Project Title: Stove and Fire Safety Monitoring System

Objective: Enhance home fire safety by detecting the status of a stove (on/off) and the presence of a fire simultaneously. Ensure that the stove is actively used for cooking when turned on, and promptly alert users in case of potential dangers.

Key Features:

1. Stove Status Detection: Utilizes YOLO-8 algorithm for real-time object detection to recognize the state of the stove (on/off). This ensures that the system can accurately identify whether the stove is in use or not.

2. Fire Detection: Monitors the area around the stove for the presence of fire using image analysis and computer vision techniques. The system must promptly identify any flames or fire sources.

3. Safety Mode: When the stove is on, and a fire is actively detected, the system indicates "Safe Mode," ensuring that the stove is being used for cooking purposes safely.

4. Danger Mode: If the stove is on, but no fire is detected for a certain continuous period (e.g., 120 frames), the system switches to "Danger Mode" to alert users about the potential risk. It calculates a percentage to reflect the level of risk based on the duration without fire detection.

5. Threshold Alert: Provides a customizable threshold for danger mode detection, allowing users to define when they consider the stove as potentially unsafe based on their preferences.

6. Real-time Monitoring: The system operates in real-time, continuously analyzing the stove area to provide immediate feedback.

Benefits:

• Increases home fire safety by reducing the risk of unattended stoves causing accidents.
• Empowers users with an automated safety monitoring system that provides real-time alerts.
• Customizable danger mode threshold enables users to adjust sensitivity based on their preferences.

Tech Stack

Language: Python

Libraries to Use: CVZone,Cv2,Ultralytics

UI: StreamLit

Model to Use: Finetuned Yolo-V8 Small Model with Custom Dataset

Run Locally

Clone the project

git clone https://github.com/Sathishmahi/Fire-Detection.git

create ,activate conda env and install requirements

 bash init_setup.sh 

run streamlit app

bash run.sh

Demo Video

first_clip.1.mp4