Fire Detection Using Image Analysis

Examensarbete för masterexamen

Please use this identifier to cite or link to this item: https://hdl.handle.net/20.500.12380/112706
Download file(s):
There are no files associated with this item.
Type: Examensarbete för masterexamen
Master Thesis
Title: Fire Detection Using Image Analysis
Authors: Krysell, Henrik
Hofflander, Fredrik
Abstract: A great issue facing many companies is the treat of fire damage and thus, it is vital to have a reliable fire detection system. To be able to detect fire at an early stage can be the difference between an insignificant smoke-damage and a total disaster. Many of these companies have special needs and require more than a regular smoke detection system. Therefore, it is important to be able to quickly and reliably detect fire in all kinds of environments. A regular standard fire detector, which often solely detects smoke, does not work well outdoors or in large premises. Nevertheless, there exists systems that can detect fires in these kinds of environments; however, these systems use expensive special equipment like heat cameras or high-tech camera filters. This thesis describes a fast and cheap way of detecting fire in large open areas, both indoors and outdoors, with the help of ordinary surveillance cameras and a computer using image analysis. The fire is detected using both color and motion properties of the flames. From these properties it is possible to distinguish between fire and humans, excavators, trucks, flags and even blinking lights. Furthermore, the system described in this thesis has the ability to learn and adapt to the different sorts of lighting conditions and cameras used in surveillance today. Experimental results shows that the method proposed in this thesis can successfully detect open flames and have the ability to distinguish between real fire and objects that look like fire.
Keywords: Datalogi;Computer science
Issue Date: 2009
Publisher: Chalmers tekniska högskola / Institutionen för data- och informationsteknik, Datavetenskap (Chalmers)
Chalmers University of Technology / Department of Computer Science and Engineering, Computing Science (Chalmers)
URI: https://hdl.handle.net/20.500.12380/112706
Collection:Examensarbeten för masterexamen // Master Theses



Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.