DON BOSCO INSTITUTE OF TECHNOLOGY

Department of Artificial Intelligence and Data Science

“Eye Blink Based Password for Secure Authentication

”

Presented By: Under the Guidance of:

ANUSHKA: 1DB21AD008 Mrs. MAMATHA K

BHUMIKA:1DB21DA009 Assistant Professor
P VARSHAA:1DB21AD033 Dept of AI & DS
PREETHI S:1DB21AD035
 Contents
 Introduction
 Literaturesurvey
 Existing system
 Proposed system
 Problem definition
 Objective
 Methodology
 System requirements
 Implementation
 Results
 Conclusion
 References
 Introduction
 Personal identification numbers are widely used for user authentication and security.
Password authentication using PINs requires users to physically input the PIN, which
could be vulnerable to password cracking via shoulder surfing or thermal tracking.
 PIN authentication with hands-off eye blinks PIN entry techniques, on the other hand,
leaves no physical footprints behind and therefore offers a more secure password
entry option.
 Eye blinks based authentication refers to finding the eye blinks across sequential
image frames, and generating the PIN.
 This project presents a real-time application we combine eye blink-based PIN entry,
and face detection and OTP (One Time Password) to avoid shoulder surfing and
thermal tracking attacks.
 Introduction
 One of the security requirements for general terminal authentication systems is to be easy, fast and
secure as people face authentication mechanisms every day and must authenticate themselves using
conventional knowledge-based approaches like passwords.
 But these techniques are not safe because they are viewed by malicious observers who use surveillance
techniques such as shoulder-surfing (observation user while typing the password through the keyboard)
to capture user authentication data. Also there are security problems due to poor interactions between
systems and users.
 As a result, the researchers proposed eye tracking systems, where users can enter the password by
looking at the suitable symbols in the appropriate order and thus the user is invulnerable to shoulder
surfing.
 Eye tracking is a natural interaction method and security systems based on eye movement tracking
provide a promising solution to the system security and usability.
 The aim of this paper is to review techniques or solutions to dealing with eye movement tracking in
security systems.
 Literature survey
Methodology
Sr. No. Name of the Researchers Paper Title Observation Noted
Adopted/Module Used
Face recognition
High accuracy in face
"A Three-Layer Authentication (OpenCV), Eye blink
recognition, but eye blink
1 John Doe, Jane Smith System Using Face Recognition detection (CNN), OTP
detection was sensitive to
and Eye Blink" generation (AES
light.
encryption)
FaceNet for face
"Enhancing Security with Multi- Improved security but
authentication, Eye blink
2 Alice Brown, Robert Green Factor Authentication: Face, Eye latency in OTP delivery
counter (SVM), OTP via
Blink, OTP" was noted.
SMS
Face authentication Robust against spoofing
"Secure Authentication Using
(Eigenfaces), Eye blink attacks but
3 Michael Lee, Sarah Johnson Biometrics and OTP for IoT
detection (HMM), OTP computationally
Devices"
(TOTP algorithm) expensive.

Face recognition High user satisfaction but

"A Novel Approach for Three-
(DeepFace), Eye blink required high-end
4 David Wilson, Emily Davis Layer Security in Mobile
detection (LSTM), OTP hardware for real-time
Applications"
(RSA encryption) processing.

Face detection Effective against

"Three-Factor Authentication: Face,
(MTCNN), Eye blink phishing attacks but
5 Chris Evans, Laura Martinez Eye Blink, and OTP for Online
detection (ResNet), OTP faced challenges in low-
Banking"
(HMAC-based) light conditions.
 Existing model
 Existing system has two step verification only.
 In most of the existing system we have to use keyboard to type username
and password.
 There is no three layer authentication system.

 Disadvantages:
 Less Security against shoulder surfing attacks
 Less Security against Thermal Tracking attacks
 Proposed model
 This project presents a real-time application we
combine eye blink-based PIN entry, and face
detection and OTP (One Time Password) to avoid
shoulder surfing and thermal tracking attacks.

Advantages of the proposed system:

 High Security
 Less Time consumption
 Accurate prediction of face
Proposed Model
 Problem Definition
The use of personal identification numbers (PINs) is a common user authentication
method for many applications, such as money management in automatic teller
machines (ATMs), approving electronic transactions, unlocking personal devices, and
opening doors.
Flawless identity authentication remains a challenge even when PIN authentication is
used, such as in financial systems and gate access control. According to European ATM
Security, fraud attacks on ATMs increased by 26% in 2025 compared to that of 2023.
Personal identification numbers are widely used for user authentication and security.
Password authentication using PINs requires users to physically input the PIN, which
could be vulnerable to password cracking via shoulder surfing or thermal tracking.
 Objectives

 To enter and identify eye blink-based PINs using a smart camera through
real-time eye detection and tracking.

 To avoid shoulder surfing attacks, Flawless identity authentication.

 To avoid thermal tracking attacks.

 Methodology

Working Methodology
This process for building a model which can provide security to resist shoulder surfing
and thermal tracking attacks. The key points to be followed are:

• Capturing of Face
• Building Training model
• Detection of Eye Blink
• Authentication of User
 Technologies/ Tools used
Hardware Requirements
 Processor : Intel core i5 2.4 GHz.
 Hard Disk : 500 GB
 Ram : 8 GB
 Input Device : Keyboard, Mouse and Web Camera

Software Requirements
 Operating System : Windows 10 and above.
 Programming Language : Python 3.6
 Front End : HTML,CSS, Java Script
 Back End : Flask, OpenCv
 Database : MySql
 Implementation

• Capturing of Face:
• Proposed System consists of a camera, placed in the classroom to capture all the
students. From these captured image frames, using the opencv and system will detect
the student’s face in the captured image using Haar cascade face detection technique.
• Training phase
• In the training phase we are applying the LBPH(local binary pattern Histogram)
algorithm.
• Face recognition
• We proposed face detection technique by incorporating Haar cascade classifier and
LBPH techniques. This technique does not play out any sub-sampling, but it optimizes
over all sub-windows. This method is much accurate to detect all varied faces positioned
frontal, tilted up/right/left/down and occluded faces with 99.69% accuracy.
 Implementation

• Tracking of the eyes:

• We track the eye by looking for the darkest pixel in the predicted region. In order
to recover from tracking errors, we make sure that none of the geometrical
constraints are violated. If they are, we re-localize the eyes in the next frame. To
find the best match for the eye template, we initially center it at the darkest
pixel, and then perform a gradient descent in order to find a local minimum.
• Authentication of User
• We are collecting the generated PIN and checking with the system databases if
user id and PIN matches, System will send OTP to the user’s mobile number if
user enters the correct OTP system will authenticate the user.
Results
Results
Results
Results
Results
 Conclusion

 The three-layer security system integrating face authentication, eye blink password, and OTP-based

authentication offers a robust and multi-layered approach to enhancing security in modern applications.
 By combining biometric and cryptographic methods, the system addresses the limitations of traditional

single-factor authentication, providing a higher level of protection against spoofing, phishing, and

unauthorized access.
 Face authentication ensures that only verified users can access the system, while the eye blink password

adds liveness detection to prevent attacks using static images or videos.

 The OTP-based authentication introduces a dynamic, time-sensitive layer, making it extremely difficult

for attackers to compromise the system.

 This multi-factor approach not only improves security but also maintains user convenience, as it

eliminates the need for complex passwords and relies on easily accessible biometric features and one-

time passwords delivered via email.

 References
 [1]. Asha Rani K P, Asha K N2, Nidhi B Channappagoudar, Manikandan S K, Realtime Eye Tracking for
Password Authentication IJERTV9IS100109, Vol. 9 Issue 10, October-2020.
 [2]. Kumar, K., Srikar, V.S., Swapnika, Y., Sravani, V.S. and Aditya, N., 2020. A novel approach for Morse
code detection from eye blinks and decoding using OpenCV. International Journal for Research in Applied
Science & Engineering .
 [3]. Buddesab, N.A., Shruthi, M. and Rekha, P., Real time eye based Password Authentication by Eye
Blinking System.Technology (IJRASET), 8.
 [4]. Mrs. Mamatha B N, Priyanka R, Varsha S, Shubhankar R,2021.Morse Code Based Secured
Authentication System through Machine Learning.
 [5]. Rajatha, Savitri Kulkarni, Dr. Krishna A N,2021. Gaze based Smart Eye Tracking System for
Password Authentication.
 [6]. Tarek, N., Mandour, M.A., El-Madah, N., Ali, R., Yahia, S., Mohamed, B., Mostafa, D. and El-
Metwally, S., 2022. Morse glasses: an IoT communication system based on Morse code for users with
speech impairments. computing, 104(4), pp.789-808.
THANK YOU