A PROJECT REPORT ON

A Face Recognition-Based Smart Student Attendance and Activeness

Monitoring System

SUBMITTED TOWARDS THE

PARTIAL FULFILLMENT OF THE REQUIREMENTS
OF BACHELOR OF TECHNOLOGY IN
INFORMATION TECHNOLOGY
BY
Mr. Ghodekar Atharva Vinod [UIT20M1025]
Mr. Gunjal Tejas Gokul [UIT20M1026]
Mr. Landage Abhijeet Ashok [UIT20M1038]
Mr. Phapale Abhijit Balasaheb [UIT20M1047]
Mr. Thorat Dinesh Ashok [UIT20M1065]
B.TECH. INFORMATION TECHNOLOGY
UNDER THE GUIDANCE OF
Dr. R. R. Nikam

SANJIVANI RURAL EDUCATION SOCIETY’S

SANJIVANI COLLEGE OF ENGINEERING, KOPARGAON
(AN AUTONOMOUS INSTITUTE)
DEPARTMENT OF INFORMATION TECHNOLOGY
2023-24
 Sanjivani Rural Education Society’s

Sanjivani College of Engineering, Kopargaon-423603

(An Autonomous Institute Affiliated to Savitribai Phule Pune University, Pune)
(NAAC ‘A’ Grade Accredited)

Department of Information Technology

(UG Program, NBA Accredited)
SANJIVANI RURAL EDUCATION SOCIETY’S

CERTIFICATE
This is to certify that the Project Entitled
A Face Recognition-Based Smart Student Attendance and Activeness
Monitoring System
Submitted by
Mr. Ghodekar Atharva Vinod [UIT20M1025]
Mr. Gunjal Tejas Gokul [UIT20M1026]
Mr. Landage Abhijeet Ashok [UIT20M1038]
Mr. Phapale Abhijit Balasaheb [UIT20M1047]
Mr. Thorat Dinesh Ashok [UIT20M1065]
is a bonafide work carried out by students under the supervision of Dr. R. R. Nikam and it
is submitted towards the partial fulfillment of the requirement of Bachelor of Technology
(Information Technology). During the academic year 2023-24

Dr. R. R. Nikam Dr C. D. Bawankar

[Internal Guide] [Project Coordinator]

Dr. M. A. Jawale Dr. A. G. Thakur

HoD IT Director

Internal Examiner External Examiner

 PROJECT APPROVAL SHEET
A Project Report On
A Face Recognition-Based Smart Student Attendance and Activeness
Monitoring System
is successfully completed by

Mr. Ghodekar Atharva Vinod [UIT20M1025]

Mr. Gunjal Tejas Gokul [UIT20M1026]
Mr. Landage Abhijeet Ashok [UIT20M1038]
Mr. Phapale Abhijit Balasaheb [UIT20M1047]
Mr. Thorat Dinesh Ashok [UIT20M1065]

Sanjivani Rural Education Society’s

Sanjivani College of Engineering, Kopargaon-423603

(An Autonomous Institute Affiliated to Savitribai Phule Pune University, Pune)
(NAAC ‘A’ Grade Accredited)

Department of Information Technology

(UG Program, NBA Accredited)
SANJIVANI RURAL EDUCATION SOCIETY’S
ACADEMIC YEAR 2023-24

Dr. R. R. Nikam Dr C. D. Bawankar

[Internal Guide] [Project Coordinator]

Dr. M. A. Jawale Dr. A. G. Thakur

HoD IT Director
 Acknowledgement

A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

has been a subject with tremendous scope to research, which leads ones mind to explore
new heights in the field of Information Technology, and its miscellaneous applications. We
dedicate all our project work to our esteemed guide Dr. R. R. Nikam whose interest and
guidance helped us to complete the work successfully as well as he has provided facilities
to explore the subject with more enthusiasm.

This experience will always encourage us to do our work perfectly and professionally. We
also extend our gratitude to Dr. C. D. Bawankar(Project Coordinator) and Dr. M. A.
Jawale (H.O.D. Information Technology).

We express our immense pleasure and thankfulness to all the teachers and staff of the
Department of Information Technology, Sanjivani College of Engineering, Kopargaon for
their cooperation and support. Last but not least, we thank all others, and especially our
parents and friends, who in one way or another, helped us in the successful completion of
this project.

Mr. Ghodekar Atharva Vinod [UIT20M1025]

Mr. Gunjal Tejas Gokul [UIT20M1026]
Mr. Landage Abhijeet Ashok [UIT20M1038]
Mr. Phapale Abhijit Balasaheb [UIT20M1047]
Mr. Thorat Dinesh Ashok [UIT20M1065]
(B.TECH IT)
 Abstract

A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

represents a transformative solution to revolutionize attendance management within edu-
cational institutions. This project introduces a new way of tracking student attendance in
schools, moving away from traditional methods like calling names or using cards. We’re
using advanced facial recognition technology, specifically the OpenFace model, to make
attendance more accurate and easier. This system corrects mistakes intelligently and refines
its data for better performance. A distinctive feature of the system involves the incor-
poration of the Haar Cascade model to monitor student engagement. Which does more
than just track attendance. It watches students’ facial expressions and head movements
in real-time. This gives teachers insights into how engaged students are in class, helping
them improve participation. Teachers get immediate feedback, allowing them to adjust
their teaching methods for a more interactive and enjoyable learning experience. Our
system doesn’t just focus on attendance, it takes a broader approach by monitoring student
engagement. By simplifying tasks for school staff and giving helpful information, this
system has the potential to change how schools operate. It uses cutting-edge technology to
make attendance and engagement monitoring easy, promising to create better classroom
experiences for everyone involved.
 Contents

Acknowledgement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I

Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . II

I Introduction and Literature Survey

1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Problem Definition 1
1.2 Relevant Theory 1
1.3 Scope 2
1.4 Objectives 3
1.5 Literature Survey 3

2 Literature Survey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.1 Literature Review 4
2.2 Comparative of Survey 7

II Project Methodology and Requirements

3 Project Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
3.1 Methods and techniques used : 9
3.2 Tools and technologies employed 10
3.3 Data collection methods 10

4 Project Requirements Specifications . . . . . . . . . . . . . . . . . . . . 11

4.1 Functional Requirements 11
4.2 Non Functional Requirements 12
4.3 System Requirements 12
4.3.1 Hardware Requirements: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
4.3.2 Software Requirements: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

III System, Architectural, and Analysis Modeling

5 System Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
5.1 Architectural Design 15
5.2 Proposed System 18

6 Analysis Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
6.1 Behavioral Modeling 21
6.1.1 Use Case Diagram: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
6.1.2 Class Diagram: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
6.1.3 Sequence Diagram: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
6.2 Functional Modeling 25
6.2.1 Activity Diagram : . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
6.3 Architectural Modeling 27
6.3.1 Component Diagram: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
6.3.2 Deployment Diagram: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
6.4 Mathematical Modeling 30

7 Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
7.1 Implementation Details 31
7.2 Challenges faced and how they were addressed 34
7.3 Code Snippets 35
7.4 Detailed Implementation 39

8 Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
8.1 Manual Testing 40
8.1.1 Performance Metrics: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
8.1.2 Test Scenarios . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
8.2 Performance Evaluation 42
 V

9 Result Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
9.1 Presentation of Results 43
9.2 Interpretation and Discussion 48

IV Conclusion and Future Scope

10 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
10.1 Achievement of Objectives 50
10.2 Contributions to the Field 50
10.3 Reflection on Limitations 51
10.4 Conclusion 51
10.5 Future Scope 52

V Reference
11 Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54

Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
Books 55
Articles 55
Project Timeline 62
Plagiarism Report 64
I
Introduction and Literature
Survey

1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . 1
1.1 Problem Definition
1.2 Relevant Theory
1.3 Scope
1.4 Objectives
1.5 Literature Survey

2 Literature Survey . . . . . . . . . . . . . . . . . . 4
2.1 Literature Review
2.2 Comparative of Survey
 1. INTRODUCTION

1.1 Problem Definition

1.Attendance Management: Traditional attendance systems are often inefficient, time-
consuming, and prone to errors. Manual methods like roll call or swipe cards can be
inaccurate and easily manipulated, leading to discrepancies in attendance records.

2.Ensuring Accuracy: Existing biometric systems might not always provide accurate
results due to factors like lighting conditions, angle variations, or image quality. Achieving
high accuracy in facial recognition systems, especially in a dynamic environment like a
classroom, poses a significant challenge.

3.Real-Time Monitoring: Monitoring student activeness in a classroom setting requires

continuous tracking of engagement levels, participation, and attentiveness. Developing a
system capable of real-time monitoring without intrusiveness or hindrance to the learning
process is essential.

1.2 Relevant Theory

Facial recognition technology operates on the principle of analysing and identifying
unique facial features from an image or video frame. It involves complex algorithms that
detect, extract, and compare facial patterns or landmarks, such as the distance between
eyes, shape of the nose, or contours of the face. These algorithms convert these features
into mathematical representations called faceprints, allowing for accurate identification
and verification. Techniques like deep learning, particularly convolutional neural networks
(CNNs), have significantly enhanced facial recognition systems by enabling them to learn
intricate patterns and variations in facial data, improving accuracy even in diverse condi-

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

 1.3 Scope A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

tions. However, challenges persist, such as variations in lighting, pose, facial expressions,
and occlusions, which can impact recognition accuracy. For student activeness monitoring,
computer vision algorithms can analyse visual cues like posture, facial expressions, and
interactions within the classroom to gauge engagement levels. Balancing the power of
these technologies with ethical considerations, especially in handling sensitive biometric
data, is crucial for ensuring privacy and maintaining trust within educational environments.

1.3 Scope
1. Facial Recognition System Development: Designing and implementing a robust
facial recognition system capable of accurately identifying and verifying students’
faces in real-time. This involves exploring and employing state-of-the-art algorithms
for face detection, recognition, and verification.

2. Attendance Tracking: Creating a mechanism to automatically record and manage

student attendance using facial recognition technology. This includes developing
a database to store attendance records and integrating the system with attendance
management software.

3. Activeness Monitoring: Implementing methods to monitor students’ engagement,

participation, and attentiveness during class sessions. This might involve analysing
facial expressions, body language, or other indicators to assess student activeness.

4. Real-Time Data Processing: Building a system capable of processing facial data

in real-time to ensure immediate and accurate attendance tracking and activeness
monitoring during class hours.

5. User Interface Development: Creating user-friendly interfaces for students, teach-

ers, and administrators to interact with the system. This includes dashboards or
applications to view attendance records, monitor student engagement, and manage
system settings.

6. Privacy and Security Measures: Incorporating robust measures to ensure data

privacy and security. This involves implementing encryption techniques, access
controls, and adherence to privacy regulations to protect students’ facial data and
personal information.

7. Integration with Educational Environment: Integrating the system seamlessly

into the educational environment, such as linking it with existing school manage-
ment systems or learning platforms, to streamline administrative tasks and enhance
classroom efficiency.

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

 1.4 Objectives A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

1.4 Objectives
The main objectives of this application are

1. Develop Robust Facial Recognition: Create a reliable and accurate facial recogni-
tion system capable of identifying students in real-time, considering various environ-
mental factors such as lighting, angles, and occlusions.

2. Automate Attendance Management: Implement an automated attendance system

that utilizes facial recognition technology to record and track student attendance
without manual intervention, reducing errors and enhancing efficiency.

3. Real-Time Activeness Monitoring: Develop algorithms or mechanisms to monitor

student engagement and attentiveness during class sessions, providing real-time
feedback to educators about participation levels.

4. User-Friendly Interface: Create an intuitive and user-friendly interface for both

students and educators to interact with the system, allowing easy access to attendance
records and activeness reports.

5. Integration with Educational Systems: Ensure seamless integration of the system

into existing educational frameworks, allowing compatibility with different platforms
and facilitating easy adoption by educational institutions.

1.5 Literature Survey

The literature survey for a Face Recognition-Based Smart Student Attendance and
Activeness Monitoring System delves into a comprehensive exploration of existing
research across key components. The review begins by scrutinizing various face
recognition techniques, ranging from traditional methods such as eigenface and
Fisherface to modern deep learning approaches like Convolutional Neural Networks
(CNNs) and Siamese Networks. Seminal works contributing to the advancements in
face recognition technology are identified and discussed. Furthermore, the survey
examines the utilization of Haar Cascade classifiers for face detection, exploring
studies that introduced Haar-like features and cascading classifiers.
This discussion encompasses the application of Haar Cascade classifiers in real-time
scenarios and their performance in diverse conditions. Moving on to attendance
management systems, the literature survey traces the evolution from manual methods
to automated systems, highlighting the advantages and challenges associated with
each approach. Noteworthy studies that integrate biometric methods, particularly
face recognition, into attendance management are explored, shedding light on the
impact of such integrations on accuracy and efficiency.

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

 2. Literature Survey

2.1 Literature Review

[start with your base paper, followed by min 10 to 15 paper details must be their]

1. Paper Details: “Deep Learning Based Real Time Face Recognition For University
Attendance System”, Mansi Singhal, Gufran Ahmad, IEEE Access, June 2023.

About Paper : Mansi Singhal and Gufran Ahmad delve into the challenges faced
by current face recognition systems, particularly in addressing pose, lighting, and
expression variations in real-world scenarios. Their research focuses on the appli-
cation of advanced algorithms and machine learning techniques, specifically deep
convolutional neural networks. The study involves the creation of a unique dataset
from real-time videos, systematically processed for face detection, labeling, tracing,
and refinement. Notably, the researchers fine-tune a convolutional neural network
using this annotated dataset, showcasing a comprehensive approach to overcoming
the limitations of existing face recognition systems.

Findings: The experimental findings from the attendance management system’s

testing dataset reveal the significant efficacy of their deep learning approach. The fine-
tuned convolutional neural network achieves an impressive accuracy of 95.3 percent,
surpassing the non-fine-tuned network’s accuracy of 81.4 percent. This underscores
the effectiveness of utilizing deep learning methodologies, particularly in real-time
video data scenarios, and highlights the potential for substantial advancements in
face recognition capabilities.

2. Paper Details: “Surveillance System for Real-Time High-Precision Recognition

of Criminal Faces From Wild Videos”, Hyun-Bin Kim, Nakhoon Choi, Hye-Jeong

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

2.1 Literature Review A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

Kwon and Heeyoul Kim, IEEE Access, volume 11 , pp. 56066-56082 May 2023.

About Paper : In their paper, Hyun-Bin Kim, Nakhoon Choi, Hye-Jeong Kwon, and
Heeyoul Kim tackle the pressing issue of preventing repeated offenses by violent
criminals, with a specific focus on locations such as schools and childcare centers.
Recognizing the limitations of current surveillance systems, the authors propose an
innovative approach leveraging deep face recognition technology. The method facil-
itates real-time identification of criminal faces from surveillance videos, ensuring
prompt notifications to relevant authorities. Their methodology involves the use of
down-sampled images for efficient face detection, coupled with the integration of
a sophisticated face tracking ID unit to improve accuracy and minimize prediction
errors.

Findings : The paper’s findings highlight the effectiveness of their proposed ap-
proach in crime prevention. By ensuring swift identification and localization of
criminal faces, the system provides an innovative solution to the challenges posed
by repeated offenses, especially in vulnerable locations. The down-sampled image
technique enhances overall responsiveness, contributing to the accuracy and reliabil-
ity of criminal identification in real-world scenarios.

3. Paper Details: “Face Recognition Attendance System Based on Real-Time Video

Processing”, Hao yang1, xiaofeng han, IEEE Accesss, volume 8, pp. 159143-159150
June 2020.

About Paper: Hao Yang and Xiaofeng Han’s research explores the potential of
face recognition technology within the context of an attendance system based on
real-time video processing. Focusing on the increasing demand for accurate and
efficient attendance management, the study addresses four key aspects: the accuracy
and stability of the face recognition system during check-in, the effectiveness of
reducing truancy, and the user interface design for a seamless attendance experience.
Through in-depth analysis and experimentation, the study introduces the concept of
a face recognition attendance system and delves into the practical implementation of
real-time video processing.

Findings: Experimental results demonstrate the impressive accuracy rate of 82

percent for the video face recognition system. This leads to a substantial reduction
of approximately 60 percent in check-in time compared to traditional methods. The
system proves effective in mitigating issues such as student truancy, early departures,
and class disruptions. These findings underscore the potential of face recognition
technology to significantly enhance overall attendance efficiency, contributing to the
streamlined functioning of educational environments.

4. Paper Details: “Face Detection and Recognition Using Face Mesh and Deep Neural
Network”, Shumbul Arifa, Keerti Chaudhary, Procedia Computer Science 218, pp.

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

2.1 Literature Review A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

741-750 2023.

About Paper: Shumbul Arifa and Keerti Chaudhary focus their paper on implement-
ing a Smart Attendance System utilizing OpenCV’s Face Recognition capabilities.
Emphasizing the human face as a crucial identifier and communication tool, their
proposed system aims to assist lecturers in automatically detecting students’ faces
within a classroom environment and recording attendance based on the recognized
faces.

Findings: The paper underscores the significance of OpenCV’s Face Recognition

capabilities in developing a practical solution for managing class attendance. While
specific experimental findings are not provided in the given text, the implication
is that leveraging OpenCV’s capabilities enhances the efficiency and automation
of attendance recording through facial recognition, contributing to a streamlined
classroom management process.

5. Paper Details: "Face Detection using Haar Cascades to Filter Selfie Face Image on
Instagram", Adri Priadana, Muhammad Habibi, IEEE Access, 13-15 March 2019

About Paper:Adri Priadana and Muhammad Habibi’s research proposes a method

to filter selfie images on Instagram by integrating web data extraction and the Haar
Cascade method for human face detection. The study addresses the issue of unrelated
hashtags overshadowing background details in images, particularly those dominated
by selfie faces. Their hybrid approach offers a promising solution for refining image
searches, enhancing content relevance, and improving the user experience on the
platform.

Findings:The experimental results of the proposed method showcase its accuracy,

with 71.48 percent for human face detection and 64.6 percent for filtering selfie
images. This hybrid approach leverages the assumption that selfies inherently contain
human faces. The findings suggest the effectiveness of the method in filtering and
enhancing image searches on Instagram, contributing to improved content relevance
and user experience by focusing on the human faces within the images.

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

 2.2 Comparative of Survey
A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

2.2 Comparative of Survey

1. Accuracy of Facial Recognition:

• Evaluate the accuracy rates of different facial recognition algorithms and

systems.
• Compare the performance in varying lighting conditions, angles, and image
qualities.

2. Real-Time Monitoring Capabilities:

• Analyse the systems’ abilities to monitor and track student activeness in real
time.
• Compare methods used for measuring engagement levels, participation, and
attentiveness.

3. Attendance Management Efficiency:

• Assess the efficiency of attendance tracking in terms of speed and accuracy.

• Compare the systems in managing large volumes of data and processing atten-
dance records.

4. User-Friendliness and Integration:

• Assess the ease of use for both students and educators.

• Evaluate how seamlessly the system integrates with existing educational frame-
works or technologies.

5. Cost and Scalability:

• Compare the initial setup costs and maintenance expenses of each system.
• Assess the scalability of the solution for different educational institutions or
classroom sizes.

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

II
Project Methodology and
Requirements

3 Project Methodology . . . . . . . . . . . . . 9
3.1 Methods and techniques used :
3.2 Tools and technologies employed
3.3 Data collection methods

4 Project Requirements Specifications

11
4.1 Functional Requirements
4.2 Non Functional Requirements
4.3 System Requirements
 3. Project Methodology

3.1 Methods and techniques used :

1. Facial Recognition Algorithms: Utilizing algorithms like Convolutional Neural
Networks (CNNs), Eigenfaces, Fisherfaces, or Deep Learning frameworks (such as
TensorFlow or PyTorch) for accurate detection, identification, and verification of
students’ faces.

2. Feature Extraction: Techniques to extract and represent facial features, such as

landmarks, textures, or key points, to create a unique representation of each student’s
face for identification purposes.

3. Data Preprocessing: Cleaning, normalization, and augmentation of facial images

to improve recognition accuracy, including tasks like resizing, grayscale conversion,
noise reduction, and alignment.

4. Machine Learning Models: Implementing classification and regression models

to analyse and predict student activeness based on various features, such as facial
expressions, head movements, eye gaze, or posture.

5. Real-Time Monitoring Integration of video processing and real-time image analysis

to continuously monitor student behaviour and engagement during class sessions.

6. Database Management: Developing a secure database to store facial data and

attendance records, implementing encryption and access control mechanisms to
ensure data privacy.

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

 3.2 Tools and technologies employed
A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

3.2 Tools and technologies employed

1. Facial Recognition Libraries and APIs:
• OpenCV: A popular open-source computer vision library that offers various
tools and algorithms for facial recognition.
• Dlib: Another powerful library with facial recognition capabilities, often used
for face detection and landmark estimation.
• FaceNet: Deep learning model for face recognition that can be integrated into
the system.
2. Machine Learning and Deep Learning Frameworks:
• TensorFlow or PyTorch: Widely used deep learning frameworks that provide
tools for building and training neural networks, essential for developing facial
recognition models.
• Keras: A high-level neural networks API, often used in conjunction with
TensorFlow for rapid development and experimentation.
3. Database Management:
• MS Excel : excel is used for the data storage.

3.3 Data collection methods

Creating a Traffic Sign Board Recognition and Alert System using R-CNN (Region-based
Convolutional Neural Network) involves collecting and preparing data for training and
evaluation. Here are some common data collection methods for such a system:
1. Facial Image Acquisition: Using cameras or webcam feeds installed in classrooms
or entry points, images of students’ faces are captured for recognition and attendance
purposes. These images need to cover various angles, lighting conditions, and facial
expressions for robust recognition.
2. Database Creation: Creating a database of facial images to train the recognition
algorithm. This database comprises images of enrolled students and possibly staff
members, collected through consent and with privacy considerations.
3. Real-Time Image Capture: Continuous real-time capture of students’ facial images
during class hours for attendance tracking and monitoring activeness. This requires
a system that can process and analyse live video feeds.
4. Data Augmentation: Techniques such as data augmentation might be employed to
increase the diversity and quality of the dataset used for training the facial recognition
model. This involves techniques like rotation, scaling, and adding noise to existing
images to make the recognition system more robust.
5. Activeness Monitoring Data: Apart from facial images, additional data might be
collected to monitor activeness, such as video feeds to track movements, interactions
with learning materials, or even sensors to detect engagement levels (e.g., eye
tracking for attention).

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

 4. Project Requirements Specifications

4.1 Functional Requirements

1. User Authentication and Registration: Registering students and faculty members
into the system with unique identifiers. Authenticating users through secure login
credentials or biometric data.

2. Face Recognition for Attendance: Capturing and analysing facial features for
attendance tracking. Identifying registered students through facial recognition.
Recording attendance automatically upon recognizing a student’s face.

3. Real-time Monitoring of Activeness: Monitoring students’ attentiveness and par-

ticipation during class sessions. Analysing facial cues or other indicators to gauge
engagement levels. Providing real-time feedback or alerts to educators about student
activeness.

4. Attendance Reporting: Generating comprehensive reports of attendance records

for individual students and classes. Offering summaries and detailed breakdowns of
attendance statistics over time.

5. Integration with Educational Systems: Integrating with existing school or univer-

sity databases for student information. Syncing attendance records with administra-
tive systems for easy access and management.

6. Adaptability and Scalability: Supporting multiple classrooms or educational set-

tings simultaneously. Scalability to accommodate varying class sizes and schedules.

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

 4.2 Non Functional Requirements
A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

4.2 Non Functional Requirements

1. Accuracy and Reliability: The system should achieve a high level of accuracy
in recognizing and logging student attendance using facial recognition, aiming for
minimal false positives or negatives.

2. Performance: It should handle simultaneous attendance tracking for multiple stu-

dents in real-time without significant delays or system lags.

3. Security and Privacy: The system must adhere to strict security protocols to
prevent unauthorized access to student data. It should comply with privacy laws and
regulations, ensuring the protection of students’ personal information.

4. Scalability: The system should be scalable to accommodate a varying number of

students and classrooms. It should perform consistently regardless of the class size.

5. Usability and User Experience: The interface should be intuitive and user-friendly
for both teachers and students. It should require minimal training for users to operate
efficiently.

6. Adaptability and Compatibility: The system should be adaptable to different

environments and lighting conditions commonly found in classrooms. It should also
integrate seamlessly with existing school management systems or databases.

7. Robustness and Error Handling: The system should be resilient to errors such as
poor image quality, variations in facial expressions, or unexpected environmental
changes. It should handle such situations gracefully without compromising accuracy.

4.3 System Requirements

4.3.1 Hardware Requirements:
1. Camera(s): High-resolution cameras capable of capturing clear facial images in
various lighting conditions. Multiple cameras may be needed for larger classrooms
or different angles.

2. Computing Device: A computer or server to process facial recognition algorithms

and store data. It should have adequate processing power and memory.

3. Network Infrastructure: Reliable internet connectivity for real-time data process-

ing and storage. Wi-Fi or Ethernet connectivity for data transfer.

4. Display Unit: If required, a display unit (such as a monitor or screen) to show

attendance and monitoring information for teachers or administrators.

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

 4.3 System RequirementsA Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

4.3.2 Software Requirements:

1. Facial Recognition Algorithm: Implementing a robust facial recognition algorithm
like OpenCV, Dlib, or TensorFlow for accurate face detection and recognition.

2. Database Management System: A database to store and manage student informa-

tion, attendance records, and activeness metrics.

3. Programming Language: Proficiency in programming languages like Python for

developing software applications.

4. Real-Time Monitoring System: Software capable of real-time monitoring and

analysis of student engagement and attentiveness. This might involve machine
learning or AI models to assess activity levels or attention span.

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

III
System, Architectural, and
Analysis Modeling

5 System Design . . . . . . . . . . . . . . . . . . . . 15
5.1 Architectural Design
5.2 Proposed System

6 Analysis Modeling . . . . . . . . . . . . . . . 21
6.1 Behavioral Modeling
6.2 Functional Modeling
6.3 Architectural Modeling
6.4 Mathematical Modeling

7 Implementation . . . . . . . . . . . . . . . . . . 31
7.1 Implementation Details
7.2 Challenges faced and how they were ad-
dressed
7.3 Code Snippets
7.4 Detailed Implementation

8 Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
8.1 Manual Testing
8.2 Performance Evaluation

9 Result Analysis . . . . . . . . . . . . . . . . . . . . 43
9.1 Presentation of Results
9.2 Interpretation and Discussion
 5. System Design

5.1 Architectural Design

Figure 5.1: System Architecture

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

5.1 Architectural Design A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

a. Image Capture Module:

• Responsible for capturing real-time images from the camera.

• Utilizes OpenCV or a similar library for interfacing with the camera.

• Transmits captured frames to the preprocessing module.

b. Preprocessing Module:

• Applies image processing techniques to enhance quality and reduce noise.

• Normalizes images, performs resizing, and applies filters for improved feature
extraction.

• Transmits preprocessed images to the Face Detection Module.

c. Face Detection Module:

• Implements a face detection algorithm to identify and localize faces in the images.

• Utilizes pre-trained models like Haarcascades or deep learning models (e.g., MTCNN).

• Sends the detected face regions to the Student Registration Module.

d. Student Registration Module:

• Allows administrators to register students by capturing facial features and associating

them with student information.

• Stores student information, facial features, and unique identifiers in the Database
Management System.

e. Feature Extraction Module:

• Extracts facial features using deep learning models (e.g., CNN) or facial landmark
detection algorithms.

• Normalizes and represents features for storage and comparison.

• Transmits the extracted features to the Database Management System.

f. Database Management System:

• Manages student information, facial features, and attendance records.

• Utilizes a relational database (e.g., SQLite, MySQL) with a well-defined schema.

• Supports efficient indexing and retrieval of data.

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

5.1 Architectural Design A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

g. Recognition Module:

• Compares the extracted features with those stored in the database using algorithms
like Euclidean distance or cosine similarity.

• Determines the identity of the recognized face.

• Sends recognition results to the Biometric Indicators and Activeness Monitoring

Modules.

h. Biometric Indicators and Activeness Monitoring Modules:

• Utilizes eye and mouth detection algorithms to monitor biometric indicators.

• Analyzes facial expressions to detect yawning and closed eyes.

• Marks attendance based on recognition results and activeness indicators.

i. User Interface:

• Provides a user-friendly interface for system control and monitoring.

• Developed using frameworks like Tkinter for Python or other GUI libraries.

• Displays real-time monitoring, attendance logs, and system settings.

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

 5.2 Proposed System A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

5.2 Proposed System

Figure 5.2: Proposed system architecture for face recognition and attendance marking

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

5.2 Proposed System A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

Figure 5.3: Proposed system for activeness detection

The proposed system is designed to be a comprehensive and intelligent Face

Recognition-Based Smart Student Attendance and Activeness Monitoring System that
leverages cutting-edge technologies for accurate and efficient monitoring in educational
environments. The system comprises several key components and features:

1. Image Capture Module:

• Utilizes OpenCV to capture high-resolution images from the camera in real-time.
• Ensures continuous and reliable image acquisition for accurate processing.

2. Preprocessing Module:
• Applies advanced image processing techniques using OpenCV to enhance image
quality.
• Implements normalization, resizing, and denoising algorithms to ensure consistency
and reduce noise.

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

5.2 Proposed System A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

3. Face Detection Module:

• Implements a robust face detection algorithm using a combination of pre-trained
models like Haarcascades and deep learning models such as MTCNN.
• Addresses challenges related to pose variations and diverse facial expressions.

4. Student Registration Module:

• Develops a user-friendly registration interface for administrators to capture facial
features and associate them with student information.
• Stores student details, unique identifiers, and facial features in a secure and scalable
database management system.

5. Feature Extraction Module:

• Utilizes state-of-the-art deep learning models, such as Convolutional Neural Net-
works (CNNs), for accurate and discriminative feature extraction.
• Ensures the normalization and representation of facial features suitable for compar-
ison.

6. Database Management System:

• uses the MS Exel for storing and managing student information, facial features,
and attendance records.
• Incorporates indexing and efficient retrieval mechanisms for quick access to data.

7. Recognition Module:
• Employs advanced recognition algorithms that consider factors like Euclidean
distance or cosine similarity for accurate identification.
• Allows for real-time recognition and marking of student attendance.

8. Biometric Indicators and Activeness Monitoring Modules:

• Integrates eye and mouth detection algorithms to monitor biometric indicators.
• Analyzes facial expressions in real-time to detect yawning and closed eyes for
activeness monitoring.

9. Attendance Marking and Logging:

• Marks attendance for recognized and active students based on recognition results
and activeness indicators.
• Logs attendance records securely and provides administrators with access to
comprehensive logs.

10. User Interface:

• Designs an intuitive and user-friendly interface using frameworks like Tkinter or
other GUI libraries.
• Displays real-time monitoring, attendance logs, and system settings for ease of
control and monitoring.

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

 6. Analysis Modeling

6.1 Behavioral Modeling

Behavioral modeling refers to the creation of models that represent and simulate the
behavior of systems, processes, or entities. This type of modeling is used in various
fields, including psychology, economics, sociology, and computer science. The goal is to
understand, predict, or analyze the behavior of individuals, groups, or systems in different
contexts.

Behavioral modeling theory seeks to understand and explain human behavior by

examining the influence of various factors on individuals and groups. At its core, this
theory posits that behavior is shaped by a combination of internal and external influences,
including cognitive processes, social interactions, and environmental stimuli. The be-
havioral modeling approach emphasizes observable actions and reactions, focusing on
the learned behaviors and responses that result from repeated exposure to certain stimuli
or experiences. Drawing from principles of conditioning and reinforcement, this theory
suggests that individuals learn through observation, imitation, and the consequences of
their actions. Social cognitive theory, a key component of behavioral modeling, highlights
the role of observational learning and emphasizes the importance of role models and the
social context in shaping behavior.

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

 6.1 Behavioral Modeling A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

6.1.1 Use Case Diagram:

An illustration of a user’s potential interactions with a technology is called a use case
diagram. A use case diagram, which is frequently supplemented by other types of diagrams
as well, illustrates the numerous use cases and user types that the system has. Either circles
or ellipses are used to symbolise the use cases. Stick representations of the actors are
frequently used.

Figure 6.1: Usecase Diagram

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

 6.1 Behavioral Modeling A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

6.1.2 Class Diagram:

A class diagram is a kind of diagram that models a system’s classes, properties, op-
erations, and relationships between them to illustrate the structure of the system. One of
the most popular kinds of diagrams in the standard method for defining object-oriented
systems, UML (Unified Modelling Language), is this one. You may build the system
architecture, describe the system specifications, and comprehend the code structure with
the aid of a class diagram.

Figure 6.2: Class Diagram

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

 6.1 Behavioral Modeling A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

6.1.3 Sequence Diagram:

A system sequence diagram (SSD), also known as a sequence diagram in software
engineering, displays process interactions ordered in a temporal sequence. The objects,
procedures, and messages that are exchanged in order to perform the functionality are all
shown in the diagram. Sequence diagrams are commonly linked to the realisations of use
cases in the 4+1 architectural perspective model of the system that is being developed.
Event diagrams or event scenarios are other names for sequence diagrams.

Figure 6.3: Sequence Diagram

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

 6.2 Functional Modeling A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

6.2 Functional Modeling

Functional modeling theory is a conceptual framework that focuses on representing
and understanding the functions of a system or process. It provides a structured approach to
analyze and describe the purpose and behavior of a system, emphasizing the relationships
between its components and how they contribute to achieving specific goals.

In functional modeling, the system is decomposed into functional elements, and

their interactions are defined in terms of inputs, outputs, and the transformation of inputs
into desired outputs. The theory often employs graphical representations, such as func-
tional flow diagrams or IDEF0 diagrams, to illustrate the flow of information and activities
within the system. By emphasizing the functions rather than the physical components,
functional modeling theory allows for a more abstract and comprehensive understanding
of complex systems, aiding in problem-solving, system design, and optimization. This the-
oretical framework is widely applied in various domains, including engineering, business
process modeling, and system analysis, to facilitate a systematic and structured approach
to understanding and improving the functionality of systems.

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

 6.2 Functional Modeling A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

6.2.1 Activity Diagram :

Activity diagrams, which allow for choice, iteration, and concurrency, are graphi-
cal depictions of workflows consisting of sequential activities and actions [1]. Activity
diagrams in the Unified Modelling Language are meant to represent organisational and
computational processes, or workflows, as well as the data flows that cross over into the
associated activities. Activity diagrams can incorporate features that illustrate the data flow
between activities via one or more data stores, even if their main purpose is to depict the
overall flow of control.

Figure 6.4: Activity Diagram

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

 6.3 Architectural Modeling
A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

6.3 Architectural Modeling

Architectural modeling theory encompasses the principles and methodologies guiding
the creation of representations that capture the essential aspects of a system’s architec-
ture. At its core, architectural modeling serves as a communication and abstraction tool,
facilitating the comprehension and analysis of complex systems by various stakeholders.
The fundamental objective is to create a conceptual blueprint that not only visualizes the
structure and behavior of the system but also aids in decision-making processes throughout
its lifecycle.

The Unified Modeling Language (UML) is a widely adopted standard for archi-
tectural modeling, providing a set of standardized diagrams and notations. Architectural
models typically consist of multiple views, each addressing specific concerns such as
functionality, structure, or behavior. Moreover, they evolve alongside the system, enabling
iterative refinement and adaptation to changing requirements. The process of architectural
modeling involves identifying key elements, their relationships, and interactions, fostering
a shared understanding among architects, developers, and other stakeholders. By adhering
to sound architectural modeling principles, such as modularity, abstraction, and trace-
ability, practitioners can enhance the clarity and effectiveness of their designs, ultimately
contributing to the success of complex software and system development projects.

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

 6.3 Architectural Modeling
A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

6.3.1 Component Diagram:

A component diagram in Unified Modelling Language (UML) shows how parts
are connected to create bigger parts or software systems. They serve as an example
of how systems of any complexity are put together. A component diagram makes it
possible to confirm that the necessary functionality of a system is appropriate. In addition,
these diagrams serve as a communication tool for system stakeholders and developers.
Programmers and developers can make better decisions regarding task assignment and
necessary skill upgrades by using the diagrams to formalise an implementation roadmap.

Figure 6.5: Component Diagram

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

 6.3 Architectural Modeling
A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

6.3.2 Deployment Diagram:

The hardware that will be used to run the software is shown in the deployment diagram.
It illustrates a system’s static deployment perspective. The nodes and their connections
are involved. It determines the hardware’s software deployment strategy. It links the
design-created software architecture to the physical system architecture, in which the
programme will operate as a node. Communication channels are used to illustrate the link
because there are numerous nodes involved.

Figure 6.6: Deployment Diagram

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

 6.4 Mathematical Modeling
A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

6.4 Mathematical Modeling

1. Image Acquisition:

• The system captures high-resolution images at regular intervals using a camera

positioned in the classroom.
• Consideration is given to lighting conditions and camera placement to ensure
optimal image quality.

2. Face Detection:

• The Haar Cascade classifier analyzes each captured image, identifying regions
that likely contain faces based on predefined patterns.
• False positives and negatives are minimized through careful tuning and testing
of the classifier.

3. Face Recognition:

• A pre-trained face recognition library is employed to compare the detected

faces with a database of known student faces.
• The system utilizes sophisticated algorithms to compute facial feature vectors
for accurate recognition.
• Facial landmarks and unique identifiers contribute to the precision of the
recognition process.

4. Attendance Management:

• Enrolled students are pre-registered in the system with their unique identifiers.
• The system updates attendance records based on the successful recognition of
student faces, including timestamps.
• Consideration is given to handling cases of multiple students in a single frame
and minimizing false attendance records.

5. Activeness Monitoring:

• Additional features like eye gaze tracking, head pose estimation, and facial
expression analysis are integrated to assess student engagement.
• These features contribute to a comprehensive activeness metric, providing
insights into the level of attention and participation.

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

 7. Implementation

7.1 Implementation Details

1. Image Capture:
Utilize a programming language like Python with libraries such as OpenCV to
capture images from the camera.
• Use OpenCV to access the camera and capture frames.
• Set up a continuous image capture process for real-time monitoring.
• Implement error handling to address issues related to camera availability or failure.

2. Pre-Processing:
Apply image processing techniques using OpenCV to enhance image quality and
reduce noise.
• Perform operations such as resizing, normalization, and grayscale conversion to
standardize image dimensions and format.
• Apply filters or denoising techniques to improve image quality.
• Consider histogram equalization for better contrast.

3. Face Detection:
Implement a face detection algorithm, possibly using pre-trained models like Haar-
cascades or deep learning models (e.g., MTCNN, OpenCV’s DNN module).
• Select a suitable face detection method based on accuracy and computational
efficiency.
• Fine-tune parameters for face detection to accommodate different lighting condi-
tions and angles.
• Consider a deep learning-based approach for more robust face detection in varying
scenarios.

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

7.1 Implementation Details
A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

4. Student Registration:
Develop a registration module to capture and store facial features along with student
information in a database.
• Create a registration interface for administrators to input student information.
• Capture multiple images per student for better feature extraction.
• Store information such as student ID, name, and facial features in a secure and
scalable database.

5. Feature Extraction (1-8-1):

Extract facial features using deep learning models (e.g., CNN) or facial landmark
detection algorithms.
• Employ pre-trained deep learning models for facial feature extraction or use facial
landmark detection algorithms.
• Extract features such as the distance between eyes, nose shape, and mouth geome-
try.
• Normalize and represent features in a format suitable for comparison.

6. Database Management:
Implement a database schema to store and manage student information and facial
features.
• Design a database schema that efficiently stores and retrieves student information.
• Ensure data integrity and implement indexing for faster queries.
• Regularly back up the database to prevent data loss.

7. Recognition:
Develop a recognition module to compare extracted features with those stored in the
database, using algorithms like Euclidean distance or cosine similarity.
• Implement a recognition algorithm that compares facial features.
• Set a threshold for similarity to determine a match.
• Consider incorporating machine learning models for improved recognition accu-
racy.

8. Biometric Indicators (Eye and Mouth):

Implement eye and mouth detection using appropriate algorithms or pre-trained
models.
• Use Haarcascades or deep learning models to detect eyes and mouth in real-time.
• Monitor changes in facial expressions for signs of attentiveness or drowsiness.

9. Yawn Detection:
Analyze mouth movements to detect yawning. This might involve measuring changes
in mouth geometry over time.

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

7.1 Implementation Details
A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

• Track changes in the mouth region over consecutive frames.

• Set thresholds for mouth opening to identify yawning.
• Integrate this information into the activeness monitoring system.

10. Attendance Marking:

Mark attendance for recognized and active students based on the results of the
recognition and activeness detection modules.
• Design a mechanism to record attendance in a centralized log.
• Integrate the attendance marking process with the recognition and activeness detec-
tion results.

11. Closed Eyes Detection: Implement a module to check for closed eyes and mark
students as "Not Active".
• Use eye detection algorithms to identify closed eyes.
• Set a threshold for eye closure to determine inactivity.
• Integrate this information into the activeness monitoring system.

12. User Interface:

Design a user interface for system control and monitoring, using frameworks like
Tkinter for Python or other GUI libraries.
• Create an intuitive interface for administrators to manage and monitor the system.
• Include features such as real-time monitoring, attendance logs, and system settings.
• Ensure responsiveness and ease of use for efficient system control.
Throughout the implementation process, conduct thorough testing and iterate on the
system design based on feedback and real-world performance. Keep security and
privacy considerations in mind, especially when dealing with sensitive biometric
data.

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

 7.2 Challenges faced and how
A Face they were
Recognition-Based addressed
Smart Student Attendance and Activeness Monitoring System

7.2 Challenges faced and how they were addressed

1. Limited Lighting Conditions:
Challenge: Inconsistent or low lighting conditions affecting the accuracy of face
detection and recognition
Addressing Strategy:Implement adaptive image processing techniques to handle
varying lighting conditions. Use image enhancement methods, and consider addi-
tional sensors like infrared for improved performance in low-light environments.

2. Pose Variability:
Challenge: Difficulty in recognizing faces at different angles or poses.
Addressing Strategy: Augment the training dataset with images captured at various
angles and poses. Consider using advanced face detection models that are robust to
pose variations.

3. Real-time Processing and Inference:

Challenge:Processing each frame in real-time can be computationally intensive.
Addressing Strategy: Optimize algorithms for speed and efficiency. Utilize hard-
ware acceleration (GPU) if possible. Implement techniques like frame skipping for
less critical frames.

4. Diverse Facial Expressions:

Challenge:Variability in facial expressions impacting the accuracy of feature extrac-
tion.
Addressing Strategy: Expand the dataset to include images of students with dif-
ferent expressions. Train the model to recognize a range of expressions for better
activeness monitoring.

5. Data Privacy Concerns:

Challenge:Balancing the need for accurate monitoring with privacy concerns.
Addressing Strategy: Implement strict access controls and encryption for stored
data. Anonymize data where possible. Ensure compliance with privacy regulations
and obtain informed consent.

6. Hardware Compatibility:
Challenge:Ensuring the system works seamlessly across different cameras and hard-
ware configurations.
Addressing Strategy: Test the system on various hardware setups. Implement
modular design principles to facilitate easy integration with different camera models
and configurations.

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

7.3 Code Snippets A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

7.3 Code Snippets

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

7.3 Code Snippets A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

7.3 Code Snippets A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

7.3 Code Snippets A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

7.4 Detailed Implementation
A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

7.4 Detailed Implementation

Initialization and Libraries
At the beginning of the script, various Python libraries are imported to support
different functionalities. These libraries include tools for subprocess management,
GUI development, computer vision tasks, data handling, and time-related operations.
Each library serves a specific purpose within the Lecture Attendance System, con-
tributing to tasks such as facial landmark detection, face recognition, data storage,
and user interface creation.

Facial Analysis and Recognition

The script utilizes advanced techniques for facial analysis and recognition. It em-
ploys algorithms from the dlib library to detect facial landmarks and faces within
images or video frames. These landmarks are crucial for identifying key facial
features like eyes, lips, and overall facial expression. Additionally, face recognition
capabilities are implemented using specialized algorithms to compare face encodings
and determine the identity of individuals, enabling accurate attendance tracking and
monitoring.

Graphical User Interface and Interaction

Tkinter, a Python library for creating graphical user interfaces, is utilized to design
an intuitive interface for users. The GUI includes entry fields for inputting lecture
details and a start button to initiate the attendance recording process. This graphical
interface enhances user interaction by providing a visually appealing and user-
friendly platform for managing lecture attendance and monitoring attentiveness in
real time.

Data Management and Automation

The script incorporates mechanisms for efficient data management and automation.
It utilizes serialization techniques to store face encodings and attendance records,
ensuring data integrity and accessibility. Additionally, the system leverages sub-
process management to seamlessly execute subsequent actions, demonstrating the
script’s capability to automate processes and integrate with external functionalities
for a streamlined lecture attendance management system.

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

 8. Testing

8.1 Manual Testing

8.1.1 Performance Metrics:
Accuracy: Percentage of correct identifications.
False Acceptance Rate (FAR): Percentage of unauthorized individuals incorrectly
identified.
False Rejection Rate (FRR): Percentage of authorized individuals incorrectly
rejected.
Recognition Speed: Time taken to process each face.
Scalability: Ability to handle large datasets and concurrent requests.

8.1.2 Test Scenarios

Table 8.1: Test Scenario: Normal Lighting Conditions

Objective Evaluate accuracy and recognition speed under normal lighting
Metric Result
Accuracy 95%
Recognition Speed Min: 0.5s, Max: 1.2s, Avg: 0.8s
FAR <1%
FRR <5%
Scalability High

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

8.1 Manual Testing A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

Table 8.2: Test Scenario: Low Lighting Conditions

Objective Evaluate accuracy and recognition speed under low lighting
Metric Result
Accuracy 80%
Recognition Speed Min: 0.8s, Max: 1.5s, Avg: 1.2s
FAR Up to 5%
FRR Up to 10%
Scalability Moderate

Table 8.3: Test Scenario: Different Camera Angles

Objective Evaluate accuracy and recognition speed with different angles
Metric Result
Frontal Accuracy 90%
Side Accuracy 70%
Overhead Accuracy 80%
Recognition Speed Frontal: 0.6s, Side: 0.9s, Overhead: 0.8s
FAR Up to 3%
FRR Up to 7%
Scalability Moderate

Table 8.4: Test Scenario: Partial Face Occlusions

Objective Evaluate accuracy and recognition speed with face occlusions
Metric Result
Glasses Accuracy 85%
Scarf Accuracy 70%
Hat Accuracy 75%
Recognition Speed Glasses: 0.7s, Scarf: 1.1s, Hat: 0.9s
FAR Up to 2%
FRR Up to 5%
Scalability Moderate

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

8.2 Performance Evaluation
A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

8.2 Performance Evaluation

Accuracy vs Training Iterations
The graph shows the trend of accuracy (%) over multiple training iterations. As the
number of training iterations increases, the accuracy of the face recognition system
improves.

False Acceptance Rate (FAR) vs Training Iterations

The graph illustrates the trend of the false acceptance rate (%) over multiple training
iterations. A decreasing FAR indicates that the system is becoming more selective in
correctly identifying authorized individuals and reducing the number of unauthorized
individuals incorrectly identified.

False Rejection Rate (FRR) vs Training Iterations

The graph demonstrates the trend of the false rejection rate (%) over multiple training
iterations. A decreasing FRR indicates that the system is becoming more accurate in
correctly identifying authorized individuals and reducing the number of authorized
individuals incorrectly rejected.

Recognition Speed vs Training Iterations

The graph shows the trend of recognition speed (seconds) over multiple training
iterations. A decreasing recognition speed indicates that the system is becoming
more efficient in processing each face for recognition.

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

 9. Result Analysis

9.1 Presentation of Results

Figure 9.1: Initial Project Start Screen

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

9.1 Presentation of Results
A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

Figure 9.2: Student Registration Screen

Figure 9.3: Registered Students Faces are Stored in Folder

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

9.1 Presentation of Results
A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

Figure 9.4: Start New Lecture Screen

Figure 9.5: Student Face Detected

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

9.1 Presentation of Results
A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

Figure 9.6: Student is Yawning so his Attentiveness is 0

Figure 9.7: Student eye is closed so his Attentiveness is 0

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

9.1 Presentation of Results
A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

Figure 9.8: The Attendance of students is recorded, and their activeness percentage is
determined based on the count of yawns and instances of closed eyes

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

9.2 Interpretation and Discussion
A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

9.2 Interpretation and Discussion

• Attendance Accuracy and Efficiency:
The system demonstrated high accuracy in tracking student attendance, achiev-
ing an average accuracy rate of over 95% across multiple testing sessions.
This level of accuracy surpassed manual attendance methods, which often
suffer from human errors and time-consuming processes. The efficiency of
the system in automating attendance recording significantly reduced adminis-
trative burdens on educators, allowing them to focus more on teaching tasks
rather than manual record-keeping.

• Activeness Monitoring Insights:

The activeness monitoring feature of the system provided valuable insights into
student engagement during lectures. By detecting closed eyes and yawning,
the system assessed student attentiveness in real-time, generating activeness
percentages for each student throughout the lecture duration. The analysis
revealed varying levels of attentiveness among students, with an average active-
ness rate of 75%, indicating a generally high level of engagement. However,
occasional dips in activeness were observed during certain lecture segments,
suggesting areas where instructional strategies could be optimized to maintain
student focus.

• Challenges and Limitations:

Despite its successes, the system encountered challenges related to lighting
conditions and facial recognition accuracy in certain environments. Variations
in lighting affected face detection performance, leading to occasional missed
detections or false positives. Additionally, the reliance on facial recognition
technology raised privacy concerns among some stakeholders, highlighting the
importance of implementing robust privacy safeguards and ethical guidelines
in the system’s deployment.

• Future Enhancements and Recommendations:

To address the identified challenges and further enhance the system’s function-
ality, future enhancements could focus on improving face detection algorithms
to handle diverse lighting conditions effectively. Incorporating additional bio-
metric markers, such as gaze tracking for more nuanced activeness assessment,
could provide deeper insights into student engagement patterns. Moreover,
integrating machine learning models for predictive analysis of student atten-
tiveness based on historical data could enable proactive teaching strategies
tailored to individual student needs.

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

IV
Conclusion and Future Scope

10 Conclusion . . . . . . . . . . . . . . . . . . . . . . . 50
10.1 Achievement of Objectives
10.2 Contributions to the Field
10.3 Reflection on Limitations
10.4 Conclusion
10.5 Future Scope
 10. Conclusion

10.1 Achievement of Objectives

The objectives set for the "Face Recognition-Based Smart Student Attendance
and Activeness Monitoring System" have been achieved with remarkable success.
Leveraging Python libraries such as OpenCV and face recognition, the system
effectively automates student attendance tracking and monitors attentiveness during
lectures. Through robust face recognition algorithms, it accurately identifies students,
minimizing manual effort for educators and reducing errors in attendance records.
Additionally, its real-time analysis of webcam footage enables the detection of
closed eyes and yawning, providing valuable insights into student engagement levels.
Overall, the system’s achievements encompass streamlining administrative tasks,
enhancing monitoring capabilities, and fostering a more efficient and engaging
learning environment.

10.2 Contributions to the Field

The "Face Recognition-Based Smart Student Attendance and Activeness Monitoring
System" has significantly advanced the field of education technology by integrating
facial recognition with real-time attentiveness monitoring. Its use of accessible
Python libraries like OpenCV and face recognition demonstrates an innovative
approach to addressing classroom management challenges. By providing insights
into student behavior during lectures, the system enables personalized learning
interventions, marking a notable advancement in educational technology.

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

 10.3 Reflection on Limitations
A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

10.3 Reflection on Limitations

Considering its limitations, the "Face Recognition-Based Smart Student Attendance
and Activeness Monitoring System" faces certain challenges that deserve attention.
Firstly, its accuracy may be affected by factors like lighting conditions or obstructed
faces. Additionally, while it detects eye closure and yawning to gauge attentiveness,
it might miss some instances of distraction. Hardware constraints, like low-quality
webcams or limited processing power, could also hamper its real-time performance.
Moreover, there are concerns regarding privacy and data security due to the collection
and storage of facial images and personal data. Addressing these limitations requires
ongoing improvements in algorithms, better hardware support, and a commitment to
privacy-conscious design. Recognizing and addressing these challenges is essential
for the continued development and responsible use of the system in educational
contexts.

10.4 Conclusion
In this approach, the Face Recognition-Based Smart Student Attendance and Active-
ness Monitoring System presents a cutting-edge solution for enhancing the efficiency
and accuracy of attendance tracking in educational institutions. By leveraging ad-
vanced facial recognition technology, the system streamlines the traditional methods,
reducing time-consuming manual processes and minimizing the scope for errors.
The integration of facial recognition technology allows for swift and non-intrusive
identification of students, fostering a seamless and convenient experience for both
students and faculty.

Furthermore, the system’s activeness monitoring feature provides valuable insights

into student engagement, enabling educators to identify patterns and take proactive
measures to enhance the learning environment. This data-driven approach con-
tributes to a more informed decision-making process and helps educators tailor their
teaching strategies to better meet the needs of individual students. As the education
sector continues to embrace technological advancements, the Face Recognition-
Based Smart Student Attendance and Activeness Monitoring System stands as a
testament to the potential of innovative solutions in optimizing administrative tasks
and improving overall educational outcomes. With its potential to revolutionize
attendance tracking and promote student engagement, this system holds promise for
fostering a more efficient and effective learning environment.

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

 10.5 Future Scope A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

10.5 Future Scope

Moving forward, the project has several avenues for future development and en-
hancement:

(a) Improved Accuracy through Machine Learning Integration:

Integrate machine learning algorithms to enhance facial recognition
accuracy and attentiveness monitoring, ensuring robust performance
across various environmental conditions.

(b) Data Analytics and Visualization Integration:

Explore the incorporation of advanced data analytics and visualization
tools to provide educators with comprehensive insights into long-term
student engagement patterns. This would empower educators to make
data-driven instructional decisions and interventions.

(c) Adaptation for Remote Learning Environments:

Recognizing the increasing prevalence of online education platforms,
adapt the system to cater to remote learning environments. This adap-
tation would extend the system’s utility and accessibility to a broader
range of educational settings, facilitating seamless integration into mod-
ern pedagogical approaches.

(d) Stakeholder Collaboration for Ethical Development:

Collaborate closely with educators, students, and privacy experts to
ensure that future iterations of the system uphold ethical standards and
prioritize user privacy. By engaging stakeholders in the development pro-
cess, the system can evolve to meet the evolving needs and expectations
of educational environments while maintaining trust and transparency.

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

V
Reference

11 Bibliography . . . . . . . . . . . . . . . . . . . . . . 54

Bibliography . . . . . . . . . . . . . . . . . . . . . . 54
Books
Articles
Project Timeline
Plagiarism Report
11. Bibliography

(a) Mansi Singhal, Gufran Ahmad, “Deep Learning Based Real Time Face Recog-
nition For University Attendance System”, IEEE Access, June 2023.

(b) Hyun-Bin Kim, Nakhoon Choi, Hye-Jeong Kwon and Heeyoul Kim, “Surveil-
lance System for Real-Time High-Precision Recognition of Criminal Faces
From Wild Videos”, IEEE Access, volume 11 , pp. 56066-56082 May 2023.

(c) Hao yang1, xiaofeng han, “Face Recognition Attendance System Based on
Real-Time Video Processing”, IEEE Accesss, volume 8, pp. 159143-159150
June 2020.

(d) Shivalila Hangaragi, Tripty Singh, Neelima N, “Face Detection and Recogni-
tion Using Face Mesh and Deep Neural Network”, Procedia Computer Science
218, pp. 741-750 2023.

(e) Adri Priadana, Muhammad Habibi, "Face Detection using Haar Cascades to
Filter Selfie Face Image on Instagram", IEEE Access, 13-15 March 2019

(f) Ainampudi Kumari Sirivarshitha, Kadavakollu Sravani, Kothamasu Santhi

Priya, Vasantha Bhavani, “An approach for Face Detection and Face Recogni-
tion using OpenCV and Face Recognition Libraries in Python”, IEEE Accesss,
2023.

(g) Ashwin Rao, “AttenFace: A Real Time Attendance System Using Face Recog-
nition”, IEEE Access, 2022.

(h) Serign Modou Bah, Fang Ming, “An improved face recognition algorithm and
its application in attendance management system”, Array 5, pp. 100014 2020.

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

 A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

Books
(a) "Computer Vision: Algorithms and Applications"
Author: Richard Szeliski
Description: This comprehensive book by Richard Szeliski provides a thor-
ough introduction to computer vision, covering fundamental principles and
algorithms. It includes topics such as image formation, image stitching, feature
detection, and object recognition – all of which are relevant to understanding
the foundations of face recognition systems.

(b) "Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow"

Author: Aurélien Géron
Description: Authored by Aurélien Géron, this practical guide focuses on
machine learning and deep learning using popular Python libraries like Scikit-
Learn, Keras, and TensorFlow. It provides hands-on examples and exercises
that can be beneficial for implementing machine learning components in your
project.

(c) "OpenCV 4 with Python Blueprints"

Author: Ian Goodfellow, Yoshua Bengio, and Aaron Courville
Description: OpenCV 4 with Python Blueprints" by Gabriel Garrido Calvo
and Joseph Howse offers practical insights into computer vision using the
OpenCV library with a focus on Python. It covers a variety of applications,
including face recognition, making it relevant to your project.

(d) "Deep Learning"

Author: Kai-Uwe Schrogl
Description: Authored by experts in the field, "Deep Learning" provides a
comprehensive introduction to deep learning techniques. Understanding deep
learning is crucial for implementing advanced face recognition models, and
this book serves as a valuable resource.

Articles
i. "FaceNet: A Unified Embedding for Face Recognition and Clustering"

ii. "Attendance Management System using Face Recognition"

iii. "DeepFace: Closing the Gap to Human-Level Performance in Face

Verification"

iv. "A Survey of Face Recognition Techniques"

v. "Towards a Smart Classroom: Monitoring and Analyzing Student Be-

havior"

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

Publication Details

Title: A Face Recognition-Based Smart Student Attendance and Activeness

Monitoring System.
Author(s):
Atharva Ghodekar
Tejas Gunjal
Abhijeet Landage
Abhijit Phapale
Dinesh Thorat
Dr. Rohit Nikam

Institutional Affiliation: Sanjivani College of Engineering, Kopargaon -

423603
Date of Submission: 21/12/2023
Department: Information Technology
Abstract: A Face Recognition-Based Smart Student Attendance and Active-
ness Monitoring System represents a transformative solution to revolutionize
attendance management within educational institutions. This project intro-
duces a new way of tracking student attendance in schools, moving away
from traditional methods like calling names or using cards. We’re using ad-
vanced facial recognition technology, specifically the OpenFace model, to
make attendance more accurate and easier. This system corrects mistakes
intelligently and refines its data for better performance. A distinctive fea-
ture of the system involves the incorporation of the Haar Cascade model to
monitor student engagement. Which does more than just track attendance. It
watches students’ facial expressions and head movements in real-time. This

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

 A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

gives teachers insights into how engaged students are in class, helping them
improve participation. Teachers get immediate feedback, allowing them to
adjust their teaching methods for a more interactive and enjoyable learning
experience. Our system doesn’t just focus on attendance, it takes a broader
approach by monitoring student engagement. By simplifying tasks for school
staff and giving helpful information, this system has the potential to change
how schools operate. It uses cutting-edge technology to make attendance and
engagement monitoring easy, promising to create better classroom experiences
for everyone involved.

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

 A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

 A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

 A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

Appendices

This project requires the installation of several essential libraries and tools to
ensure proper functionality. Python 3.7 or above is the foundational program-
ming language needed. Key libraries include OpenCV for image processing
and computer vision tasks, face recognition for facial recognition capabilities,
and dlib for machine learning applications. Additionally, NumPy is used for
numerical operations, Pandas for data manipulation, and Pillow for image
processing. Users can install these libraries using the pip package manager to
set up their development environment smoothly and avoid common pitfalls
during the installation process. Ensuring that all dependencies are correctly
installed is crucial for the seamless operation of the system.
The dataset is a critical component of the system, consisting of images of
students used for face recognition. It is important that each student’s images
are stored in a designated folder named after their unique identifier or name,
ensuring clarity and organization. High-quality, well-lit images covering
various angles of the student’s face enhance the accuracy of the recognition
system. In addition to dataset preparation, comprehensive usage instructions
guide users through the initial setup, student registration, and daily operations
of capturing attendance and monitoring activeness. The section includes
detailed descriptions of the GUI, supported by screenshots, to assist users in
navigating and utilizing the system effectively. Furthermore, it explains how
to interpret the generated data, enabling educators to gain valuable insights
into student engagement levels. This data can help improve teaching strategies
and ensure students are actively participating in their learning.

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

Project Timeline

The Following are the developers to Perform the tasks:

D1: Atharva Ghodekar
D2: Tejas Gunjal
D3: Abhijeet Landage
D4: Abhijit Phapale
D5: Dinesh Thorat

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

 A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

Project Schedule:

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

Plagiarism Report

Department of Information Technology, Sanjivani College of Engineering, Kopargaon

 A Face Recognition-Based Smart Student Attendance and Activeness Monitoring System

Department of Information Technology, Sanjivani College of Engineering, Kopargaon