2020 International Conference on Emerging Smart Computing and Informatics (ESCI)

AISSMS Institute of Information Technology, Pune, India. Mar 12-14, 2020

Multimodal Biometric Authentication System using

Deep Learning Method
Sandeep Singh Sengar U. Hariharan K. Rajkumar
Department of Computer Science and Department of Information Technology Department of Information Technology
Engineering Galgotias College of Engineering and Galgotias College of Engineering and
SRM University-AP, Amaravati Technology Technology
Andhra Pradesh, INDIA Greater Noida, Uttar Pradesh, India Greater Noida, Uttar Pradesh, India
sandeep.iitdhanbad@gmail.com *udhayakumarhariharan@gmail.com rajonline7@gmail.com

Abstract—For specific identification process, Identity fingerprints. A biometric confirmation structure works within
Management details an ailment of supplying authorized 2 modes, for instance, enlistment function as well as examine
owners with secure and easy admittance to information and function. Inside enlistment function, the device recognizes an
solutions. For choosing the individual's identity, the primary individual by looking the layouts of significant amount of
goal is actually executing secured identification feature. PINs, clientele within the data source for a match up. Within the
keys, gain access to cards, passwords, tokens are actually the confirmation function, structure approves character of
private determining elements which are actually utilized within specific by evaluating the found biometric details together
standard methods which may have a tendency to drawbacks with the possession of biometric template(s) that are stored
such as cracking, stealing, copying and posting. Biometrics
method repository.
grounded identification is needed having a perspective to stay
away from the drawbacks. Due to intra category variants, non- Biometric methods that rely on the evidence associated
universality, sound as well as spoof strikes are impacted. with an one particular wellspring of information for
Multimodal biometrics are actually employed to get rid of the validation are actually labelled as Unimodal method [4].
episodes which are actually a grouping of countless modalities. Unimodal biometric methods endure an assortment of
For an authentication supply, Fingerprint and Palmprint problems, for instance, Commotion within detecting
identification are popular systems these days. For minutiae information, Intra class variations, Uniqueness capability.
thing detection as well as attribute extraction, with this paper,
The confinements of unimodal biometric structure lead to
rich neural community (DNN) were definitely projected. The
substantial False Acceptance Rate (FAR) along with False
confinements of unimodal biometric structure lead to
substantial False Acceptance Rate (FAR) along with False Rejection Rate (FRR), limited splitting up skill, top bound
Rejection Rate (FRR), limited splitting up skill, top bound within delivery therefore the multimodal biometric product is
within delivery therefore the multimodal biometric product is designed to satisfy the strict delivery demands. A biometric
designed to satisfy the strict delivery demands. For minutiae os and that is determined by the nearness of several pieces of
corresponding, values of Euclidean distance are actually used. evidence of specific identification is known as multimodal
The better identification pace is actually attained throughout biometric structure [5]. A biometric os and that is determined
the suggested procedure & it's extremely safe only in loud by the nearness of several pieces of evidence of specific
problem. identification is known as multimodal biometric phone
system. It fuses some variety of autonomous biometrics as
Keywords—DNN, Biometrics, FAR, FRR, 2D Gabor Filter, well as overcomes a percentage on the restrictions created by
Deep Neural Network, Euclidean distance. making use of Unimodal biometric. Multimodal biometrics
are progressively important to phony advances, since it's a lot
I. INTRODUCTION more difficult to generate numerous biometric characteristics
The zone of unique identification is actually mistreating when compared with fashioning an one particular biometric
computer aided devices as a far more protected also trademark hence provides larger accuracy fee as well as from
progressively impressive way as well as biometrics is spoofing steps by which makes it troublesome for an
actually a standout amidst most reliable characteristics which interloper to simultaneously spoof the a number of biometric
could be used for specific recognition. For conventional attributes of a real customer [6]. The delivery markers for
identification method passwords [1,2] ID cards are already specific recognition feature include things like False
used to moderate permission to access restricted methods. Be Acceptance Rate (FAR) along with False Rejection Rate
this since it might, protection could be broken off when an (FRR).
unapproved prospect or maybe fake endeavours to abuse
them. Inconvenience with making use of conventional False Acceptance Rate (FAR): When the biometric
solutions triggered fast increment from the usage of telephone system improperly accepts a gain access to try of
biometrics. Biometrics are actually an instant identification an unauthorized pc user is actually recognized as False
which alludes to a searching of decomposing, estimating as Acceptance Rate (FAR). A system's FAR says when the ratio
well as evaluating an individual's persona influenced by the involving the variety of phony acceptances as well as variety
social or physiological characteristics associated together of identification tries.
with the prospect [3]. You will find two types of biometrics ‫ܣܨ‬
to become Multimodal, Unimodal, and specific. The typical ‫ܴܣ݁ݏ݈ܽܨ‬ሺΨሻ ൌ  ൬ ൰ ‫ͲͲͳ כ‬ሺͳሻ
ܰ
area biometric capabilities are palm prints, hand geometry, Wherever FA - Lots of incidents of phony acceptances, N
discourse designs, iris designs, facial features, and - Total amount of samples.

978-1-7281-5263-9/20/$31.00 ©2020 IEEE 309

Authorized licensed use limited to: Cornell University Library. Downloaded on August 19,2020 at 04:33:25 UTC from IEEE Xplore. Restrictions apply.
 False Rejection Rate (FRR): When the biometric carried out to split up the parallel picture in w *w squares
telephone system improperly rejects a gain access to try of an [8]. Fig. 2 shows the information stream chart for picture
authorized pc user is actually recognized as False Rejection securing.
Rate (FRR).A system's FRR states since the ratio in between
quantity of phony acceptances as well as variety of
identification tries.
‫ܴܨ‬
‫ܴܴܨ‬ሺΨሻ ൌ  ൬ ൰ ‫ͲͲͳ כ‬ሺʹሻ
ܰ
FR - Lots of incidents of phony acceptances. N - Total
amount of samples. Therefore, the equations (one) as well as
(two) are actually utilized for general performance analysis
within biometric methods.
With this paper, fingerprint or palm print documents
biometrics are actually utilized for multimodal biometric
phone system. The characteristics are actually extracted
making use of the distinction algorithm Deep Neural
Networks (DNN). Minutiae based mostly extraction as well
as detection is completed by using Multilayer Perceptron
(MLP) as well as Backpropagation Algorithm.
II. PROPOSED MODEL Fig. 2. Flow Sequence of Image Acquisition
The suggested method depends upon Multimodal
biometrics whereby all fingerprint as well as palmprint B. Feature Extraction
biometric is actually used like a wellspring of authentication. Extraction of appropriate feature of the brother ql is
Fig. 1 suggests multimodal biometrics of palmprint as well easily the most crucial task. DNN is actually used for
as fingerprint. They're viewed as essentially the most element extraction. It's a distinction algorithm which is made
popular, reliable as well as top biometrics. Automated up of big amount of neurons (units) interconnected as levels.
Fingerprint Identification System (AFIS) makes use of Information is delivered only one means and possesses
methods generally determined by specifics concentrates. absolutely no cycles or even sectors. The information
concentrates are actually divided by DNN show in Fig. 3
Palmprint has edges, central lines, wrinkles, for example, and
coming from the edges, valleys, wrinkles [14]. Deep-learning
features by what specifics concentrates could be extricated networks and rule lines are actually realized using the much
[7]. Because of abundant floor information fingerprint as more normal single concealed level neural networks by the
well as palmprint are believed to be as incredible techniques profundity of theirs; that's, the volume of hub levels whereby
to come down with around house identification. Artificial info needs to go within a multistep process of instance
Neural Networks acknowledgment. Prior versions of neural networks, for
(ANN), one of several Soft Computing methods [8] used for instance, the very first perceptron’s had been superficial,
Multimodal biometric identification. produced from 1 info as well as one particular yield level, as
well as a maximum of 1 shrouded level within the centre.
Several levels (counting info as well as yield) qualifies as
"deep" learning [9]. Therefore deeply is not merely a
fashionable phrase to affect computations to show up as
although they come across Tune and also Sartre directly into
organizations you have not understood around just yet. It's a
meticulously recognized phrase which suggests greater than
a person shrouded level.
Within deep learning networks, every level of nodes
preps on an unmistakable plan of spotlights determined by
history layer's yield [15]. The additional you progress
Fig. 1. Multimodal Biometric straight into the neural web, the greater perplexing the
spotlights the nodes of yours are able to perceive, as they
A. Image Acquisition total as well as recombine spotlights as a result of history
The info impression received is actually for starters coating.
standardized to reduce variations to come down with dark C. Minutiae Detection
dimension esteems on the tips as well as valleys. Histogram
feature is actually carried out to enhance the clearness of this Multilayer Perceptron (MLP) strategy is used to
distinguish the specifics, regardless of whether it's on
picture [15]. Binarization is actually carried out to alter 8 bit
advantage department or even edge finishing. It contains
dim scope picture directly into twofold picture with zero
several levels, for instance, feedback, covered up as well as
share for tips as well as one share for valleys. Advantage yield levels [12]. The networking is actually divided using
segmentation is completed to determine advantage finishing, Back propagation algorithm that is cooking or even
edge department as well as distinguish return on first capital mastering algorithm in which a community process is
expense. Within that time, the launch treatment is actually actually done by isolating the picture directly into 3*3

310

Authorized licensed use limited to: Cornell University Library. Downloaded on August 19,2020 at 04:33:25 UTC from IEEE Xplore. Restrictions apply.
windowpane square measurement, if the windowpane is
actually placed on specifics the self-esteem is actually one or
even different zero. If perhaps within pixel worth is actually
one and possesses simply 1 one self-esteem neighbor, then
simply particulars factor is actually on advantage completing
[11]. On the away possibility which center pixel worth is
actually one and possesses simply 3 one self-esteem
neighbor, then specifics factor is actually on advantage
department. With lengthy final, Euclidean splitting up is
actually discovered in between specifics concentrates and set
away within a repository.

Fig. 4. System Architecture

IV. RESULTS AND DISCUSSIONS

Making use of Matlab, the pictures of all fingerprint as
well as palmprint are actually selected and set away within
an MS - Access repository. At this point amid the
identification process, the pictures are actually packed for
picture enhancement in which standardization, histogram
getting used to as well as advantage launch are actually
performed [10]. With the much better photos, the specifics
concentration are actually divided. Within the first coat level
Fig. 3. A Sample DNN the bifurcation as well as advantage endings are actually
realized by executing community process. Euclidean
D. Minutiae Matching splitting up computation is actually performed [13]. The
Specifics coordinating are actually completed by information esteems coordinating are actually completed by
evaluating the specifics esteems place at bay within data computing the scores as well as validation is actually
source using the specified info picture. Specifics scores are examined. The delivery of Gabor channel switches for both
actually produced having a stove someplace inside all the real component as well as fanciful component as much as
different zero as well as one. If perhaps the coordinating FAR and also FRR influenced by the amount of assessments.
score is much more noteworthy compared to predefined cap The variations are showing up Table I as well as Table II.
worth, the prospect is actually accredited mostly is a faker The delivery on the suggested means with different
[5]. limitations, Tar, and they limit the phony accept pace as well
as bogus refuse pace is actually been seen [3]. Therefore, the
III. SYSTEM IMPLEMENTATION suggested strategy grounded techniques to enhance
distinction exactness involving frauds and also genuine
The structure consists of 2 phase forms: Identification clientele the prediction of FAR and FRR shown in Table III.
and also Enlistment. Fig. 4 suggests System Design for
Particulars Coordinating Based Individual Identification TABLE I. RELIABILITY MEASURES FOR GABOR REAL PART
System. Amid enlistment procedure, the prospect enlists
equally palmprint as well as fingerprint as character. The No. of Samples FRR FAR ACCURACY
functions, for instance, tips, valleys, main collections, lines
and wrinkles are actually extricated of specifics concentrates 25 1.10 3.0 91.00
are actually acquired. With, specifics recognition, the
information concentrates are actually divided as end as well 125 0.70 2.3 96.23
as bifurcation (edge endings) concentrates [10,11]. At this
600 0.02 1.3 97.60
point together with the information concentrates, the
Euclidean splitting up esteems are actually computed and set
away inside repository. Amid identification process, TABLE II. RELIABILITY MEASURES FOR GABOR IMAGINARY
fingerprint as well as palmprint are actually procured for PART
identification. Similar process occurs till particulars
No. of Samples FRR FAR ACCURACY
recognition. With specifics coordinating, the coordinating
score esteems are actually computed by evaluating the 25 1.10 2.0 91.00
specifics contained in the specified information and also the
camera put away inside repository. The option is performed 125 0.70 1.4 96.23
influenced by the coordinating score. On the away possibility
which coordinating score tends to be more noteworthy 600 0.02 01.4 97.60
compared to the advantage worth, the prospect is actually
accredited as well as recognized. Another thing, prospect is
actually unapproved as well as rejected.

311

Authorized licensed use limited to: Cornell University Library. Downloaded on August 19,2020 at 04:33:25 UTC from IEEE Xplore. Restrictions apply.
 TABLE III. TEST RESULTS ON DESIGNATED THRESHOLD [3] A.K. Jain, A. Ross & S. Prabhakar, “An Introduction to Biometric
STANDARDS Recognition”, In: IEEE Transactions on Circuits and Systems for
Video Technology, Vol. 14, N. 1, pp 4-20, January 2004.
Threshold Tar False Accept Rate(%) False Reject Rate(%) [4] Sengar, S.S. and Mukhopadhyay, S. (2017). Moving object detection
based on frame difference and W4. Signal, Image and Video
0.325 0.00 1.46 Processing, 11(7), pp.1357-1364.
[5] Mir. A.H, Rubab .S, Jhat .Z.A,” Biometric Verification: A Literature
Survey “, International Journal of Computing & ICT Research,
0.335 0.00 0.81 Vol.05, No.02, December 2011.
[6] Fawaz Alsaade,” Neuro-Fuzzy Logic Decision in a Multimodal
0.345 0.27 0.55 Biometric Fusion System”, Scientific Journal of King Faisal
University (Basic and Applied Sciences), Vol.11, No.02, 2010.
0.355 0.82 0.55 [7] MuthuKumar .A, Kasturi .C and Kannan .S, ”Multimodal Biometric
Authentication using Particle Swarm Optimization Algorithm with
Fingerprint and Iris”, ICTACT Journal on Image and Video
0.365 2.40 0.29 Processing, Vol.02, No.03, Feburary 2012.
[8] Sengar, S.S. and Mukhopadhyay, S. (2016). Moving object area
0.375 5.26 0.11 detection using normalized self adaptive optical flow. Optik, 127(16),
pp.6258-6267.
[9] K.Shanmugapriyal, M.Karthika, Dr.S.Valarmathy, M.Arunkumar
V. CONCLUSION “Performance Evaluation of Contourlet Transform based Palmprint
Recognition using Nearest Neighbour Classifier” International
Biometrics assumes a crucial task within around house Journal of Emerging Technology and Advanced Engineering
identification. The goal is actually actualizing a secured Website: www.ijetae.com (ISSN 2250-2459, ISO 9001:2008 Certified
unique identification device continues to be dependably Journal, Volume 3, Issue 1, January 2013.
achieved by techniques for making use of Multimodal [10] Sengar, S.S., Mukhopadhyay, S. (2017). Foreground Detection via
Biometrics. Fingerprint as well as Palmprint that are wealthy Background Subtraction and Improved Three-Frame
within floor information achieves top exactness. DNN Differencing, Arabian Journal for Science and Engineering, vol. 42,
pp. 3621.
grounded distinction becomes excellent identification fee.
[11] SitalakshmiVenkataraman,”A Grid-Based Neural Network
Together the collections, the recommended technique betters Framework for Multimodal Biometrics”, World Academy of Science ,
noteworthy delivery as well as distinction accuracy within Engineering and Technology,2010.
factors that are fresh . The results provide far better accuracy [12] Murugan, B.S., Elhoseny, M., Shankar, K. and Uthayakumar, J.,
for your offered approach. 2019. Region-based scalable smart system for anomaly detection in
pedestrian walkways. Computers & Electrical Engineering, 75,
REFERENCES pp.146-160.
[1] Shankar, K., Elhoseny, M., Kumar, R.S., Lakshmanaprabu, S.K. and [13] Sengar, S. (2017). Detection of moving objects based on enhancement
Yuan, X., 2018. Secret image sharing scheme with encrypted shadow of optical flow. Science Direct, Vol. 145, pp.131-141.
images using optimal homomorphic encryption technique. Journal of [14] Uthayakumar, J., Metawa, N., Shankar, K. and Lakshmanaprabu,
Ambient Intelligence and Humanized Computing, pp.1-13. S.K., 2018. Financial crisis prediction model using ant colony
[2] Shankar, K., Elhoseny, M., Chelvi, E.D., Lakshmanaprabu, S.K. and optimization. International Journal of Information Management.
Wu, W., 2018. An Efficient Optimal Key Based Chaos Function for [15] Sengar, S. S., and Mukhopadhyay, S. (2017). Motion detection using
Medical Image Security. IEEE Access, 6, pp.77145-77154. block based bi-directional optical flow method. Journal of Visual
Communication and Image Representation, 49, 89-103.

312

Authorized licensed use limited to: Cornell University Library. Downloaded on August 19,2020 at 04:33:25 UTC from IEEE Xplore. Restrictions apply.