2024 Second International Conference on Inventive Computing and Informatics (ICICI)

Secured Voting System based on Multilayered

2024 Second International Conference on Inventive Computing and Informatics (ICICI) | 979-8-3503-7329-5/24/$31.00 ©2024 IEEE | DOI: 10.1109/ICICI62254.2024.00087

Biometric Authentication

Gunda Sathwik Pendayala Druva Teja Reddy *Manu Gupta

Department of Electronics and Computer Department of Electronics and Computer Department of Electronics and Computer
Engineering, Engineering, Engineering,
Sreenidhi Institute of Science and Technology, Sreenidhi Institute of Science and Technology, Sreenidhi Institute of Science and Technology,
Hyderabad, India Hyderabad, India Hyderabad, India
[email protected] [email protected] *[email protected]

Myadam Rahul Mohan Dholvan

Department of Electronics and Computer Engineering, Department of Electronics and Computer Engineering,
Sreenidhi Institute of Science and Technology, Sreenidhi Institute of Science and Technology,
Hyderabad, India Hyderabad, India
[email protected] [email protected]

Abstract— The right to vote is a crucial aspect of a system that In the current era of fast-growing technology, using
is democratic and allows people to select the leader or people manual election procedures is an old-fashioned method,
they prefer. The traditional offline voting system is time- especially since online platforms are much faster, more
consuming, costly and requires lot of manpower, making it accurate, and more effective. A nation's democracy is based
prevalent to frauds and human errors. There is a need to on an open electoral system that allows individuals to pick
upgrade and develop efficient methods for voting ensuring their representatives. The current manual voting system in
system authenticity and voter verification. In the proposed work India overloads voters and the government, and an online
a multi-layered approach based voting system using biometric counterpart is urgently required [4]. The idea is to use an
features is developed. The voter verification is performed using online voting system to facilitate a rapid process of
facial recognition, thumbprint and Aadhaar number validation. registration, casting, counting, and the announcement of
The use of unique biometric features and user details, results. In the proposed work a secured voting system with
strengthen voting system security and ensure authenticity while facial recognition, and thumb detection is developed.
decreasing the chances of fraudulent activities during the voting Additionally, Aadhaar card verification for voters is also
process. A web application through three-step verification using
included, which increases system robustness and lowers
biometric features and Aadhaar card is developed that
facilitates a user-friendly and cost-effective voting environment
sensitivity to fraud and duplicate votes. A comprehensive
under administrative guidance. Furthermore, this work creates three-step verification system to address prevalent challenges
a secured model by eliminating fraudulent or duplicate votes in traditional voting methods, including accessibility issues,
cast during elections and also displays user votes cast so they security vulnerabilities, and logistical constraints is devised.
may be verified more efficiently for faster counting procedures Employing cutting-edge biometric technologies and modern
allowing admin to release results more promptly. web development methodologies, the proposed work
integrates facial recognition, thumb detection and Aadhaar
card number authentication to ensure robust authentication
Keywords— Voting system, Aadhaar Number, Face protocols. By providing voters with a multi-layered
Recognition, Thumb detection, Biometric.
verification process, the proposed method aims to promote the
voting system's security and integrity.

I. INTRODUCTION The proposed model is primarily concerned with putting

in place a multi-tiered authentication system to guarantee
Elections are the cornerstone of democracy and can serve increased security and integrity. This method lowers the
to promote it if conducted fairly and transparently; possibility of unwanted access or tampering by adding
unfortunately, however, the current method for conducting numerous stages of verification, which increases the voting
elections contains flaws that make democracy harder for many system's dependability [3]. Firstly, a web application for
of its constituents to access their government representatives providing a voting platform is designed using Django thereby
directly [1]. Current election systems contain numerous flaws improving its accessibility and scalability. It comprises of
that allow contestants and political parties to exploit it to their admin page which handles all data and maintains security and
benefit, including duplicate votes, rigging electronic voting a user page for voters to register. Facial recognition for voters
machines (EVMs) and falsifying results [2]. Electronic voter is performed using image processing through OpenCV, which
machines for implementing elections are time and energy- provides extensive libraries and tools for computer vision and
consuming and must take place in designated places at certain image processing applications. The use of OpenCV for facial
intervals. This process also needs extra manpower for recognition improves the voting system's overall efficacy and
transport and supervision purposes and may not generate integrity while guaranteeing dependable authentication
accurate results[3].

979-8-3503-7329-5/24/$31.00 ©2024 IEEE 502

DOI 10.1109/ICICI62254.2024.00087
Authorized licensed use limited to: VTU Consortium. Downloaded on October 23,2024 at 09:38:44 UTC from IEEE Xplore. Restrictions apply.
procedures for registered voters. Optical Character and vote counting process [6]. Providing access to mobile
Recognition (OCR) Method [5] is used for Aadhaar card phone to voters during the voting procedure for accessing
verification. The users are first asked to register in the web OTP is a risk to election security. Additionally, allowing the
application with their face, thumbprints and Aadhaar card admin to access the result section using an OTP introduces a
details. At the time of voting, the users have to log in using the potential point of failure or manipulation. If the admin
username and password provided at the time of registration. credentials are compromised or misused, it could undermine
Then multilayered verification is performed by comparing the the integrity of the voting process. In the proposed system
current biometrics captured. If the data matches the user is these limitations are addressed by developing a secure voting
allowed to vote and choose the candidates. The entire system system using a multilayered biometric authentication process.
is based on a web application and controlled by the admin.
This process is more secure, inexpensive and consumes less III. PROPOSED MODEL
manpower. The proposed multilayer secure voting system using
biometric analysis incorporates a three-step authentication
II. RELATED WORKS process for ensuring voter legitimacy. Firstly, the web
There are many works in literature based on developing a application is developed with Admin and user platform. The
voting system using advanced technological approaches. admin will add party details and voter details. The admin
Sirenjeevi et al. [6] developed a voting system using deep generates the database for voters by making users register
learning. The face authentication is made using the through the user login page. The user has to register and
Convolutional Neural Network (CNN). This system assumes provide the thumb impression and face images using a
a two-level authentication technique. Initially, the registered webcam and an Aadhaar card. The generated database by the
user information is published in the login page and the user is admin is then used for voter verification at the time of voting.
validated based on both face authentication and One Time This comprehensive approach significantly mitigates risks
Password (OTP). In the next phase, the vote count also associated with fraudulent voting activities, bolstering safety
requires proper verification. At this time, the admin is asked and integrity in voting. The flow of the proposed system from
to produce the OTP and then the admin is logged in to the the Admin web page is depicted in Fig. 1.
result section. Abimanyu et al. [7] developed Iris-based
Authentication for secure voting. With the help of Daugman’s
algorithm, a user's Iris is scanned and checked with the image
stored in the database. After verification, the user is allowed
to cast the vote.
Govindaraj and Kumaresan [8] developed a voting system
using the cloud. The software was developed using the
Microsoft SQL Server 2012 as a database platform, C#
programming language and the Microsoft Azure cloud. The
major advantage is this system is the user can cast a vote from
anywhere. Rajesh and Priyanka [9] developed an online
voting system using fingerprint detection. The Minutiae
Extraction Algorithm is used for fingerprint matching.
Gangadurai et al.[10] proposed a fingerprint-based voting
system. Aadhaar card data is used for authentication of voter
information. Nalayini et al. [11] presented a mobile voting
application using fingerprints for voter identity authentication.
Python framework is used to implement the application. The
existing systems have the limitation that one person can vote
more than once as admin control is not provided. Users need
mobile phones and an internet facility has to be available to
perform voting. However, in the proposed method the web
application is developed and users have to register in the
system generating a username and password. The database is
maintained by the admin and used for multi-layered Fig. 1. Working stages of the system from the admin page
verification through face, thumb impression and Aadhaar card
data authentication at the time of elections when users have to
cast votes. The Major components of the proposed system from
the admin page are detailed below:
In the existing systems, the models based on iris detection
i. Admin Login page: Using the username and
[7] and fingerprint-based e-voting systems [9] include
password the admin can access the system.
potential issues with accuracy, privacy concerns, and the need
for specialized hardware, which may limit accessibility and ii. Add Party Details: After login, the admin will add
increase costs. Additionally, factors such as physical candidates and party details such as party symbol
disabilities, eye conditions, or simply lack of access to the images.
required technology could hinder certain demographics from iii. Add Voter with Face: The admin will add voter
exercising their voting rights.Cloud based model presented in details by taking input such as Aadhaar card
[8] makes the voting process easy but have limitations of not number, thumb image and face image captured
including any authentication process to verify the users using the webcam. All the user details are then saved
identity. Some of existing models used OTP during the voting

503

Authorized licensed use limited to: VTU Consortium. Downloaded on October 23,2024 at 09:38:44 UTC from IEEE Xplore. Restrictions apply.
 in the database. Face recognition is performed using IV. TECHNIQUES USED
the OpenCV library in Python. Various techniques used in the proposed model for
iv. View Party Details: Admin can view all available developing the proposed model are explained in this section.
party details at any time. Fig. 3 demonstrates the complete mechanism of proposed
v. View Votes: The Admin can view all vote counts secured voting.
received by candidates.
The flow of the proposed System from the Voter (or)
user login page is depicted in Fig. 2. The major components
of the proposed system from the user login page are detailed
below:
i. User Login Page: This is where users start the
process of logging into an application or system.
ii. Username: To access the system, the user must
enter their name which is already given by the
admin.
iii. Password: The user then inputs their password,
which is a private word or code that only they know.
The user’s identity is verified at this phase.
iv. Aadhaar Number: Here the user has to scan their
Aadhaar card the system will scan the Aadhaar
number and verify it with the number given by the
admin. Fig. 3. Mechanism of proposed system
v. Fingerprint Image: The user is asked to upload
their fingerprint images. This picture is utilized for A. MySQL for Database Management: In the proposed
biometric authentication, in which the device work MySQL is used for the collection of the details of the
confirms the user's identification through the user (voter) casting the vote and also for adding the party
distinctive pattern of their fingerprints. details who are contesting in the elections. MySQL is an
vi. Face-verification (Face Recognition): Lastly, the open-source relational database management system
user's face image is captured by the system using a (RDBMS) that is widely used for building web-based
webcam. The user's identification is then confirmed applications and managing data. The MySQL command line
by utilizing facial recognition, a technique that is used for the execution of the SQL code. MySQL 5.5
analyzes the user's facial traits using the image. command line interface allows users to interact with the
MySQL database system directly by typing commands.
B. Django, HTML and CSS for Web Application
development: The web application in the proposed work is
developed using the Django framework with the front end
created using HTML, and CSS. Only Admin can add
candidates and positions for the elections using the Django
inbuilt Admin module. Django is a high-level Python web
framework that encourages rapid development and clean,
pragmatic design. HTML stands for Hyper Text Markup
Language. It is used to design the front-end portion of web
pages using markup language. It acts as a skeleton for a
website since it is used to make the structure of a website [12].
Cascading Style Sheets fondly referred to as CSS is a simply
designed language intended to simplify the process of making
web pages presentable. It is used to style the website.
C. OpenCV and Haar Cascade algorithm for Face
Detection: OpenCV offers extensive libraries and tools for
image processing and computer vision tasks, enabling us to
achieve superior results in facial recognition. In the proposed
work, Haar feature-based cascade classifiers from OpenCV
are used to perform facial recognition [13]. To identify the
features in the face training data is used. The images are
converted into gray scale and then features include distance
between forehead and nose etc. computed using haar filters
shown in Fig. 4. The pixel value [14] is obtained as follows:
Fig.2 Working stages of the system from User login page
ܵ‫ݏ݈݁ݔ݅ܲ݇ݎܽܦ݂݋݉ݑ‬
ܲ݅‫ ݁ݑ݈ܸ݈ܽ݁ݔ‬ൌ 
ܰ‫ݏ݈݁ݔ݅݌݇ݎ݂ܽ݀݋ݎܾ݁݉ݑ‬

504

Authorized licensed use limited to: VTU Consortium. Downloaded on October 23,2024 at 09:38:44 UTC from IEEE Xplore. Restrictions apply.
 ܵ‫ݏ݈݁ݔ݅݌ݐ݄݃݅ܮ݂݋݉ݑ‬
െ
ܰ‫ݏ݈݁ݔ݅݌ݐ݄݈݂݃݅݋ݎܾ݁݉ݑ‬



Fig.5 Flow Diagram of the OCR method

Fig. 4. Haar filters used for face recognition
E. Oriented FAST and Rotated BRIEF (ORB) Method for
Thumb Verification: ORB algorithm is utilized in proposed
The data used for training includes positive images and work for thumb verification of voters. It is a popular
negative images. Images with human face are the positive technique in computer vision for feature detection [15]. The
images, and the images where a human face is not present are voters has to submit their thumb scans during registeration
negative images.The features extracted from these training process which is then stored in database. At the time of voting
samples are the used to detect the voters face during the users thumb scans are verified by matching with database
verification process. records using ORB method. Firstly the user thumb images are
loaded and converted to grayscale while preserving essential
D. Optical Character Recognition (OCR) Method for infromation. Next, ORB is employed to detect keypoints and
Aadhaar Card Verification: OCR is used in the proposed compute descriptors for both query image of voter and
work for the detection of the Aadhaar number from the database images. Finally, a brute-force matcher is created to
Aadhaar card, shown by the user for verification before they find matches between the descriptors of the two images.
cast the vote. After scanning the Aadhaar number it validates Matches are sorted based on their distances, with closer
with the Aadhaar number which is given by the admin while matches considered more reliable. If the similarity score is
registering the user. OCR method converts scanned images same then the thumb image authentication is completed. The
or handwritten text into editable and searchable digital text. sample images in the databse for thumb scans are shown in
It works by analyzing the shapes and patterns of characters in Fig. 6.
the image, identifying them, and then translating them into
machine- readable text data. OCR methods typically involve
several steps, including image preprocessing, text detection,
character segmentation, and character recognition [5]. These
methods aim to accurately identify and convert text from
images into machine-readable formats. The flow of the OCR
Model is depicted in Fig.5.

Fig.6 Sample images in database for thumb scans

V. RESULTS AND DISCUSSION

In this section, the developed web application "eVoting”
and the results for the admin and user login are described. The

505

Authorized licensed use limited to: VTU Consortium. Downloaded on October 23,2024 at 09:38:44 UTC from IEEE Xplore. Restrictions apply.
home page of the web application is shown in Fig. 7,
comprising of Admin login and user login options. The
"Admin Login" option is illustrated in Fig. 8.

Fig.10 Admin adding party details

Fig.7 Home page of the web application

Fig.11 Candidate details added

The administrator can add party and candidate information
and an image as shown in Fig.10. After clicking "Submit,"
the page shown in Fig.11will appear, where candidate details
are added. As shown in Fig.12, the admin will enter all voter
details along with the Aadhaar number and thumb image.
Fig. 8. Admin Login Screen
When the ‘Register’ button is selected, voter details are added
and the page shown in Fig. 13 appears to capture the voter's
After logging in, the page displayed to the administrator is face.
shown in Fig. 9, where the admin will be given options to add
party details, and voter details and view the votes. The
screenshots for the same are described in Fig. 10 ,11 and 12.

Fig.12 Admin will enter all voter details

Fig. 9 Admin page view

506

Authorized licensed use limited to: VTU Consortium. Downloaded on October 23,2024 at 09:38:44 UTC from IEEE Xplore. Restrictions apply.
 Fig.13 User image on webcam for registration

The face will be captured by the webcam and displayed to the

Fig.16User with Face, Aadhaar number and thumb details added by admin
user. Next, the user will be prompted to complete the Sign-up
Task to finish adding voter information as demonstrated in
Fig.14.
The user with their thumb, Aadhaar number, and face details
entered into the database is seen in the blue text on the screen
above. Likewise, the admin can add a large number of voters
to the database. By selecting “View Party Details", the admin
can see a list of all the available party candidates as illustrated
in Fig.17.

Fig.14 Face captured and Sign up complete

If the light is dim or blurred image is taken by a webcam, at

the time of entering user details, the error message “unable to
predict” is given, as shown in Fig. 15.

Fig.17 Admin can view a list of all available candidate

The user login screen is shown in Fig. 18. It asks the user to
log in with the username, password and thumb image used at
the time of registration. Also, the user will be asked to capture
Fig.15 Face capturing through webcam failed
their face and Aadhaar card using a webcam. If the valid
details are not entered the error message as shown in Fig. 18
will appear. If any of the details are not matched while
logging to cast the vote. It will show invalid login details.

507

Authorized licensed use limited to: VTU Consortium. Downloaded on October 23,2024 at 09:38:44 UTC from IEEE Xplore. Restrictions apply.
 Fig.18 User login screen Fig.21 Face Recognition and Verification Stage

After the three-stage verification is completed the user will

be allowed to cast a vote by prompting with the "Cast Your
Vote" option. The identified Voter and the voted candidate
details are shown to the user, as illustrated in Fig. 22. The
predicted voter based on thumb verification, face recognition
and Aadhaar card verification of the voter is displayed on
screen as demonstrated. The authenticated user can now give
their vote to any registered candidate. The accepted vote
details are prompted to the voter as shown in Fig. 23.

Fig.19 User login screen if invalid details provided by the user

In the screen shown in Fig.19, the webcam asked to show the

Aadhaar number and then clicked on the 'Take Snapshot'
button to check number is coming clearly in the image as seen
in the right side panel. After the Aadhaar card is successfully
verified, the user will be asked for face verification as shown
in Fig. 20.

Fig. 22 Voter Identified details and candidate options to caste vote

Fig.20 Aadhaar card verification stage

Fig.23 Accepted Vote status for candidate

508

Authorized licensed use limited to: VTU Consortium. Downloaded on October 23,2024 at 09:38:44 UTC from IEEE Xplore. Restrictions apply.
After the voting procedure is completed the admin can view accuracy and reliability of the voting system but also greater
the vote count for authenticated voters and announce the confidence among voters in the fairness and transparency of
election results. the electoral process. Overall, the incorporation of these
authentication highlights the commitment to conducting free,
fair, and secure elections in the digital age. The future scope
of this work includes adding the Chatbot to the web
application which can help users during registration process
or while casting vote.
REFERENCES
[1] M. Mochta, C. Lesschaeve, J. Glaurdić, “Voting and winning:
perceptions of electoral integrity in consolidating democracies,”
Democratization, 2021 Nov, 17; vol. 28(8), pp. 1423-41.
[2] S.A. Sheikh, A. Syed, O. Saleh, M. A. Khan, A. Bali A., “Online
Voting System,” International Journal of Advanced Research in
Science, Communication and Technology (IJARSCT). 2023. vol. 3(2),
pp. 477-483.
[3] M. Daud, I.R. Khan, “Issues & Vulnerabilities Related to Online
Voting and Its Mitigation,” International Journal of Advanced
Research in Computer Science, 2017 May, vol. 8(5).
[4] G. Kumar, S. Gupta, D. Agarwal D, A. Tiwari, “Virtual Voting
System,” International Journal of Informatics, Information System and
Computer Engineering (INJIISCOM), 2021, vol. 2(1), pp. 77-82.
[5] S. Duth, B. Amulya, “Recognition of hand written and printed text of
cursive writing utilizing optical character recognition”. In 2020 4th
international conference on intelligent computing and control systems
Fig. 24 Admin view for vote count (ICICCS), 2020 May 13, pp. 576-581. IEEE.
[6] P. Sirenjeevi, C. Preethi, S. Shaminee, J. Gayathri, “Smart Voting
System using Deep Learning Techniques and Facial Authentication,”
In comparison to existing systems [6-9] the proposed International Journal Of Engineering Research & Technology (IJERT)
multilayered voting system based on biometric analysis is ICONNECT. 2022, vol. 10(09).
secure, cost-effective, consumes less time and maintains [7] S. Abimanyu, R.Anbarasan, S. Dhinakaran, M. K. Abhirami, “Iris
Detection based Authentication for Secure Voting System”.
voters' authenticity. In the model presented by Abhimanyu et
[8] R. Govindaraj, P. Kumaresan, “Online voting system using cloud,” In
al.[7] and Rajesh and Priyanka [9] based on Iris and 2020 International Conference on Emerging Trends in Information
fingerprint detecion, voter verification is performed but there Technology and Engineering (ic-ETITE), 2020 Feb 24, pp. 1-4. IEEE.
is no provision to check duplicate votes. However, the [9] M. G. Rajesh and S. Priyanka, “Online Voting System,” International
proposed model carried out multistage voter verification Research Journal of Engineering and Technology (IRJET),2017, vol.
adding additional security and also doesn’t allow duplicate 4(12), pp. 1411-1413.
votes to caste vote. The proposed system does not require any [10] E. Gangadurai, R. Divakaran, U. Aruneshwaran, “Fingerprint-Based
Voting System,” Journal of Telecommunication Study, 2023, vol. 8(2),
OTP for voter and admin verification as used in model pp 30-38.
developed by Sirenjeevi et al.[6], thus making the voting [11] C. M. Nalayini, K. Vishnupriya, A. Dhivyabharathi, H.Yuvapriya,
system more secure. The cloud based system developed by “Biometric based Mobile Voting Application,” Journal of Information
Govindaraj and Kumaresan[8] permits users to caste vote Technology and Digital World, 2023, pp.159-171.
online using email and no additional authentication of voters [12] M. Gupta, S. V. Amirishetty, B. Nithin, H. Kurakula, “Mind Frame
Music and Movie Recommendations to uplift the current mood using
identity is performed. The proposed model allow only the Deep Learning,” In: 2021 International Conference on System,
voters who are verified through biometric features and Computation, Automation and Networking (ICSCAN), 2021 Jul 30,
Aadhar card details to caste vote, the users with non-matching pp. 1-6. IEEE.
features are not allowed to cast the votes. Thus providing a [13] S. M. Cheruvu, D. Sesank, K.S. Sandilya, M. Gupta, “Vituperative
secure and authenticated voting system using the unique high Content Detection: A Multidomain Architecture Using OpenCV,” In:
International Conference on Advances and Applications of Artificial
precision features of the voters. Intelligence and Machine Learning, 2022 Sep 16, pp. 409-421.
Singapore: Springer Nature Singapore.
VI. CONCLUSION AND FUTURE SCOPE [14] A. B. Shetty and J. Rebeiro, J, “Facial recognition using Haar cascade
This research concluded that it ensures the security and and LBP classifiers,” Global Transitions Proceedings, 2021, vol. 2(2),
integrity of the electoral process. By leveraging biometric pp.330-335.
authentication methods namely thumb and face recognition, [15] A.K. Al Tamimi, A. Qasaimeh and K. Qaddoum, “Offline signature
recognition system using oriented FAST and rotated BRIEF,”
along with Aadhaar Number validation, this system enhances International Journal of Electrical and Computer Engineering (IJECE),
voter verification and reduces the risk of fraudulent activities. 2021, vol. 11(5), pp.4095-4103.
Moreover, this multi-layered approach not only enhances the

509

Authorized licensed use limited to: VTU Consortium. Downloaded on October 23,2024 at 09:38:44 UTC from IEEE Xplore. Restrictions apply.