = 𝛿 𝑘0
𝑛
Now consider that 𝜙 𝑥 and 𝜙 𝑥 are two
scaling function with their own hierarchy of
approximation spaces, then we will generate
function of wavelet in a method of analogous to the
orthogonal case. We now define the scaling
function as follows:
𝜙 𝑥 = 2 𝑕 𝑛 𝜙 2𝑥 − 𝑛𝑛𝑛 (7)
𝜙 𝑥 = 2 𝑕 𝑛 𝜙 2𝑥 − 𝑛𝑛 (8)
So, finally the bi-orthogonal wavelet functions can
be defined as follows:
𝜓 𝑥 = 2 𝑔 𝑛 𝜙 2𝑥 − 𝑛𝑛 (9)
𝜓 𝑥 = 2 𝑔 𝑛+1 𝜙 2𝑥 − 𝑛𝑛 (10)
3.2. Testing
The second module in the proposed
system is testing process which includes that the
database palm image will be selected for testing
with the registered palm image by applying
morphological processing; ROI extraction, distance
transform and UDBW transform there by
calculating the statistics to get the test feature
vector.
3.3. Matching Process
In this step, Euclidean distance will be
calculated between both the feature vectors i.e.,
train and test to obtain the most matched image that
is stored in database to found that whether
authorized person’s identification is available or
not. If the distance is zero then the person will be
identified otherwise it displays that the match not
found.
IV. SIMULATION RESULTS
Experimental results have been done in
MATLAB 2014a version with various palm images
by using proposed palm print identification model
with high security. We achieved 100% accuracy
and more efficiency with the proposed model. Fig1
shows that the original palm image for registration
process described in section 3.1, 3 (a) shows the
original palm image, 3(b) shows it’s binary image
with morphing, 3(c) shows that the distance
transformed image of a binary image and finally 3
(d) is a registered palm print for authenticating a
person for authorization into a particular task.
(a) (b)

(c) (d)
Fig3. (a) original palm image for
registration (b) morphed image (c) distance
transformed image and (d) registered palm image
Fig4. Message box for saving the registered palm
filename with mahe7
Fig5. distance transform of a test image
Fig6. Registered palm print of a test image
Fig7. Message box displayed after completion of
test and matching process
(a) (b)
(c) (d)
Fig.8un saved file from data base (a)
binary image (b) distance transform (c) registered
palm print and (d) message box after testing with
data base files
(a)
(b)
Fig.9 screen shots of test image 4.jpg which has
been saved with a specific file name in database

Fig.10 screen shots of test image 2.jpg which has
not saved with a specific file name
V. CONCLUSIONS
Here, we introduced a novel and highly
secured biometric authentication model with palm
print identification system using morphological
ROI extraction with distance transform and un-
decimated biorthogonal wavelet transform. Due to
its multi scaling functionality, two different
wavelet filter banks will be used to extract the
features of distance transformed image to obtain
the most effective feature factor for comparing with
a test feature vector. The proposed model has
proven that it has achieved 100% accuracy with
several test images from the database.
REFERENCES
[1] D. Zhang, W. K. Kong, J. You, M. Wong,
“Online Palmprint Identification,” IEEE
Transaction on Pattern Analysis and
Machine Intelligence, Vol.25, No. 9, pp.
0162-8828, Sept 2003.
[2] G. Lu, D. Zhang, K. Wang, Palm print
recognition using eigenpalms features,
Pattern Recognition Letters 24 (9) (2003)
1463–1467.
[3] J. Dai, J. Feng, J. Zhou, “Robust and
Efficient Ridge-Based Palmprint
Matching,” IEEE Transaction on Pattern
Analysis and Machine Intelligence,
Vol.34, No. 8, pp. 0162-8828, August
2012
Kong and D. Zhang, “Competitive coding
scheme for palmprint verification,” in
Proc ICPR, 2011,pp. 520–523.
[4] Huang,W. Jia, D. Zhang, “Palmprint
verification based on robust line
orientation code,” Pattern Recognition,
Science Direct, pp. 1504 – 1513, 2008.
[5] Huang,W. Jia, D. Zhang, “Palmprint
verification based on principal lines,”
Pattern Recognition, Science Direct,
pp.1316 – 1328, 2008.
[6] J. You, W. Kong, D. Zhang, and K.
Cheung, “On hierarchical palmprint
coding with multiple features for personal
identification in large databases,” IEEE
Trans. Circuits Syst. Video Technol., vol.
14, no. 2,pp. 234–243, Feb. 2009.
[7] W. Li, J. You, and D. Zhang, “Texture-
based palmprint retrieval using a layered
search scheme for personal identification,”
IEEE Trans. Multimedia, vol. 7, no. 5, pp.
891–898, Oct. 2009.
[8] S. M. Prasad, V. K. Govindan , P. S.
Sathidevi, “Palmprint Authentication
Using Fusion of Wavelet Based
Representations,” IEEE, pp. 978-1-4244-
5612-3, 2009.
[9] R. Cappelli, M. Ferrara, and D. Maio, “A
Fast and Accurate Palmprint Recognition
System Based on Minutiae,” IEEE
Transaction on System, Man and
Cybernetics- Part B: Cybernetics, Vol. 42,
No. 3, pp. 1083-4419, June 2012.
[10] Gyaourova and A. Ross, “Index codes for
multibiometric pattern retrieval,” IEEE
Trans. Inf. Forensics Security, vol. 7, no.
2, pp. 518–529, Apr. 2012.
Raja Murali Prasad received
his bachelor’s degree from
the Institution of Engineers
in 1989 and M. Tech in
ECE from Pondicherry
Engineering College in
1993. He worked in various
engineering colleges as
faculty member. Presently,
he is working as faculty member in the
Department of Electronics and Communication
Engineering, Vardhaman College of
Engineering, Hyderabad. He completed PhD at
JNT University Anantapur. His research
interests include digital communications,
control systems and wireless communications.

Highly Secured Bio-Metric Authentication Model with Palm Print Identification

More Related Content

What's hot (20)

Viewers also liked (19)

Similar to Highly Secured Bio-Metric Authentication Model with Palm Print Identification (20)

Recently uploaded (20)

Highly Secured Bio-Metric Authentication Model with Palm Print Identification