In this paper, we exploit the possibility to have SAR data with different ground resolution to characterize different fusion methodologies. We consider fuzzy algorithms, based on the fuzzy-c-means procedure, and applied to a data set of AIRSAR and SIR C-band SAR images. First, we consider a pyramidal approach, starting from coarse data analysis and using the higher details to add precision to the classification map. Then, a spatial enhancement algorithm has been implemented to provide a guess of the details of the coarse resolution data. The second approach allows obtaining better classification results as long as we consider only the soil classes that it is possible to identify in both the low and high resolution data. No serious advantage is instead found for the investigated procedure when a more detailed classification map is searched
Fuzzy pyramidal joint classification of SIR-C and AIRSAR data
DELL'ACQUA, FABIO;GAMBA, PAOLO ETTORE
2001-01-01
Abstract
In this paper, we exploit the possibility to have SAR data with different ground resolution to characterize different fusion methodologies. We consider fuzzy algorithms, based on the fuzzy-c-means procedure, and applied to a data set of AIRSAR and SIR C-band SAR images. First, we consider a pyramidal approach, starting from coarse data analysis and using the higher details to add precision to the classification map. Then, a spatial enhancement algorithm has been implemented to provide a guess of the details of the coarse resolution data. The second approach allows obtaining better classification results as long as we consider only the soil classes that it is possible to identify in both the low and high resolution data. No serious advantage is instead found for the investigated procedure when a more detailed classification map is searchedI documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.