DESIGN AND IMPLEMENTATION OF OBJECT

DETECTION AND ALARM MANAGEMENT SYSTEM

BY

HAMISU SALISU
02200470475

BSC. COMPUTER SCIENCE

JUNE, 2025

i
 DECLARATION

I hereby declare that this project work titled "Object Detection and Alarm Management
System" is my original work. This work has not been submitted previously for any degree
or qualification at any other academic institution. All information derived from the work of
others has been acknowledged in the text and a list of references is provided.

__________________________
HAMISU SALISU
02200470475

ii
 CERTIFICATION

This is to certify that the project titled "Object Detection and Alarm Management System"
was carried out by HAMISU SALISU with Matriculation Number 02200470475, in partial
fulfilment of the requirements for the award of Bachelor of Science (B.Sc.) in Computer
Science at the Department of Computer Science, Faculty of Natural and Applied Science,
Nassarawa State University, Keffi.

_________________________ ________________
DR. ANYACHEBELU K. T Date
Supervisor

_________________________ ________________
Dr Binyamin A. Ajayi Date
HOD, Computer Science

_________________________ _________________
Internal Examiner Date

_________________________ _________________
Prof M.D Makut Date
Dean, FNAS

_________________________ _________________
External Examiner Date

iii
 DEDICATION

This project is dedicated to my beloved parents, whose unwavering support, sacrifices, and
encouragement have been the foundation of my academic journey; my siblings for their
constant motivation and belief in my abilities; and to all aspiring computer scientists who
strive to make a difference in the world through technology.

iv
 ACKNOWLEDGEMENTS

I would like to express my sincere gratitude to my project supervisor, DR.

ANYACHEBELU K. T, for her guidance, expertise, and invaluable feedback throughout
this project; the entire department of the Computer Science Department at Nasarawa State
University, Keffi, for imparting knowledge and skills that have been instrumental in
completing this project; my classmates and friends, particularly Theresa Eleojo Ibrahim
and Joseph Abraham Dangana, for their support, collaborative spirit, and engaging
discussions that enriched my understanding of the subject; the staff of the school library,
for providing access to necessary resources and technical support; my family, for their
unconditional love, patience, and support throughout my academic journey; and finally, I
thank the Almighty Allah(SAW) for the strength, wisdom, and perseverance granted to me
to complete this project successfully.

v
 ABSTRACT

This study aims to develop a secure and efficient web-based Object Detection and Alarm
management system for law enforcement agencies. The system addresses limitations of
traditional manual and disconnected digital systems by leveraging modern web
technologies and incorporating advanced security measures. Key features include user-
friendly interfaces for both law enforcement personnel and the public, secure
authentication and data encryption, real-time information sharing, and comprehensive data
analytics tools. This project uses a approach of combining system performance metrics
with feedback from law enforcement users. The proposed system aims to enhance
operational efficiency, improve data-driven decision making in crime prevention and
investigation, and provide a model for modernizing crime management processes in law
enforcement agencies. By integrating various aspects of crime management into a single,
secure platform, this study seeks to contribute to the advancement of law enforcement
technologies and improve overall public safety efforts

vi
 TABLE OF CONTENTS

Declaration.............................................................................................................................ii
Certification..........................................................................................................................iii
Dedication.............................................................................................................................iv
Acknowledgements................................................................................................................v
Abstract.................................................................................................................................vi

CHAPTER ONE
INTRODUCTION
1.1 Background to the Study..............................................................................................1
1.2 Statement of the Problem.............................................................................................2
1.3 Objectives of the Study................................................................................................2
1.4 Significance of the Study..............................................................................................3
1.5 Scope of the Study........................................................................................................3

CHAPTER TWO
LITERATURE REVIEW
2.1 Conceptual Framework................................................................................................4
2.2 Empirical Review.........................................................................................................5
2.2.1 The Problems or Issues..........................................................................................5
2.2.2 Solutions Explored Previously...............................................................................5
2.2.3 Gaps in Previous Solutions....................................................................................6

CHAPTER THREE
RESEARCH METHODOLOGY
3.1 Research Design...........................................................................................................8
3.2 System Design..............................................................................................................9
3.2.1 Current System......................................................................................................9
3.2.2 Challenges in the Current System..........................................................................9
3.3 Methods of Data Collection........................................................................................10
3.3.1 Justification of Methods.......................................................................................11
3.2 System Design............................................................................................................12
3.2.1 Current Systems...................................................................................................12
3.2.2 Challenges in the Current System........................................................................12
3.2.3 Proposed System or Solution...............................................................................13
System Specifications:......................................................................................................13
Non-Functional Requirements:........................................................................................13
3.2 System Architecture...................................................................................................14
3.3 Methods of Data Collection........................................................................................15
Goal of the Research........................................................................................................16
Data to be Gathered..........................................................................................................16
Methods and Procedures...................................................................................................16
3.3.1 Justification of Methods.......................................................................................16
Key reasons for choosing Agile include:......................................................................16
System Algorithm Description.........................................................................................17
3.4 Data Architecture....................................................................................................20
REFERENCES.....................................................................................................................25

vii
 LIST OF TABLE
Table 3.1 Users table
Table 3.2 Cameras table
Table 3.3 Detection event tableTable 3.4 Alarms tableTable 3.5 Notification table

viii
 LIST OF FIGURES

Figure 3.1 Agile methodology 11

Figure 3.2 System architecture

Figure 3.3 Flowchart
Figure 3.4 Usecase diagram
Figure 3.5 Entity Relationship diagram

ix
 CHAPTER ONE
INTRODUCTION

1.1 Background to the Study

The rapid evolution of surveillance technologies and computer vision methods has led to

significant improvements in security systems globally. Object detection plays a vital role in

automating the recognition of unauthorized activities, thereby enhancing safety in public

and private spaces (Wang, 2020). University campuses, such as Nasarawa State University,

Keffi (NSUK), face growing security challenges, including unauthorized access, theft, and

vandalism, which call for smarter and more proactive surveillance solutions.

Recent developments in deep learning, particularly convolutional neural networks (CNNs),

have greatly improved the accuracy and speed of object detection systems (Oguine,

Oguine, & Bisallah, 2022). These systems are capable of real-time monitoring and can

trigger alarms automatically upon detecting suspicious activities, thus mitigating risks

associated with manual surveillance (Abbas, Tahir, & Khan, 2021). However, many

institutions in Nigeria and similar developing contexts still rely heavily on manual or semi-

automated monitoring methods, which are often inadequate.

Prior research by Babalola and Olufemi (2024) demonstrated how smart intrusion

detection systems could reduce vandalism in academic environments by integrating IoT

and machine learning technologies. Similarly, Zhang, Chen, and Yang (2023) emphasized

the importance of AI-driven object detection frameworks to provide timely alerts and

improve security response. Implementing such systems at NSUK will contribute to

safeguarding campus facilities and students’ welfare, addressing the limitations of current

practices.

1
1.2 Statement of the Problem

Security on university campuses, especially in student hostels and restricted areas, remains

a challenge due to the inefficiency of existing monitoring methods (Ajayi et al., 2021).

NSUK relies mainly on conventional CCTV systems, which require constant human

supervision, leading to potential oversights and delayed incident reporting (Nwankwo &

Uche, 2022). This manual approach is inadequate for timely detection of threats or

unauthorized activities. Although automated security solutions exist internationally, many

are either too costly or technologically complex for local adaptation. Moreover, existing

systems often lack integration between real-time object detection and alarm management,

limiting their effectiveness in rapidly alerting security personnel (Okoro & Chukwuemeka,

2023). There is therefore a critical need for a tailored, cost-effective system that integrates

object detection with automatic alarms, suited to NSUK’s security context. This study

proposes to design and implement such a system to enhance security monitoring at NSUK

by detecting potential threats automatically and issuing immediate alerts. The approach

aims to reduce reliance on manual supervision, thereby improving the university

community's overall security posture and safety.

1.3 Aim and Objectives of the Study

This study aims to design and implement an object detection and alarm management

system.

The main objectives of this study are to;

i. Design a framework for an Object Detection and Alarm Management System.

ii. Develop an efficient algorithm for the system

iii. Implement the designed algorithm

2
1.4 Significance of the Study

This study is important for several reasons. First, it will enhance the security infrastructure

of Nasarawa State University, Keffi by providing a smart system capable of real-time

detection and alerting, which is crucial for preventing unauthorized access and criminal

activities (Ajayi et al., 2021; Zhang et al., 2023). Second, the reduction in dependence on

human monitoring will minimize errors and increase efficiency, thereby allowing security

personnel to focus on rapid response rather than constant surveillance (Nwankwo & Uche,

2022). Third, the research will contribute to academic knowledge and practical solutions in

Nigeria by developing an affordable and adaptable system for institutions with similar

security challenges (Okoro & Chukwuemeka, 2023). Finally, the proposed system will

serve as a model that can be replicated across other universities and institutions, supporting

broader efforts to leverage AI and machine learning for campus safety.

1.5 Scope of the Study

This research focuses on the design, implementation, and evaluation of an Object

Detection and Alarm Management System specifically for Nasarawa State University,

Keffi. The system will target key locations such as student hostels, administrative offices,

and main entrances—areas identified as most vulnerable to security threats (Ajayi et al.,

2021).

The study involves developing an algorithm capable of detecting unauthorized persons or

suspicious objects in real-time and triggering an alarm for immediate action. The

performance of the system will be tested under realistic campus conditions to evaluate its

accuracy and responsiveness. However, the study does not extend to a full campus-wide

deployment or integration with other university IT infrastructure.

3
 CHAPTER TWO
LITERATURE REVIEW

2.1 Conceptual Framework

The conceptual framework for this study is anchored on the integration of computer vision

and alarm management systems to enhance security through real-time detection and alert

mechanisms. Object detection, a critical subset of computer vision, involves the

identification and localization of objects within images or video frames (Redmon et al.,

2016; Oguine et al., 2022). This technology is employed in surveillance to automatically

identify potential threats or unauthorized activities without human intervention.

Alarm management systems work as complementary tools that trigger notifications or

alarms once suspicious activities are detected (Al-Ali et al., 2021). The interaction between

object detection (independent variable) and alarm activation (dependent variable) forms the

basis for proactive security monitoring. The system’s efficiency relies on its ability to

accurately detect relevant objects and instantly communicate alerts to security personnel

for rapid response.

In the context of Nasarawa State University, Keffi, the framework focuses on identifying

unauthorized persons or suspicious items within restricted zones (hostels, administrative

offices) and immediately notifying the security team. This aligns with the general aim of

leveraging automation to mitigate human limitations in surveillance (Zhang et al., 2023).

Thus, the conceptual framework involves the following components:

4
 i. Input: Real-time video feeds from surveillance cameras.

ii. Process: Object detection algorithm analyzes frames to identify potential threats.

iii. Output: Automatic triggering of alarms and notifications upon detection.

This interaction aims to reduce manual monitoring efforts and improve security efficiency,

as suggested by previous works emphasizing automation in campus security (Abbas et al.,

2021; Babalola & Olufemi, 2024).

2.2 Empirical Review

This section critically examines prior research related to object detection and alarm

systems, particularly their applications in security and surveillance. It also highlights

challenges faced and gaps that this study intends to address.

2.2.1 The Problems or Issues

A major issue in security systems across academic institutions in Nigeria, including

Nasarawa State University, is the reliance on traditional surveillance systems that require

constant human supervision (Ajayi et al., 2021). This approach is error-prone and

inefficient, as manual monitoring can lead to missed security breaches and delayed

responses (Nwankwo & Uche, 2022).

Furthermore, conventional CCTV cameras do not provide real-time analytics or automated

alerts, leaving security personnel overwhelmed and less effective during critical situations

(Okoro & Chukwuemeka, 2023). The absence of intelligent detection results in

vulnerability to theft, vandalism, and unauthorized access (Babalola & Olufemi, 2024).

In addition, while some institutions have started integrating computer vision into their

surveillance, these systems are often costly, complex, and not tailored to the local context,

limiting their deployment (Oguine et al., 2022). This gap has prevented the widespread

adoption of effective security automation in Nigerian universities.

5
2.2.2 Solutions Explored Previously

Various studies have attempted to address these challenges by designing automated

surveillance systems using deep learning and IoT technologies. Abbas et al. (2021)

proposed a real-time campus security system utilizing convolutional neural networks

(CNNs) for accurate object detection combined with an alert management module. Their

system demonstrated improved detection accuracy and reduced human supervision needs.

Babalola and Olufemi (2024) developed an IoT-based intrusion detection system for

polytechnic campuses that combined sensors with object recognition algorithms to trigger

alarms upon detecting suspicious movements. Their solution highlighted the benefit of

integrating multi-sensor data for increased reliability.

Additionally, Zhang et al. (2023) focused on AI-based frameworks that provided scalable

and cost-effective solutions adaptable to different campus environments. Their work

underscored the importance of context-specific customization to meet unique institutional

security requirements.

Despite these advances, most solutions have been limited by hardware costs,

computational complexity, and lack of integration with existing campus security protocols

(Okoro & Chukwuemeka, 2023).

2.2.3 Gaps in Previous Solutions

Although significant progress has been made, several gaps remain. Firstly, many existing

systems are designed for developed countries with access to advanced infrastructure and

budgets, which makes them unsuitable for institutions like Nasarawa State University

(Ajayi et al., 2021).

Secondly, integration between object detection algorithms and alarm management remains

limited or rudimentary in many studies, reducing the overall effectiveness of the security

system (Nwankwo & Uche, 2022).

6
Thirdly, the majority of prior work focuses on either object detection or alarm systems

independently, neglecting the synergy that arises from combining both in a unified system

(Okoro & Chukwuemeka, 2023). Lastly, limited empirical evaluation of these systems in

real campus environments creates uncertainty about their practical utility in Nigerian

contexts, which tend to have unique security challenges and resource constraints (Babalola

& Olufemi, 2024).

This study aims to fill these gaps by designing a tailored, integrated object detection and

alarm management system that is both cost-effective and contextually suitable for NSUK.

The system will be evaluated under realistic conditions to provide evidence of its utility in

enhancing campus security.

7
 CHAPTER THREE
RESEARCH METHODOLOGY

3.1 Research Design

This section outlines the methodological framework employed to develop a practical

solution addressing the security challenges faced by Nasarawa State University, Keffi

(NSUK). The research adopts a design and implementation-based approach aimed at

creating an Object Detection and Alarm Management System to automate surveillance and

improve incident response on campus.

i. Measurement Objective

The primary goal of this study is to design, develop, and implement a robust

system capable of detecting unauthorized objects or suspicious activities within

the university environment and generating real-time alerts to security personnel.

The system’s success will be measured based on detection accuracy, response

time reduction, user interaction efficiency, and system reliability in real-world

conditions.

ii. Data Collection Process

Data for this research was gathered through a combination of methods:

i. Observation of existing security measures and incident management

practices at NSUK, particularly focusing on manual patrol routines and

response protocols.

ii. Informal interviews and discussions with campus security officers and

university administrative staff to understand security gaps and operational

challenges.

iii. Review of incident reports and security logs where available to identify

common threat scenarios and response delays.

8
These data sources provided comprehensive insight into the operational context and

informed system requirements and design considerations.

iii. Data Analysis and Reporting Plan

Collected qualitative data were analyzed thematically to identify key

vulnerabilities, operational inefficiencies, and user needs. These findings

guided the system architecture, including sensor placement, algorithm design,

and user interface development.

The reporting plan encompasses detailed documentation of the system design

process, the algorithms developed for object detection, the integration of the

alarm management module, and results from prototype testing on campus

scenarios.

3.2 System Design

System design involves outlining the architecture, components, and operational workflow

of the Object Detection and Alarm Management System to meet the specified requirements

for NSUK’s campus security enhancement.

3.2.1 Current System

The current security framework at Nasarawa State University is fully manual, relying

solely on security personnel who patrol the campus and monitor entry points without

electronic surveillance tools such as CCTV cameras (Ajayi et al., 2021). This approach is

limited by human factors, coverage gaps, and delayed incident detection.

3.2.2 Challenges in the Current System

The manual system faces multiple challenges:

i. Inadequate coverage due to limited manpower, leaving some campus areas

unmonitored (Nwankwo & Uche, 2022).

9
 ii. Delayed detection and response to incidents owing to reliance on human

observation and reporting (Okoro & Chukwuemeka, 2023).

iii. Susceptibility to human error and fatigue, affecting vigilance levels (Babalola &

Olufemi, 2024).

iv. Absence of real-time alert mechanisms to notify security staff immediately of

suspicious events.

3.2.3 Proposed System or Solution

The study proposes an integrated Object Detection and Alarm Management System

designed to automate surveillance by:

i. Deploying cameras across critical campus locations for continuous monitoring.

ii. Implementing computer vision algorithms to detect unauthorized persons,

unattended objects, or unusual activities.

iii. Triggering automated alarms and mobile notifications for rapid security response.

iv. Providing a centralized dashboard for security personnel to monitor live feeds and

manage alerts efficiently.

This design bridges the existing security gaps by combining automation with human

oversight, improving campus safety and operational efficiency.

3.3 Methods of Data Collection

Data collection focused on acquiring relevant information to understand the current

security landscape and specify system requirements.

i. Observation: Researchers conducted on-site observations of security patrol routines

and campus layout.

ii. Interviews: Informal discussions with security personnel and campus administrators

captured experiential insights and identified security pain points.

10
 iii. Document Review: Examination of incident logs and security protocols to

contextualize system needs and validate identified challenges.

These methods ensured a comprehensive understanding of the operational context and

informed a user-centric system design.

3.3.1 Justification of Methods

The combination of qualitative data collection methods was chosen to ensure both practical

and contextual accuracy in system development. Observations and interviews provide real-

world insights that cannot be captured through theoretical research alone, while document

reviews offer a factual basis for identifying system requirements and validating security

challenges. This triangulation enhances the robustness and relevance of the final design

and implementation.

Figure 3.1 Agile methodology

Retrieved from: Beck, K., et al. (2001). Manifesto for Agile Software development

11
3.2 System Design

System design refers to the process of defining the architecture, modules, interfaces, and

data flow for the Object Detection and Alarm Management System tailored for Nasarawa

State University, Keffi. This system aims to enhance campus security by automating object

detection and triggering alarms to alert security personnel in real-time. The design process

identifies stakeholders, functional requirements, and interaction pathways to ensure an

efficient, user-friendly, and reliable security solution.

3.2.1 Current Systems

Currently, campus security at Nasarawa State University primarily relies on manual

surveillance conducted by security personnel stationed across different locations. There is

no automated system for continuous monitoring or rapid detection of suspicious objects or

activities. This human-dependent security setup lacks technological support such as CCTV

cameras integrated with alarm systems, which leads to delays in detecting and responding

to security threats.

3.2.2 Challenges in the Current System

The current manual security framework faces several critical challenges;

i. Absence of automated surveillance tools, resulting in limited monitoring coverage

and delayed threat detection

ii. Over-reliance on human vigilance, which can be inconsistent due to fatigue or

distraction.

iii. Lack of integration between detection and alarm notification systems, reducing the

speed of response.

iv. Difficulty in maintaining records of security incidents due to absence of systematic

data logging.

v. Inability to analyze patterns of security threats for proactive prevention.

12
 vi. These challenges underscore the urgent need for an automated object detection and

alarm system to strengthen campus security and improve responsiveness.

3.2.3 Proposed System or Solution

The proposed solution is a web-based Object Detection and Alarm Management System

designed to provide continuous, real-time monitoring of key university locations. The

system will automatically detect suspicious objects or movements using advanced

computer vision techniques and immediately notify security personnel through alarms and

alerts.

System Specifications:

i. Users: Security Administrators, Security Personnel, System Supervisors

ii. Modules: Real-time Object Detection, Alarm Notification, Incident Logging, User

Management, Report Generation

Functional Requirements:

i. Continuous video feed processing for object detection

ii. Real-time alarm triggering on detection of predefined threats

iii. Secure login and role-based user authentication.

iv. Incident reporting and historical log access.

v. Notification system (SMS, email, or app alert) to alert security teams immediately

Non-Functional Requirements:

i. High system availability and fault tolerance to ensure continuous monitoring.

ii. Scalability to cover multiple locations or expand to other university campuses.

iii. User-friendly and responsive interface for easy interaction by security personnel.

iv. Data security and privacy protection for video feeds and user data.

v. Compatibility with existing network infrastructure at the university

13
 Figure 3.2 System architecture

3.2 System Architecture

The system adopts a three-tier architecture designed for robustness, scalability, and real-

time responsiveness:

i. Presentation Layer:

 User Interface for monitoring detected objects and alarm statuses.

 Technologies: HTML, CSS, JavaScript (React.js or Vue.js) for responsive

dashboards and alerts display.

ii. Application Layer:

 Core business logic including object detection processing, alarm triggering, user

authentication, and real-time notification handling.

14
  Technologies: Python (Flask/Django REST API) integrated with TensorFlow or

OpenCV for object detection; WebSocket for live alarm updates.

iii. Data Layer:

 Persistent storage of detected events, user profiles, alarm logs, and system

configurations.

 Technologies: PostgreSQL or MongoDB, chosen for their reliability and flexibility

in handling time-series and event data.

This layered approach ensures separation of concerns, allowing efficient maintenance and

future expansion such as mobile app integration or advanced analytics.

3.3 Methods of Data Collection

To develop a system that meets actual security monitoring needs in Nsuk, the following

data collection methods were employed:

i. Observational Study: Monitoring current security practices in Nsuk, noting how

security personnel handle object detection and alarms manually without automated

systems.

ii. Semi-Structured Interviews: Engaging local security personnel and facility

managers to understand their challenges, expectations, and operational

requirements for an automated detection and alarm system.

iii. Document Review: Reviewing security logs, incident reports, and existing alarm

system protocols to identify gaps and key functional requirements.

iv. Comparative Study: Analyzing existing object detection and alarm systems to

benchmark features such as detection accuracy, notification speed, and user

interface design.

15
Goal of the Research

To design and implement an effective Object Detection and Alarm Management System

that automates surveillance, improves real-time threat detection, and enhances the security

response capabilities in Nsuk.

Data to be Gathered

a. Current security procedures and workflows

b. Specific object detection and alarm notification needs

c. System functional and technical requirements

d. Benchmark features and best practices from existing security systems

Methods and Procedures

a. Field observations of security operations in Nsuk

b. Recording and analyzing interview responses

c. System modeling using UML diagrams (use case, activity, sequence diagrams)

d. Iterative system design and implementation using Agile methodology to incorporate

continuous feedback

3.3.1 Justification of Methods

The Agile methodology was selected for the development of the Object Detection and

Alarm Management System due to its flexibility, iterative nature, and emphasis on

continuous user involvement, which align well with the project’s dynamic requirements

and real-world application.

Key reasons for choosing Agile include:

i. Flexibility to adapt to changing requirements: Security systems often require

adjustments during development based on real-time feedback from end-users

16
 (security personnel) and evolving threat scenarios. Agile’s iterative sprints allow

for incremental improvements and quick incorporation of new insights.

ii. Frequent user feedback and collaboration: Agile encourages regular interaction

with stakeholders, which ensures the system addresses actual needs, usability

concerns, and operational challenges faced by Nsuk security personnel, leading to a

more user-centered design.

iii. Early and continuous delivery of functional components: The system can deliver

usable features early (e.g., basic object detection, alarm triggers) and progressively

add advanced functionality (e.g., alarm management, reporting), enabling timely

evaluation and refinement.

iv. Risk reduction through incremental development: By breaking the project into

manageable sprints, potential technical risks (like real-time video processing

challenges) are identified and resolved early, minimizing costly failures at later

stages.

v. Enhanced quality through iterative testing: Continuous testing during each sprint

ensures that the system’s core functionalities, such as object detection accuracy and

alarm responsiveness, meet quality standards before final deployment.

This methodology suits the complexity and security-critical nature of the project, ensuring

a practical, reliable, and user-friendly solution tailored to the operational environment of

Nsuk.

System Algorithm Description

The following algorithm outlines the core logical flow of user interaction with the Object

Detection and Alarm Management System:

i. Start: User (security personnel) launches the monitoring platform.

ii. Login: Secure user authentication verifies credentials and assigns roles.

17
 iii. Dashboard Access: User lands on the main interface displaying live camera feeds

and alarm statuses.

iv. Monitoring:

i. The system continuously analyzes video streams for predefined objects (e.g.,

intruders, suspicious packages).

ii. Upon detection, an alarm is triggered immediately.

v. Alarm Handling:

i. System sends real-time notifications (visual and audio alerts).

ii. User can acknowledge, silence, or escalate the alarm.

vi. Event Logging: All detection events, alarms, and user actions are recorded in the

database for audit and analysis.

vii. Reporting: Users can generate reports summarizing alarm history, detection

accuracy, and response times.

viii. Logout: User securely exits the system.

This algorithm forms the basis for your UML diagrams and system modeling in the next

sections, representing the key interactions and workflows within your project.

18
Figure 3.3 flowchart

19
Figure 3.4 Usecase diagram

3.4 Data Architecture

The database schema for the Object Detection and Alarm Management System defines the

structure and organization of data crucial for managing real-time surveillance, alarm

events, user roles, and security logs within Nasarawa State University, Keffi. The schema

outlines the key tables, relationships, and data types necessary to store information such as

detected objects, alarm triggers, user credentials, notification history, and system

configurations.

This structured data organization ensures efficient retrieval and integrity of information,

enabling seamless tracking of security events and management of user interactions. Proper

relational mappings between entities, such as users and alarms or cameras and detection

events, allow for comprehensive reporting and audit trails. Ultimately, the data architecture

supports the system’s goals of providing accurate detection, timely alarm notifications, and

effective monitoring to enhance campus security.

20
 Field Name Data Type Description
user_id INT (PK) Unique identifier for each user
username VARCHAR(50) User login name
password_hash VARCHAR(255) Hashed password
full_name VARCHAR(100) Full name of the user
role ENUM('admin', 'security', 'operator') User role in the system
email VARCHAR(100) User email address
phone_number VARCHAR(15) Contact phone number
created_at DATETIME Account creation timestamp

Table 3.1 Users table

Field Name Data Type Description

camera_id INT (PK) Unique camera identifier
location VARCHAR(100) Physical location of the camera
ip_address VARCHAR(15) IP address of the camera
status ENUM('active', 'inactive', 'maintenance') Operational status
installation_date DATETIME Date camera was installed

Table 3.2 Cameras table

Field Name Data Type Description

event_id INT (PK) Unique identifier for detection event

21
 Field Name Data Type Description
camera_id INT (FK) Foreign key referencing Cameras
Type of object detected (e.g., person,
detected_object VARCHAR(50)
vehicle)
confidence_score FLOAT Confidence score of detection (0 to 1)
event_timestamp DATETIME When the detection occurred
BLOB or
image_snapshot Path or binary data of captured image
VARCHAR(255)

Table 3.3 Detection event table

Field Name Data Type Description

alarm_id INT (PK) Unique alarm identifier
Detection event that triggered the
event_id INT (FK)
alarm
Type of alarm (e.g., intrusion,
alarm_type VARCHAR(50)
unauthorized access)
ENUM('active', 'acknowledged',
alarm_status Current alarm state
'resolved')
triggered_at DATETIME Timestamp when alarm was triggered
User ID of person who acknowledged
acknowledged_by INT (FK)
the alarm
Timestamp when alarm was
acknowledged_at DATETIME
acknowledged
resolved_at DATETIME Timestamp when alarm was resolved

Table 3.4 Alarms table

Field Name Data Type Description

notification_id INT (PK) Unique notification identifier
alarm_id INT (FK) Related alarm triggering notification

22
 Field Name Data Type Description
user_id INT (FK) Recipient of the notification
notification_type VARCHAR(50) Email, SMS, in-app alert
message TEXT Notification content
sent_at DATETIME Time notification was sent
status ENUM('sent', 'failed', 'read') Notification delivery status

Table 3.5 Notification table

23
Figure 3.6 Entity relation diagram

24
 REFERENCES

Ahmed, M., Paul, A., & Ahmed, S. (2021). Real-time object detection for smart
surveillance using deep learning. Procedia Computer Science, 199, 184-191

Alghamdi, A., & Khan, R. A. (2023). Deep learning-based object recognition systems
for smart surveillance: A survey. IEEE Access, 11, 31584–31603.

Ali, H., & Mohamed, M. A. (2020). IoT-based intelligent alert system for intrusion
detection using computer vision. International Journal of Advanced Computer
Science and Applications, 11(3), 248–253.

Anwar, M., & Siddique, M. U. (2020). Review of object detection systems for
surveillance. International Journal of Computer Applications, 177(20), 1–6.

Bhattacharya, S., & Roy, A. (2022). YOLO-based object detection in embedded

surveillance systems. Journal of Ambient Intelligence and Humanized
Computing, 13, 7435–7445.

Chen, J., Lin, G., & Zhang, Z. (2020). Real-time fire and human detection using deep
convolutional neural networks. Fire Technology, 56, 911–930.

Dey, R., & Choudhury, R. (2022). Intelligent object classification for smart city
surveillance using AI. ICT Express, 8(1), 95–100.

Hassan, R., & Hossain, M. S. (2021). Internet of Things-based smart alert system using
object recognition. Journal of Network and Computer Applications, 178,
102989.

He, K., Gkioxari, G., Dollar, P., & Girshick, R. (2020). Mask R-CNN. IEEE
Transactions on Pattern Analysis and Machine Intelligence, 42(2), 386–397.

Khan, Z., & Zubair, M. (2023). Edge AI for real-time object detection in urban
monitoring. Sensors, 23(1), 55.

Liu, W., Anguelov, D., Erhan, D., Szegedy, C., & Reed, S. (2021). SSD: Single shot
multibox detector. IEEE Transactions on Pattern Analysis and Machine
Intelligence, 43(4), 957–971.

Mahmood, T., & Naz, S. (2021). Deep learning techniques for visual surveillance
applications: A review. Artificial Intelligence Review, 54, 1211–1252.

Mohamed, I. Y., & Abdi, F. A. (2020). Design and implementation of a smart motion-
detection alarm system using Raspberry Pi. International Journal of Computer
Science and Network Security, 20(9), 126–134.

Nair, P. S., & Raj, R. (2020). Object detection using YOLO for smart surveillance
applications. International Journal of Recent Technology and Engineering, 8(6),
1743–1747.
25
Nguyen, H. T., & Le, M. H. (2022). AI-powered security monitoring with CNN and
IoT integration. Journal of Intelligent & Robotic Systems, 106(3), 62.

Park, C., & Kim, Y. (2021). Implementation of a real-time video surveillance system
with object detection. Multimedia Tools and Applications, 80(18), 28465–28484.

Patil, R., & Bhandari, S. (2020). Motion-triggered object detection system using
OpenCV and Python. International Research Journal of Engineering and
Technology, 7(7), 2162–2165.

Saleh, A. H., & Alzubaidi, A. (2023). Hybrid deep learning models for real-time object
tracking in surveillance systems. Computers, 12(3), 64.

Sharma, A., & Singh, R. (2021). Human activity recognition for smart surveillance
using ML. International Journal of Intelligent Systems and Applications, 13(5),
1–10.

Wang, J., & Yu, W. (2020). Object detection algorithms for smart security
environments: A comprehensive survey. Information Fusion, 54, 261–280.

26