Home Smoke detection
Smoke detection algorithms PDF Print
Written by Toni Jakovcevic   
Tuesday, 09 February 2010 13:52

Smoke-detection is a complex task and most of the algorithm authors have their own specific approach to solving such problem. Most of the algorithms have multiple phases or methods of detection (e.g. motion detection followed by region energy estimation) so the different phases of detection are categorised. Some of main categories are specified here, and for each category methodologies from different authors are referenced.

Main categories are:

The number of alarms depends largely on manually set detection sensitivity parameters that are specific for each phase of the detection. To improve the detection, it is possible to adaptively estimate those parameters using GIS and augmented reality:
Video database used in this article is available at the following link.
Last Updated on Wednesday, 05 February 2014 20:09
Wildfire Observers and Smoke Recognition Homepage, Powered by Joomla!; Joomla templates by SG web hosting