Burned area mapping system and fire detection system, based on neural networks and NOAA-AVHRR imagery

New automatic systems for mapping burned areas and for fire detection, base d on neural networks, were developed. The Supervised ART-II artificial neur al network was employed. These two newly developed systems were applied for mapping burned areas and for fire detection in the eastern part of Spain, which suffered damage during July 1994. Images from the National Oceanic an d Atmospheric Administration Advanced Very High Resolution Radiometer (NOAA -AVHRR) were used. These systems were tested using different sizes of train ing sets and different dynamic parameters. The data description and methodo logy are discussed. The full algorithm is listed.