IMAGE PROCESSING BASED FIRE DETECTION
Project Report
ET3491 : Embedded Systems and IoT Design
( Project Driven Course)
Submitted by
Aditya P (311522106005)
Kishore M (311522106052)
Manoj G (311522106061)
DEPARTMENT OF ELECTRONICS AND COMMUNICATION
ENGINEERING
MEENAKSHI SUNDARARAJAN ENGINEERING COLLEGE,
(An Autonomous Institution)
KODAMBAKKAM, CHENNAI-24
ANNA UNIVERSITY: CHENNAI 600 025
MAY 2025
MEENAKSHI SUNDARARAJAN ENGINEERING COLLEGE
(An Autonomous Institution)
ANNA UNIVERSITY: CHENNAI 600 025
BONAFIDE CERTIFICATE
Certified that this project report
“_______________________________________________________________
__________________________________” is the bonafide work of “Aditya P
(311522106005), Kishore M (311522106052), Manoj G (311522106061)” who
carried out the project work under my supervision.
SIGNATURE SIGNATURE
HEAD OF THE DEPARTMENT COURSE INSTRUCTOR
Electronics and Communication Engineering Electronics and Communication Engineering
Meenakshi Sundararajan Engineering Meenakshi Sundararajan Engineering
College, College,
No.363, Arcot Road, Kodambakkam, No.363, Arcot Road, Kodambakkam,
Chennai -600 024. Chennai -600 024.
Submitted for the end semester project review of Embedded System and IoT
Design in the Department of ________________________ held on .
INTERNAL EXAMINER EXTERNAL EXAMINER
ACKNOWLEDGEMENT
We, the project team, would like to express our sincere gratitude to all those
who supported and contributed to the successful completion of our project titled
"Image Processing Based Fire Detection".
First and foremost, we are deeply thankful to our project guide, Arif, for their
valuable mentorship, constant encouragement, and insightful suggestions
throughout the course of the project. Their expertise played a vital role in
guiding us through both the conceptual and practical aspects of the system.
We also express our appreciation to the faculty and staff of the Department of
Electronics and Communication Engineering, whose teaching and resources
greatly supported our research and development process.
Our heartfelt thanks go to the developers and documentation contributors of the
tools and services used in our project—ESP32-CAM, DeepAI, Twilio, n8n,
and Google Drive API—whose platforms enabled seamless integration of
components for real-time operation.
We are also grateful to our fellow classmates, peers, and friends for their helpful
feedback, technical support, and encouragement during testing and debugging
phases.
Lastly, we would like to thank our families for their patience, motivation, and
continuous support throughout the project duration. Their belief in our
capabilities gave us the strength to work with full dedication.
This project has been an incredible learning journey for all of us and has
significantly enriched our understanding of embedded systems, IoT, and
artificial intelligence in real-world applications.
ABSTRACT
With the rapid evolution of smart technologies and the Internet of Things (IoT),
safety systems have seen substantial improvement in terms of efficiency and
responsiveness. One of the critical safety challenges faced today is the early and
reliable detection of fire, particularly in open areas or large indoor spaces where
traditional smoke sensors are often ineffective. This project presents a collaborative
effort to design a smart, image-based fire detection system that utilizes embedded
systems and cloud-based automation for real-time detection and alerting.
Our system is built around the ESP32-CAM, a compact microcontroller with an
integrated camera, which captures environmental images either periodically or based
on motion. These images are automatically uploaded to a dedicated Google Drive
folder using n8n, an open-source automation platform. The uploaded images are then
analyzed using an AI model provided by the DeepAI Fire Detection API. Upon
detecting fire, the system immediately sends SMS alerts via Twilio and email
notifications to designated users, ensuring prompt awareness and action.
This project integrates key technologies such as edge computing, AI-based image
recognition, and cloud workflow automation to deliver a practical, low-cost fire
detection solution. The system was tested in both indoor and outdoor environments
and achieved reliable fire detection with minimal false positives. The results
demonstrate that our system can serve as an efficient tool for fire surveillance in
homes, offices, factories, and public places. The modularity of the design also allows
for future enhancements such as mobile app control, offline processing, and thermal
imaging.
`
TABLE OF CONTENTS
S. NO. TITLE PAGE NO.
1
INTRODUCTION
1.1 GENERAL
1.2 PROBLEM STATEMENT
1.3 OBJECTIVES
2 LITERATURE REVIEW
3 THEORETICAL BACKGROUND
4 CONCEPT & METHODOLOGY
DESIGN, MODELLING &SYSTEM
5 ARCHITECTURE
6 RESULTS AND DISCUSSION
7 CONCLUSION AND SCOPE OF FUTURE WORK
8 REFERENCES
Image Processing Based Fire Detection
1. Introduction
1.1 General
The rapid advancements in the Internet of Things (IoT) have revolutionized
safety and automation systems. Among these, early fire detection is critical to
preventing catastrophic losses. Traditional fire detection systems often rely on
smoke or heat sensors, which are limited in open or large spaces. This project
proposes a real-time image-based fire detection system using the ESP32-CAM,
integrated with AI and cloud automation for intelligent monitoring and alerting.
1.2 Problem Statement
Conventional fire detection systems may fail in large or open environments
where smoke disperses quickly. Additionally, they lack smart image analysis
and remote notification capabilities, which are essential for modern safety
systems.
1.3 Objectives
● Develop a cost-effective, smart fire detection system using ESP32-CAM.
● Capture and upload images to Google Drive using n8n.
● Integrate AI-based fire detection via DeepAI or custom models.
● Enable real-time alerts through SMS (Twilio) and email.
2. Literature Review
Various fire detection techniques such as smoke detectors, IR sensors, and
thermal cameras have been employed historically. While effective, these
methods are either costly or limited in scope. Recent studies show that AI-based
image processing using CNNs or APIs significantly enhances detection
accuracy. Projects integrating ESP32-CAM with cloud-based tools like n8n
have proven effective for remote and automated monitoring.
3. Theoretical Background
● ESP32-CAM: A compact microcontroller with built-in camera, ideal for
IoT and image processing tasks.
● n8n: An open-source workflow automation tool used for managing image
uploads, integrating APIs, and triggering alerts.
● AI-Based Fire Detection: Utilizes services like DeepAI for analyzing
images for fire patterns using pre-trained deep learning models.
● Communication Tools: Twilio API for SMS and SMTP for sending
email alerts.
4. Concept & Methodology
The project involves the following steps:
1. ESP32-CAM captures images either periodically or upon motion
detection.
2. Images are uploaded to a specified Google Drive folder using n8n.
3. n8n triggers a workflow:
○ Fetches the uploaded image.
○ Sends it to an AI service (e.g., DeepAI).
○ If fire is detected, an alert is sent via SMS and email.
This system ensures autonomous fire detection and instant alerting without
human involvement.
5. Design, Modelling & System Architecture
Hardware:
● ESP32-CAM module.
● IR sensor for motion-triggered image capture.
Workflow Diagram:
[ESP32-CAM] → [Google Drive Upload (via n8n)] → [AI Fire Detection] → [Twilio/Gmail Alert]
1. ESP32-CAM captures an image.
2. n8n detects the new file.
3. The image is analyzed via AI.
4. Alerts are triggered if fire is detected.
6. Results and Discussion
Extensive testing was conducted in diverse indoor and outdoor conditions. The
AI model showed a high detection rate with minimal false positives. The
integration with Twilio and Gmail worked reliably for sending instant alerts.
Overall, the system effectively met its goal of early fire detection with real-time
alerts.
7. Conclusion and Scope of Future Work
Conclusion
This project successfully integrates embedded systems, image processing, and
cloud automation to deliver an effective fire detection system. It ensures safety
through real-time monitoring and alerting.
Future Scope
● Enhance accuracy using thermal cameras.
● Develop custom AI models for offline (on-device) processing.
● Add mobile app support for live feed and system control.
8. References
1. DeepAI Fire Detection API Documentation
2. ESP32-CAM Datasheet
3. Twilio API Documentation
4. Google Drive API Documentation.
