A Project Report

Submitted on

Project Title
In Partial Fulfillment of award of the

Bachelor of Technology (B.Tech) Degree

in
Information Technology
by

Tanishq Gupta Janakinath Verma

(2001920130167) (2001920130074)

Abhinav Tyagi Akash Nishad

(2001920130003) (2001920130016)

Under the Supervision of

Ms. Jyoti Raj Singh
Assistant Professor

G.L. BAJAJ INSTITUTE OF TECHNOLOGY & MANAGEMENT

GREATER NOIDA

DR. A.P.J. ABDUL KALAM TECHNICAL

UNIVERSITY, UTTAR PRADESH, LUCKNOW
MAY, 2024
 Declaration

We hereby declare that the project work presented in this report entitled “Food-O-Holic
food delievery System”, in partial fulfillment of the requirement for the award of the degree
of Bachelor of Technology in Information Technology, submitted to A.P.J. Abdul Kalam
Technical University, Lucknow, is based on our own work carried out at the Department of
Information Technology, G.L. Bajaj Institute of Technology & Management, Greater
Noida. The work contained in the report is true and original to the best of our knowledge
and project work reported in this report has not been submitted by us for award of any
other degree or diploma.

Signature:
Name:

Signature:
Name:
Roll No:

Signature:
Name:
Roll No:

Date:
Place: Greater Noida

ii
 Certificate

This is to certify that the Project report entitled “Project Title” done by Student Name (Roll

No), Student Name (Roll No), and Student Name (Roll No) is an original work carried out

by them in Department of Information Technology, G.L. Bajaj Institute of Technology &

Management, Greater Noida under my guidance. The matter embodied in this project work

has not been submitted earlier for the award of any degree or diploma to the best of my

knowledge and belief.

Date:

Supervisor Name HoD Name

Signature of the Supervisor Head of Department

i
 Acknowledgement

The merciful guidance bestowed to us by the almighty made us stick out this project to a
successful end. We humbly pray with sincere heart for his guidance to continue forever.

We pay thanks to our project guide Guide Name who has given guidance and light to us
during this project. His/her versatile knowledge has helped us in the critical times during the
span of this project.

We pay special thanks to our Head of Department HoD name who has been always present
as a support and help us in all possible way during this project.

We also take this opportunity to express our gratitude to all those people who have been
directly and indirectly with us during the completion of the project.

We want to thanks our friends who have always encouraged us during this project.

At the last but not least thanks to all the faculty members of the Department of
Information Technology who provided valuable suggestions during the period of project.

ii
 Abstract

Traffic accidents are a serious problem everywhere, and it is essential to spot them quickly to
reduce the effects. The goal of this project is to create a system that uses machine learning to
identify accidents in real-time using CCTV footage photos and notifies authorities and
emergency services via push notifications. Convolutional neural networks (CNNs), a type of
deep learning model, will be used by the system to evaluate the CCTV footage images and
find any patterns that would indicate an accident. In order to recognize accidents
effectively, the system will be trained using a collection of CCTV footage images of
accidents and non-accident occurrences. The project will also put an emphasis on reducing
false alerts, which can waste time and resources. The security of the system will be
considered. The privacy of the system will be protected by making sure that only pertinent
data is gathered and that it is stored safely. This project will be carried out as a proof-of-
concept system that can be expanded for implementation in the actual world. The findings
of this project will help to speed up the reaction time to accidents and ensuring that
emergency services respond promptly, improving the situation for those who are impacted.

i
 TABLE OF CONTENT

Declaration............................................................................................................................. (ii)
Certificate ....................................................................................................................... (iii)
Acknowledgement ................................................................................................................ (iv)
Abstract ....................................................................................................................... (v)
Table of Content.................................................................................................................... (vi)
List of Figures …………………………………………………………………………….(ix)
List of Tables ……………………………………………………………………………… (x)

Chapter 1. Introduction .......................................................................................... Pg. No.

1.1 Background & Motivation.............................................................................10
1.2 Objective………………………............................................................. 11
1.3 Delimitation of research …………………………………...........................11
1.4 Benefits of research …………………………………………………..........12
1.4.1 Faster Response Time …………………………………………… 12
1.4.2 Accurate Detection ……………………………………………… 12
1.4.3 Resource Efficiency ……………………………………………... 13
1.4.4 Scalability ……………………………………………………….. 13
1.4.5 Privacy …………………………………………………………... 13
1.4.6 Improved Outcomes …………………………………………….. 13
Chapter 2. Literature Survey.......................................................................................14
2.1 Introduction …………………………………………………………… 14
2.2 Literature Review.........................................................................................15

2.3
2.2.1 Some Previous Work …………………………………………….

15 Inferences Drawn from Literature Review………………. 18

vi
Chapter 3. Problem Formulation and Proposed Work………………………. 22
3.1 Introduction …………………………………………………………… 22
3.2 Problem Statement ……………………………………………………. 25
3.3
Proposed Augmentation ………………………………………………. 26
3.4
Data Augmentation ……………………………………………………. 27
3.5
Convolutional Neural Network (CNN) ……………………………….. 28
3.6
PushBullet Notification API …………………………………………... 29
Chapter 4. Methodology…....................................................................................... 31

4.1 Introduction …………………………………………………………… 31

4.2 Implementation Strategy (Flowchart, Algorithm etc.) ………………... 31
4.2.1 Data Collection ………………………………………………….. 31
4.2.2 Data Preprocessing ……………………………………………… 33
4.2.3 Model architecture ………………………………………………. 33
4.2.4 Training …………………………………………………………. 36
4.2.5 Validation ………………………………………………………... 38
4.2.6 Deployment ……………………………………………………... 39
4.3 Tools/Hardware/Software Requirements..……………………………... 41
4.3.1 Hardware Requirement ………………………………………….. 41
4.3.2 Software Requirement …………………………………………... 41

Chapter 5. Result & Discussion ……………………….......................................... 42

5.1
Result……………………………………………………………………. 42
5.2
Epoch Conclusion………………………………………………………. 43
5.3 45
Output……………………………………………………………………
5.4 48
CNN model explained…………………………………………………...
5.5 49
Stats on Training Data………….………………………………………..
5.6 49
Stats on Validation Data…………………………………………………
5.7 49
Scope……………………………………………………………………. 49
vii
Chapter 6.
Chapter 7. Conclusion & Future Scope.………………………………………………. 51
Appendix I References………………………………………………………………. 51
53
Appendix II Plagiarism Report……………………………………………………...
Appendix III Research Paper………………………………………………………... 54
Plagiarism Report for Research Paper…………………………………..
55

i
 LIST OF FIGURES

Figures Description Page

No. Figure 2.1 Computer Vision and Understanding…………………

14
Figure 2.2 Basic Understanding of CNN………………………. 15
Figure 2.3
Convolutional Operation b/w images and filters……. 18
Figure 2.4
Convolutional Op b/w images and filter(Stribe = 1)... 19
Figure 3.1
Proposed Work Architecture………………………… 23
Figure 4.1
Collection of Data…………………………………… 31
Figure 4.2
Data Processing……………………………………… 32
Figure 4.3
Data Augmentation…………………………………. 33
Figure 4.4
CNN model Architecture…………………………… 34
Figure 4.5
Activation function…………………………………. 35
Figure 4.6
Sample of Model working…………………………. 39
Figure 5.1
Epoch output………………………………………. 42
Figure 5.2
CNN layered model……………………………….. 46
Figure 5.3
Stats on Training Data…………………………….. 47
Figure 5.4
Stats on Validation Data…………………………... 47

i
 LIST OF TABLES

Tables Description Page No.

Table 5.1 Output Table……………………………………… 43

ii
 Chapter 1
Introduction

Automobile accidents are a leading source of fatalities and serious injuries worldwide, and they
may have a terrible effect on the victims and their families. The severity of these mishaps can be
reduced by quickly identifying accidents and offering aid. The goal of this project is to create a
system that can use photos from CCTV footage to instantly identify accidents and notify
authorities and emergency services via push notifications.
The technology will use deep learning models from machine learning algorithms to examine
CCTV footage photos and find patterns that could indicate an accident. The system's goal is to
correctly pinpoint incidents while eliminating false warnings, which can waste time and
resources. To make sure that only pertinent data is gathered and data is maintained safely, the
system's privacy will also be a crucial factor.
The importance of this project rests in its ability to speed up emergency service reaction times to
accidents and ensuring that they happen on time, benefiting those who are impacted. The project
will be carried out as a proof-of-concept system that can be expanded for implementation in the
actual world. The only goal of this project is to create a machine learning-based system that can
use CCTV footage to instantly identify accidents and notify authorities and emergency services
via push notifications.

1.1 Background and Motivation

Governments and communities all over the world are very concerned about traffic accidents. The
World Health Organization estimates that 1.35 million people per year pass away in traffic-
related deaths, and millions more sustain injuries and disabilities. Road accidents cause major
economic losses in addition to the human cost, such as medical costs, lost productivity, and
infrastructure damage.
Accident’s effects can be reduced by detecting them quickly and offering assistance when
needed. Accident detection systems have traditionally relied on human observation; however,
this method has its limitations due to the availability of people and their capacity to
simultaneously monitor multiple areas.

Automated accident detection systems have been created thanks to developments in computer
vision and machine learning. In order to identify accidents in real time and notify authorities and
emergency services, these systems can analyze CCTV footage photos. These devices still have
issues with accuracy and false alarms, though.

The goal of this project is to create a machine learning-based system that can quickly and
effectively identify accidents in CCTV footage photos and warn authorities and emergency
services via push notifications. Rapid and precise accident detection by the system can speed up
response times and result in improved results for those who are impacted. Furthermore, by
reducing false alarms, the system can avoid resource waste and guarantee effective use of
emergency services.

1.2 Objective

Develop an automated system that can quickly detect accidents in real-time: The primary
objective of the project could be to develop a system that can quickly and accurately detect
accidents in real- time using CCTV footage. This system should use machine learning
algorithms to analyze the images and identify any accidents that have occurred.

Improve response time in emergency situations: The project could aim to improve response time
in emergency situations by pushing notifications to emergency services, such as hospitals or
police departments, as soon as an accident is detected. This would help emergency services to
arrive at the scene of the accident as quickly as possible.

Reduce the number of fatalities and injuries resulting from accidents: The project could aim to
reduce the number of fatalities and injuries resulting from accidents by providing faster response
times and improving the overall effectiveness of emergency services.
Enhance CCTV camera technology: The project could aim to enhance CCTV camera technology
by developing algorithms that can better analyze and interpret the footage captured by cameras.
This could help to improve the overall quality of CCTV footage and make it easier to detect
accidents.

Enhance public safety: The ultimate objective of the project could be to enhance public safety by
reducing the number of accidents and injuries that occur in public places. By providing a faster
response time and improving the overall effectiveness of emergency services, the project could
help to ensure that people feel safer in public places.

1.3 Limitation of Research

The goal of this project is to create a machine learning-based system that can identify accidents
using CCTV camera photos in real time and notify authorities and emergency services via push
notifications. Convolutional neural networks (CNNs), in particular, will be used in the research to
evaluate the CCTV footage images and find patterns that are suggestive of an accident.

Additionally, false alarms, which can waste time and resources, will be reduced through research.
The privacy of the system will be protected by making sure that only pertinent data is gathered
and that it is stored safely. The research will concentrate on creating a proof-of-concept system
that can be expanded for use in the actual world.

Other elements that affect traffic accidents, such as driver conduct, road conditions, and vehicle
maintenance, won't be included by the study. Additionally, the research will not cover the
deployment and upkeep of the system in real-world circumstances, including the requirements for
hardware and software as well as other logistical issues.

Last but not least, the study will not touch on the ethical and legal ramifications of utilizing
CCTV camera photos to detect accidents, such as data privacy and permission.
1.4 Benefits of Research

1.4.1 Faster Response Time

One of the primary benefits of using a machine learning-based system with push notifications for
accident detection is a faster response time. The system can detect accidents in real-time using
CCTV footage images, and push notifications can be sent immediately to alert authorities and
emergency services. This can lead to a quicker response time, potentially saving lives and
reducing the severity of injuries.

1.4.2 Accurate Detection

Another benefit of using a CNN for accident detection is increased accuracy. The system can be
trained to detect specific patterns and features that are indicative of an accident, reducing the
chances of false alarms.

1.4.3 Resource Efficiency

By minimizing false alarms, the system can prevent the wastage of resources and ensure efficient
use of emergency services. This can lead to cost savings and more efficient use of resources.

1.4.4 Scalability

The system can be scaled up for real-world deployment, potentially covering a large number of
locations and increasing the chances of detecting accidents.

1.4.5 Privacy

The system can be designed to ensure privacy concerns are addressed. Only relevant information
is collected, and data is stored securely.

1.4.6 Improved Outcomes

The faster response time and increased accuracy of accident detection can lead to better
outcomes for the affected parties, including reducing the severity of injuries and potentially
saving lives.

Overall, a machine learning-based system with push notifications for accident detection using
CNN has the potential to improve response times, increase accuracy, and ultimately lead to better
outcomes for those involved in road accidents.
]=

Chapter 2 Literature Review

Chapter 3 Problem Formulation and Proposed Work

Chapter 4 Methodology

Chapter 5 Result and Discussion

Chapter 6 Conclusion and Future Work

Note: Form chapter 1 you have to start page number from 1, 2, 3……..
]=

50
References
[1] Khare, S., Mittal, A., & Goyal, P. (2020). Accident Detection and Notification using Deep
Learning. In 2020 11th International Conference on Computing, Communication and
Networking Technologies (ICCCNT) (pp. 1-5). IEEE.

[2] Roy, N., Bhattacharyya, D., & Kalita, J. K. (2019). Real-time Road Accident Detection using
Deep Learning Techniques. In 2019 10th International Conference on Computing,
Communication and Networking Technologies (ICCCNT) (pp. 1-6). IEEE.

[3] Thanh, P. V., Le, T. P., & Nguyen, T. A. (2021). Real-time Road Accident Detection using
Deep Learning and Cloud Computing. In 2021 13th International Conference on Knowledge and
Systems Engineering (KSE) (pp. 24-29). IEEE.

[4] Singla, P., Singh, P., & Gupta, S. (2020). Real-time Detection of Road Accidents using Deep
Learning Techniques. In 2020 4th International Conference on Intelligent Computing and
Control Systems (ICICCS) (pp. 1-5). IEEE.

[5] Chen, W., Chen, W., & Wang, Y. (2019). Real-time Detection of Road Accidents using Deep
Learning and Internet of Things. In 2019 IEEE International Conference on Smart Internet of
Things (SmartIoT) (pp. 215-220). IEEE.

[6] Zheng, H., Chen, H., & Cui, L. (2020). Deep Learning-based Road Accident Detection
System for Traffic Video Surveillance. In 2020 International Conference on Computer
Information and Big Data Applications (CIBDA) (pp. 166-170). IEEE.

[7] Han, J., Zhang, C., Li, B., & Wang, X. (2019). Road Accident Detection and Notification
System using Deep Learning Techniques. In 2019 IEEE International Conference on

1
Computational Science and Engineering (CSE) and IEEE International Conference on Embedded
and Ubiquitous Computing (EUC) (pp. 295-300). IEEE.
[8] Karki, U., Jung, H., & Park, D. (2018). Real-time Detection of Road Accidents using
Convolutional Neural Networks. In 2018 International Conference on Information and
Communication Technology Convergence (ICTC) (pp. 1405-1408). IEEE.

[9] Mahajan, A., Mukherjee, S., & Sharma, S. (2019). Road Accident Detection and Notification
System using Machine Learning Techniques. In 2019 International Conference on Intelligent
Sustainable Systems (ICISS) (pp. 515-519). IEEE.

[10] Wang, J., Zhang, J., & Liu, Z. (2020). Real-time Detection of Road Accidents using Deep
Learning and Computer Vision Techniques. In 2020 4th International Conference on Big Data
and Computing (ICBDC) (pp. 229-233). IEEE.

[11] Bhattacharjee, D., Baishya, K., & Rahman, T. (2020). Deep Learning-based Road Accident
Detection System for Smart Cities. In 2020 IEEE 17th India Council International Conference
(INDICON) (pp. 1-5). IEEE.

[12] Pal, A. K., Kar, S., & Maity, A. (2020). Real-time Road Accident Detection using
Convolutional Neural Networks and Raspberry Pi. In 2020 6th International Conference on
Advanced Computing and Communication Systems (ICACCS) (pp. 1-6). IEEE.

[13] Poudel, S., Timilsina, S., & Dahal, S. (2018). An Efficient Road Accident Detection System
using Deep Learning Techniques. In 2018 IEEE 4th International Conference on Computational
Intelligence and Applications (ICCIA) (pp. 1-5). IEEE.

[14] Zhang, H., Yang, Y., & Li, X. (2019). Real-time Detection of Road Accidents using Deep
Learning and Edge Computing. In 2019 IEEE International Conference on Edge Computing
(EDGE) (pp. 184-191). IEEE.

2
 Appendix I
Plagiarism Report of Project Report

3
 Appendix II
Research Paper Based on Project

4
 Appendix III
Plagiarism Report of Research Paper

55