Learning algorithm of fire detection from
remote sensing data
S. M. Zraenko, M. A. Mymrina, V. V. Ganzha
Ural Federal University named after the first
President of Russia B.N.Yeltsin
The paper is received on November 10, 2014
Abstract. An efficiency lack of regular algorithm MOD14 of temperature anomalies
detection and its modifications for the detection of forest fires are shown. The
training threshold algorithm for the analysis of the fire situation on the base
of remote sensing data obtained from a sub-satellite is developed. The
effectiveness of the proposed algorithm by night and day snapshots of spectroradiometer
MODIS is demonstrated.
Key words: forest fires, forest fire internet services, MODIS spectroradiometer,
MOD14 algorithm, trained threshold algorithm.