Fire Detection by Image Analysis ECE Project Report

Introduction to Fire Detection by Image Analysis Project:

The main aim of this project is to detect the fire by analyzing images. The process is carried out by comparing y, cb, and cr of YcbcR color model. The sensors are not used in this project. For creating Y, CB, Cr components from a RGB image. The color space transformation equation is used for transformation of each RGB pixels in to the Y, CB, and Cr image. Y indicates Luma that is black or white achromatic portion of the image and CB, Cr are red, blue difference between the Chroma components.

These colors are chosen initially why because illumination can be separated easily from the image than other colors. Based on these conditions the alarm will rang. 

 If the fire pixel has y(x,y)>=Cr(x,y)>=Cb(x,y).

  Whenever a non- pixel does not satisfy this condition. That is a fire pixel.

 If certain number of pixels like 10 to 50 satisfy above condition the alarm will rang

Design & Structure:

This fire detection alarm consists of a camera with 12 megapixels for high quality resolution and picture analysis. This is connected to the computer which receives the video bit sequence with 640×480 resolutions and then it analyzed based on the above conditions specified.

The model supposed to detect 70% of fire and 10% of false alarm.


  It can be used railway stations

 It can be used in nuclear power plants

  It can be used in bus-station as well as airports

  It can be used in the fireworks manufacturing company

  It can be used in forest which are most frequently lit out

  Where we can find a place for frequent fire catchable it can be used their


By this model we can only some amount of fire but not accurately.

Download  Fire Detection by Image Analysis ECE Project Report .

Leave a Reply

Your email address will not be published. Required fields are marked *