1

International Conference on Communication and Information Processing

(ICCIP-2023) Available on:
Elsevier-SSRN

Online Voting System with Face Recognition and One Time Password
Uma Hombal1, Kalpesh Chaudhari2, Sanika Utpat2, Shruti Yadav2, Parth Joglekar2
1
Research Scholar KSIT, Bengaluru-India
2
Department of Computer Science and Engineering, Nutan College of Engineering and Research, Pune,
Maharashtra, India

uhombal@gmail.com, kdchaudhari2000@gmail.com, sanikautpat2@gmail.com, svyadav4801@gmail.com

parthjoglekar213@gmail.com
________________________________________________________________________________

Abstract
The electoral process is the foundation of democracy and governance. Within the last few decades, the election
system has effectively undergone a number of modifications. Despite having the largest majority rule
government in the world, India still uses either secret ballot voting (SBV) or electronic voting machines
(EVM), both of which are expensive, labor-intensive, and wasteful. The current method merely checked
identifying documents, increasing the likelihood of fraudulent votes. The main goal of this system is to provide
an online voting system that, by using a camera for face recognition and OTP generation, will help to reduce
voter fraud in manual voting systems and earlier iterations of online voting. Additionally, we are building a
method of remote voting. to voters who are unable to travel to their hometown's polling place. To guarantee
the dependability of the device, we are supplying software that has multiple layers of verification, including
face recognition or verification, OTP verification, and validation data. Only after being authenticated and
matched with the provided voter database may a single voter access the system. The voter will be able to
continue choosing their chosen candidate from the panel once the corresponding face has been matched with
the data provided.

Keywords: Smart Voting System, Facial Recognition, OTP, Voter ID, Candidate Registration.

© 2023 –Authors

1. Introduction

Elections will unavoidably occur in a democratic society, and it is the responsibility of the government and the
people to guarantee that they take place in a secure environment. It occurs in a very effective way. This method requires
the voter to register his face ahead to the election, and the results are compared when they cast their ballot. The
Individual Database management falls under the purview of the web application coder. When a user casts a ballot, the
website confirms that the vote was properly registered by displaying the phrase "voted successfully."
expressing their ideas, casting a ballot, or stating their preferences. The primary objective of this endeavor is to
establish a voting system that uses face recognition technology and an OTP system to enable voters to cast ballots.
ballots can be cast online from any place on Earth. The voting results are stored in the server database. Since adaptation
is necessary in order to survive and uphold international norms in a world that is always growing. With this innovative
technology, elections can be held both online and offline, but an efficient data transfer and precise result calculation
are made possible by a central database.

Electronic copy available at: https://ssrn.com/abstract=4670374

 2

Therefore, an electronic voting system must be developed and implemented in order to have a fair election. Online
voting has become more popular recently as a way to increase accessibility and voter turnout. The usage of online
voting technologies raises questions about security and the impartiality of the voting process. a two-step verification
process that offers a new position of security.
This investigation paper will look at the underlying benefits and drawbacks of using two-step verification in an
online voting system to ensure the security and integrity of the electoral process. The paper will discuss the current
state of online voting systems, their challenges, and the risks associated with technological breakthroughs in the field.

2. Related work

This project development consists of the four modules administrator, voter registration, nominated candidate, and
field officer. Here, a one-time password algorithm is used to confirm the voters. The functionality of the proposed
system enables a voter to cast their vote from the polling place website online. E-voting incorporates sociological,
sociolegal, and monitoring aspects of the present system [7].
With this system, the voter's identity will be physically verified. The purpose of this implementation is to confirm
that the one-time password (OTP) method is being used by the voter. In the section next, we'll discuss the one-time
password algorithm. The voter initially registers on the electronic voting website by providing personal information
such as their address, phone number from their Adhaar card, etc. The report will be created by the field officer and
sent to the administrator. It is the administrator's duty to activate the voter so they can cast their ballot or leave a
note. The admin starts the voter registration procedure if the field officer's report is accurate; otherwise, they stop it.
Voters must ensure that the mobile number they submit is accurate because it is required to verify their voting status.

[2] did research and made observations of the drawbacks of the manual or traditional voting systems used by several
nations. They made the following points to explain their observation:

Vote tallying and results release:

Manual ballot counting is used, and bottom-up results collation is used to compile the results. Accordingly, counting
starts at the lowest level (i.e., the polling station level), and results are then tallied up to the constituency level,
which is followed by the national level. The total number of votes cast for each contender is determined at the
highest level. The EC then releases the final results.

They tried putting into practice such strategies as:

Installing a database management system (DBMS):

To manage system data, a DBMS called the OVIS Database Manager (ODM) would be installed. Voter data could
be simply and quickly retrieved, added, updated, removed, sorted, and navigated thanks to ODM. ODM will also
keep track of election results data and other system-related information.

Implementation of safeguards to prevent duplicate voting:

The system includes user identification and access control features to prevent double voting and vote manipulation.
Additionally, these characteristics would aid in preventing unauthorized users from gaining access to the system. As
a result, a voter who makes a second effort to cast a ballot will be denied entry and prompted by the computer
system.

Report generation:

The system would enable the creation and printing of a variety of reports, from information about voter registration
to election results. According to [3], a study was undertaken using internet voting to better understand human
behavior. In the essay, they claimed Our empirical research was based on two context-free, simple-to-understand
interactive voting games. In both cases, voters' private preferences are automatically allotted among a predetermined
group of candidates (unknown to the other voters). The first configuration is a one-shot voting scenario with a
predetermined preference order for a single human voter. By giving the voter a (non-binding) pre-election poll of
other voters' votes, we come to control the data available to the voter and record her voting behavior under
circumstances that change the poll's contents.
A group of human participants in an iterative voting game makes up the second setting. Similar to the previous
scenario, the voters are given their preference profile but are allowed to change their votes whenever they like until a
consensus is reached (or a timeout). We documented each voter's choice and the facts she had access to at the
moment, just like in the poll game. In both games, voters are uncertain, but in the first game, the only information

Electronic copy available at: https://ssrn.com/abstract=4670374

 3

available to them is an unreliable poll. In the second game, a participant directly sees the votes of her peers, but she
is unaware of how they will ultimately cast their ballots or when they will do so.

The notion that we are recommending three layers of verification is quite effective in lowering the possibility of fake
voting, according to [6] online voting using facial recognition. The voter would receive a unique ID after registering,
which is included in the first. After that, the voter must proceed to the third level of security, where their
identification will be cross-checked by an Election Commission officer. This level of verification will significantly
increase security because it involves comparing the voter's current facial features to those in the database, which will
lessen the likelihood of a voter casting a vote in error and make the system safer.

Three different degrees of security are being used, according to [7], and they are as follows:

Level 1: -Unique ID number (UID):

During voter registration, the system will ask the voter for their unique ID. The election commission's database is
used to verify the entered unique ID.
Level 2: The ID number from the election commission.
The voter must enter either the voter's id number or the electoral commission identification during the second level
of verification. The election commission's database is used to verify the entered ID number.
Level 3: Face recognition with the appropriate ID number from the election commission. In this level, the facial
images of the voters from the database provided by the electoral commission are verified using the Eigen face
algorithm and Gabor Filters Algorithm.
Additionally, they contrasted the Eigen face method with the Gabor filter technique.

[8] Construction of efficient machine smart voting machine with liveness detection model, They propose that
numerous spoofing fingerprints are recognized during electronic voting due to the rising use of technology. In other
words, there is just one method for determining whether something is alive, and that method is liveness detection.
Any functional information for the person or any functional information in the system will be identified by liveness
detection. The results of the vote might be found at any point during the voting process, in contrast to conventional
voting, which allows the results to be announced as soon as the voting machine completes the subsequent phase.

Their designed module yields effective results for fake fingerprint images. In addition, it has been predicted that the
recommended module has a significantly higher efficiency and accuracy than other traditional modules. However,
they are observing a problem with the total response time with the suggested fraud analysis module. To reduce
reaction time over the course of its execution, the planned job should be changed so that it uses fewer electronic
devices.

[9] Smart Voting, a web-based tool that uses face, Aadhar, and OTP verification. Their effort aims to create an
interactive voting system where people can take part while creating an Aadhar ID using data that has already been
pre-stored in the database. They have a system where anyone with Indian nationality who is at least 18 years old and
of any gender can cast a ballot online without going to a real voting place. The voter's photo ID and image stored in
the database will be used to confirm the user at each login. Through this initiative, they hope to create a safe website
that unifies all available voting options.

[11] claims that their suggested solution is a two-part one that includes voting on a website and voting on mobile
devices. The voter may choose from either of the two options as it suits him. A robust internet connection is required
for the software's initial fold, which has a straightforward and self-explanatory GUI. On the other side, voters can
cast ballots using regular mobile phones with interactive voice response (IVR). They have also made use of the iris
scanning and verification technology to further increase the security of their system. When a voter registers with the
system, his voter ID and Aadhar number should match. Their technology also offers high transparency of voters'
details at that moment alone.

According to [13], they are creating the system for a company with offices spread across several cities. Their first
priority is to protect cast votes as they are being transported from the voter to the voting server to be stored. They are
concentrating on providing security from both passive and active intruders.
They are utilizing digital signatures to provide protection against active intruders who could change or tamper with
the cast vote as it is being transferred from the voter to the voting server. When votes are transferred between voting
clients and voting servers, their technology offers protection from all types of threats. Security threats from both
passive and active intruders are included in these attacks.

Electronic copy available at: https://ssrn.com/abstract=4670374

 4

In line with [15], Android phones can be equipped with their application, and on election day, the administrator will
turn on voting. If there are different polling phases, the programme will only make the user available on election
day. By entering his voter ID or Aadhar number (a unique identifier) and a secret password that is given to him, the
user can access the application. The user chooses his or her constituency and booth number or location as soon as
they log in, after which they can cast their votes and exit out of the application. The vote should be confidential and
accessible in the counting centre on election day. Encryption and decryption algorithms are utilized to keep the
vote secret. Therefore, as soon as a user casts a ballot, it is encrypted to ensure its secrecy, and this encryption is
maintained on election day as well.

3. Methodology

Helen Chan Wolf and Charles Bisson created the first ever computer-based Face Detection. The widows peak in the
hairline, the inside and outside corners of the eyes, and the pupil centre all needed to be manually located. Twenty
distances were calculated using the coordinates, including the width of the mouth and the eyes. It is possible to analyze
40 images in an hour in this way, creating a database of the computed distances. It is possible to analyze 40 images
in an hour in this way, creating a database of the computed distances. After a computer automatically checked the
distances for each image, determined how far away they were, and did so, the closed records would then be returned
as a probable match.

The approach makes use of a Motivation Increases of Basic Features for Object Recognition. When the algorithm is
being trained. It utilizes an Object Detection Algorithm to find faces in still photos or moving movies. Given a
considerable number of positive images with faces and a sizable number of unfavorable photos, the edge or line
detection capabilities presented in Viola and Jones' 2001 study.

4. Proposed Work

Event after so many improvements in technology, the voting system in India has not been upgraded event after so
many years. Our work takes into account all the advancements and combines them to use the latest and greatest
innovations in technology to upgrade Voting Systems.
In our System, face data and phone number of an user are stored on registration which then cannot be changed by
the user until he/she contacts an Admin to change this data. The registration is done by taking information about the
voter, such as their Voter Id, Adhaar Card Number, Phone Number, Email, etc. and finally an image of the user is
stored to use during the time of voting and thus, giving us our training data as well as data to match while user is trying
to vote to a certain party/candidate.
Before voting, user has to login to their accounts and then verify using a One Time Password which is sent to their
respective phone number provided when registering. Once user is verified using the One Time Password, he/she then
has to verify their Face before they are finally allowed to vote.

Electronic copy available at: https://ssrn.com/abstract=4670374

 5

Another layer of protection we have is that, if someone else tries to vote from someone else‟s account, even if
they‟ve been successful in retrieving their Phone Number and OTP for initial verification, they also need to pass the
face verification. After all the voting procedure has been successful, the user can see all the votes and take a look at
the results.

4.1. Use Case Diagram

Fig. 4.1. Use Case Diagram

4.2. DFD Diagrams

Fig. 4.2.1. DFD-0

Electronic copy available at: https://ssrn.com/abstract=4670374

 6

Fig. 4.2.2. DFD-1

Fig. 4.2.3. DFD-2

5. Implementation
We are glad to inform that the Online Voting System has been successfully built.

Electronic copy available at: https://ssrn.com/abstract=4670374

 7

6. Further Work

As technology develops and advances face recognition and OTP (One Time Password)-based internet voting
systems are expected to have a wider range of applications. Following are some potential hotspots for fluorescence
intensity:
1. As face recognition technology advances, the accuracy of face recognition algorithms is likely to rise, which
will likely result in more accurate online voting systems. This can mean adding new algorithms or data
sources to increase the dependability of the system.
2. Strengthening security measures: It is crucial to defend online voting systems from threats like fraud or
hacking, thus future developments might focus on improving this defense. This may need the use of more
complex encryption techniques or extra authentication methods, such as the use of biometric data.
3. Improving accessibility: Future online voting platforms may work to make their platforms more accessible
to those with disabilities or those who might find it difficult to use traditional voting methods. This may
need developing fresh user interfaces or integrating assistive technology in order to make the system more
usable for all users.
4. Increasing the use of online voting: As the dependability and security of online voting systems improve,
more elections and decision-making processes are anticipated to use them. This may entail employing
online voting for corporate governance or integrating it with other forms of e-government.
5. Integration of Cloud computing, WSN, IoT and Bigdata technology based systems with PCA/ICA
techniques to improve the identification efficiency and to achieve location independent operations [4, 5,
10, 12, 14, 16].

7. Conclusions
As we can see, the electoral system has several shortcomings, including a lengthy process that consumes a lot of
time, is unreliable, permits false voting, and lacks any kind of security. We can now, however, assert that our
mechanism is more advantageous and secure than the existing one. The three levels of safeguards in this proposed
system make it simple to identify the bogus voters. By identifying fraudulent voters with the aid of facial
authentication methods, we can stop fake ballots from being cast during the election commission. By integrating into
our recommended online smart voting system, people can cast their votes from any location. Because internet is
required for every operation, it is a one-time expenditure for the government. Voting is more significant than where
someone lives. Data is continuously accessible, and backup copies are also kept in a centralized repository. Data is
plausible. The sophisticated voting system updates the results every minute. also employs less laborer and resources.
The database needs to be updated annually or just before an election to allow for the insertion of new persons
becoming eligible to vote as well as the removal of deceased voters.
Raising the system's security and efficiency standards is part of the system's potential future scope. It would also be
interesting to meet the other undiscovered primitives to improve system efficiency.
Future improvements may also handle system failures and power outages to increase voter confidence.

References

[1] Patidar, K., & Jain, S. (2019, July). Decentralized e-voting portal using blockchain. In 2019 10th International Conference on
Computing, Communication and Networking Technologies (ICCCNT) (pp. 1-4). IEEE.
[2] Rezwan, R., Ahmed, H., Biplob, M. R. N., Shuvo, S. M., & Rahman, M. A. (2017, December). Biometrically secured electronic voting
machine. In 2017 IEEE Region 10 Humanitarian Technology Conference (R10-HTC) (pp. 510-512). IEEE.
[3] Usmani, Z. A., Patanwala, K., Panigrahi, M., & Nair, A. (2017, March). Multi-purpose platformindependent online voting system. In
2017 International Conference on Innovations in Information, Embedded and Communication Systems (ICIIECS) (pp. 1-5). IEEE
[4] Mukund Kulkarni, Prachi Deshpande, Sanjay Nalbalwar, Anil Nandgaonkar, "Taxonomy of load balancing practices in the cloud
computing paradigm", International Journal of Information Retrieval Research, Vol.12, no. 3, pp. 1-15, 2022.
[5] Kulkarni, M., Deshpande, P., Nalbalwar, S., Nandgaonkar, A. (2022). Cloud Computing Based Workload Prediction Using Cluster
Machine Learning Approach. Smart Innovation, Systems and Technologies, vol 303, pp.591 -601. Springer, Singapore.
https://doi.org/10.1007/978-981-19-2719-5_56

Electronic copy available at: https://ssrn.com/abstract=4670374

 8

[6] Hongyu Zhang., Qianzi You, and Junxing Zhang (2015), „A lightweight electronic voting scheme based on blind signature and
kerberos mechanisms‟, International conference on advanced networks and tele-communications systems,pp. 978- 4799.
[7] Ashwini Mandavkar, Prof. Rohini Agwane, “Mobile based facial recognition using OTP verification for voting system”, 2015
IEEE, IACC, pp. 644-649
[8] Z. A. Usmani, K. Patanwala, M. Panigrahi, and A. Nair, “Multi-purpose platform independent online voting system,” in 2017
International Conference on Innovations in Information, Embedded and Communication Systems (ICIIECS), 2017, pp. 1–5.
[9] N. Keerthi, A. Raghuram, and R. Jayaraman, “Interfacing of online and offline voting system with an E-voting website,” in 2022
6thInternational Conference on Devices, Circuits and Systems (ICDCS), 2022, pp. 223–228.
[10] Patil, M.N., Brijesh Iyer, Rajeev Arya (2016). Performance Evaluation of PCA and ICA Algorithm for Facial Expression
Recognition Application. Advances in Intelligent Systems and Computing, vol 436. Springer, Singapore.
https://doi.org/10.1007/978-981-10-0448-3_81
[11] A. Shankar, Harsshanth, J. Solai, Geetha, S. Haran, and D. Jemima D, “Migrant voting system-solution to gain the lost votes,” in
20212nd International Conference on Smart Electronics and Communication (ICOSEC), 2021, pp. 770–774.
[12] S. Panbude, B. Iyer, A. B. Nandgaonkar, and P. S. Deshpande, “DFPC: Dynamic Fuzzy-based Primary User Aware clustering
for Cognitive Radio Wireless Sensor Networks”, Eng. Technol. Appl. Sci. Res., vol. 13, no. 6, pp. 12058 –12067, Dec.
2023.doi:10.48084/etasr.6279
[13] G. Kalaiyarasi, K. Balaji, T. Narmadha, and V. Naveen, “E-voting system in smart phone using mobile application,” in 2020 6th
International Conference on Advanced Computing and Communication Systems (ICACCS), 2020, pp. 1466–1469.
[14] S. Panbude, B. Iyer, A. B. Nandgaonkar, and P. S. Deshpande, “DFPC: Dynamic Fuzzy-based Primary User Aware clustering for
Cognitive Radio Wireless Sensor Networks”, Eng. Technol. Appl. Sci. Res., vol. 13, no. 6, pp. 12058–12067, Dec.
2023.doi:10.48084/etasr.6279
[15] S. Ganesh Prabhu, A. Nizarahammed., S. Prabu., S. Raghul., R. R. Thirrunavukkarasu, and P. Jayarajan, “Smart Online Voting
System,” in 2021 7th International Conference on Advanced Computing and Communication Systems (ICACCS), 2021, vol. 1,
pp. 632–634.
[16] Shirke, K., Iyer, B. Through the wall human detection & surveillance sensor. Int J Syst Assur Eng Manag 14 (Suppl 1), 569–
574 (2023). https://doi.org/10.1007/s13198-022-01853-5
[17] S. Panbude, B. Iyer and A. Nandgaonkar, "Clustering solutions and analysis for cognitive radio Wireless Sensor Networks," 7th
International Conference on Computing in Engineering & Technology (ICCET 2022), Online Conference, 2022, pp. 267-273,
doi: 10.1049/icp.2022.0630.
[18] K. Malladi, S. Sridharan, and L. T. JayPrakash, “Architecting a large-scale ubiquitous e-voting solution for conducting
government elections,” in 2014 International Conference on Advances in Electronics Computers and Communications, 2014, pp.
1–6.
[19] X. I. Selvarani, M. Shruthi, R. Geethanjali, R. Syamala, and S. Pavithra, “Secure voting system through SMS and using smart
phone application,” in 2017 International Conference on Algorithms, Methodology, Models and Applications in Emerging
Technologies (ICAMMAET), 2017, pp. 1–3. 20. S. Damle, S. Gujar, and M. H. Moti, “FASTEN: Fair and secure distributed
voting using smart contracts,”in 2021 IEEE International Conference on Blockchain and Cryptocurrency (ICBC), 2021, pp. 1–3.
[20] M. A. Cheema, N. Ashraf, A. Aftab, H. K. Qureshi, M. Kazim, and A. T. Azar, “Machine learning with blockchain for secure E-
voting system,” in 2020 First International Conference of Smart Systems and Emerging Technologies (SMARTTECH), 2020, pp.
177–182

Funding Information: The reported work did not receive any funding from any Institutions or
Individuals.

Competing Interest Declaration: The authors do not have any competing interest with any
Institutions or Individuals.

Ethical Statement: No human/animal clinical trials were conducted for this research. Further, this
paper had used publicly available data sets/information.

Electronic copy available at: https://ssrn.com/abstract=4670374