Fire-detection

Bangkit Capstone Project
(B21-CAP0484)

This project implemented machine learning to detect fire from HLS video stream and then will send notifications to the android app. We used dataset from kaggle

Table of Contents

1. Requierements
2. Installation
3. Flowchart
4. Documentation
5. Code information
6. Member

Requierements

To run this project, ensure you have:

• Video stream (you can put the video stream to hls-server/streams)
• Installed requirements.txt

Installation

• run HLS server node hls-server/index.js
• run main server python server/main.py
• install the android app
• to get notifications, please click turn on first in the app. You will be notified if fire was detected. (or you can request to localhost/turn_on)
• if you get a notification there is a fire detected, go to Menu → Notification, then you can click get Image to get the frame of the video. (or you can request to localhost/get_status)

Flowchart

Documentation

Code information

+ hls-server
  + streams			# video streams folder
  + cctv.mp4
  + index.js			# code to run hls server
  + file stream.py		# code to generate video stream
  + requirements.txt
+ Model
  + fire_detection_v2.h5
  + Fire_Detection_v2.ipynb
+ server
  + main.py			# code to run main server
  + requirements.txt
  + detector.py
+ Firewatch
  + app				# App Resources
  + asset			# Logo Firewatch
  + gradle/wrapper
  + Firewatch.apk		# Application Firewatch

Member

1. Mobile Programming
   • A3322963 - Ahmad Try Bayu Al Haq
   • A3322967 - M Alfin Nurdiansyah
2. Cloud Computing
   • C0111137 - Barin Khalifamadani Anantapasha
   • C3252917 - Nathanael Victorious
3. Machine Learning
   • M0111144 - Muhamad Syifa Nurhidayat