Abstract | ||
---|---|---|
This paper addresses the problem of classifying altimetric waveforms backscattered from different kinds of surfaces including oceans, ices, deserts and forests. Appropriate features associated with altimetric radar waveforms are first introduced for this classification. These features are completed by radiometer temperatures and pre-processed using a linear discriminant analysis for dimensionality reduction. The classification of altimetric waveforms is finally achieved using the resulting pre-processed vector with reduced dimension. Different classification strategies are finally considered. These strategies are based on the nearest mean rule, the nearest neighbor method or on the multilayer perceptron. Various simulation results illustrate the performance of the proposed classifier. |
Year | DOI | Venue |
---|---|---|
2008 | 10.1109/IGARSS.2008.4779286 | Geoscience and Remote Sensing Symposium, 2008. IGARSS 2008. IEEE International |
Keywords | Field | DocType |
geophysical signal processing,multilayer perceptrons,radar altimetry,remote sensing by radar,Ku band,altimetric radar waveforms,altimetric signal classification,backscatter,classification strategies,deserts,dimensionality reduction,forests,ices,linear discriminant analysis,multilayer perceptron,oceans,radiometer temperatures,Altimetry,classification,linear discriminant analysis,multilayer perceptron,nearest neighbor rule | Nearest neighbour algorithm,Altimeter,Dimensionality reduction,Computer science,Waveform,Remote sensing,Multilayer perceptron,Linear discriminant analysis,Classifier (linguistics),Radiometer | Conference |
Volume | ISBN | Citations |
3 | 978-1-4244-2808-3 | 3 |
PageRank | References | Authors |
0.47 | 1 | 4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Jean-Yves Tourneret | 1 | 835 | 64.32 |
Corinne Mailhes | 2 | 99 | 20.48 |
Laiba Amarouche | 3 | 3 | 0.47 |
Nathalie Steunou | 4 | 3 | 0.47 |