Towards DMC microsatellites use in forest fire remote sensing: Case of Alsat-1 product-based false alarm rate assessment

Abstract

In this paper, we studied the contribution of the Algerian Alsat-1 satellite image and its effects on reducing false alarm rates when detecting or monitoring forest fires. We used the classical Support Vector Machines classification method which required positive and negative database training sets. Experiments demonstrate that, such Alsat-1 images, similar products of nearest characteristics satellites ensure very lower rates of false alarm rates without treating about detecting rates.