Assessment of different classification algorithms for burnt land discrimination

Author

Zammit, Olivier ; Descombes, Xavier ; Zerubia, Josiane

Author_Institution

INRIA-I3S, Sophia-Antipolis

fYear

2007

fDate

23-28 July 2007

Firstpage

3000

Lastpage

3003

Abstract

In this paper, satellite-based remote sensing techniques are used for assessing the damage after a forest fire. Here, burnt land mapping is based on a single after-fire satellite image (SPOT 5). Both support vector machines (SVM) and traditional classification algorithms such as the K-nearest neighbours or the K-means are used to discriminate burnt from unburnt areas. An automatic method combining K-means and SVM is presented and its performances are compared to more classical methods. Maps produced by the different classifiers are also compared to official ground truth provided by the French Space Agency (CNES).

Keywords

fires; geophysical techniques; image classification; learning (artificial intelligence); support vector machines; terrain mapping; vegetation; CNES; French Space Agency; K-means algorithm; K-nearest-neighbours; SPOT 5 satellite image; Support Vector Machines; burnt land discrimination; burnt land mapping; forest fire; satellite-based remote sensing techniques; supervised learning method; Classification algorithms; Constraint optimization; Ecosystems; Fires; Image classification; Remote sensing; Satellites; Supervised learning; Support vector machine classification; Support vector machines;

fLanguage

English

Publisher

ieee

Conference_Titel

Geoscience and Remote Sensing Symposium, 2007. IGARSS 2007. IEEE International

Conference_Location

Barcelona

Print_ISBN

978-1-4244-1211-2

Electronic_ISBN

978-1-4244-1212-9

Type

conf

DOI

10.1109/IGARSS.2007.4423476

Filename

4423476