Title :
A data mining approach to discover collections of homogeneous regions in satellite image time series
Author :
Mougel, Pierre-Nicolas ; Selmaoui-Folcher, Nazha
Author_Institution :
PPME, Univ. de Nouvelle Caledonie, Nouville, New Caledonia
Abstract :
Our work aims at analyzing satellites images time series using a pattern based data mining approach. We consider structures formed by a collection of regions sharing similar pixel properties in images. Such structure enable to discover hidden relations between disconnected regions corresponding to similar objects. The approach has been applied to the analysis of an area in New Caledonia and we present an example of pattern corresponding to several regions having an eroded soil.
Keywords :
data mining; erosion; geophysical techniques; remote sensing; soil; New Caledonia; data mining; eroded soil; homogeneous regions; pixel properties; region sharing collection; satellite image time series; Abstracts; Data mining; Gold; Lakes; Shape;
Conference_Titel :
Geoscience and Remote Sensing Symposium (IGARSS), 2013 IEEE International
Conference_Location :
Melbourne, VIC
Print_ISBN :
978-1-4799-1114-1
DOI :
10.1109/IGARSS.2013.6723800
