Skip to content

A Computer Vision based system to detect fire at an early stage in real-time and alert the user through a mobile application.

License

Notifications You must be signed in to change notification settings

ninadnaik10/FireSense

Repository files navigation

FireSense

FireSense is a Computer Vision based system to detect fire at an early stage in real-time and alert the user through a mobile application. It is built by training YOLO model on a dataset consisting 2400 images with and without fire. The system consists of Python program running YOLO model, a Flask server and an Android application to receive instant notification and alert the user by playing siren sound.

Features

  1. Real-time Early Fire Detection
  2. Instant Notification
  3. Snapshot of detected fire

Tech Stack used

  1. Python
  2. YOLOv8
  3. Flask
  4. Flutter
  5. Firebase Cloud Messaging

Flowchart

FireSense flow chart

Screenshot

Notification alert on Android device Screenshot demo

Further Development

  1. Integrate with the existing CCTV software solution.