Thursday, 28 May, 2020




Deep Learning Based Fire Detection System

International Journal of Knowledge Based Computer Systems

Volume 6 Issue 1

Published: 2018
Author(s) Name: Mathew Regi, Renju George Varghese, Sidharth V. and Jency Thomas | Author(s) Affiliation: Dept. of Computer Science, Muthoot Institute of Technology and Science, Ernakulam, Kerala, India.
Locked Subscribed Available for All


The fundamental necessity of a fire alarm system is to report the occurrence of fire at the earliest. With the advent of computer vision this has become more advanced and reliable. A prerequisite of a Fire Detection System is the detection of fire conditions as early as possible, to provide enough time for Automated Systems/Fire personnel for effective counter actions. Conventional fire detection systems detect and forecasts fire by using by-products of combustible such as smoke, flame, temperature which takes a significant time to develop the required level to trigger heat sensors and smoke sensors. All these situations motivated us to think of a new method of fire detection which uses the technique of computer vision. Computer vision based fire detection systems overcome these limitations since it detects the combustible instead of it are by products. Furthermore, it detects through a camera, which is a volume sensor and covers a wide range from a single camera. The primary objective of this computer vision based fire detection system is to detect the fire, which is done by the method of deep learning, and produce warning alarms if the fire is detected. All of the above clues are combining to form a more efficient fire detection system compared to the conventional systems.

Keywords: Conventional fire detection systems, Convolutional neural network, Computer vision.

View PDF

Refund policy | Privacy policy | Copyright Information | Contact Us | Feedback ©, All rights reserved