Vegetation map by satellite imagery in Socotra (Yemen) The present study has produced a high resolution vegetation map of Socotra Island (Yemen) by combining vegetation classification with remote sensing analysis. The satellite data source was represented by multi-temporal sets of RapiEyeTM satellite images with a pixel resolution of 5 m and 5 spectral bands. More than 370 vegetation surveys, carried out with the phytosociological method and used to identify the main vegetation type, were used to obtain the training and evaluation sets. To produce the vegetation map, spectral signatures of the vegetation classes were obtained through Gaussian mixture distribution model. A Sequential Maximun “ a Posteriori” classification method was applied to take into account the heterogeneities in the signatures of some classes and the spatial pattern of vegetation types. Post-classification sorting was performed to adjust the classification through various rule-based operations. A total of 28 classes were mapped with an accuracy greater than 80%. The resulting map and data will represent a fundamental tool for the elaboration of conservation strategies and the sustainable use of natural resources.
Le immagini satellitari ad alta risoluzione e la classificazione delle comunità vegetali per la creazione della carta della Vegetazione dell'Isola di Socotra (Yemen).
ALTOBELLI, ALFREDO
2011-01-01
Abstract
Vegetation map by satellite imagery in Socotra (Yemen) The present study has produced a high resolution vegetation map of Socotra Island (Yemen) by combining vegetation classification with remote sensing analysis. The satellite data source was represented by multi-temporal sets of RapiEyeTM satellite images with a pixel resolution of 5 m and 5 spectral bands. More than 370 vegetation surveys, carried out with the phytosociological method and used to identify the main vegetation type, were used to obtain the training and evaluation sets. To produce the vegetation map, spectral signatures of the vegetation classes were obtained through Gaussian mixture distribution model. A Sequential Maximun “ a Posteriori” classification method was applied to take into account the heterogeneities in the signatures of some classes and the spatial pattern of vegetation types. Post-classification sorting was performed to adjust the classification through various rule-based operations. A total of 28 classes were mapped with an accuracy greater than 80%. The resulting map and data will represent a fundamental tool for the elaboration of conservation strategies and the sustainable use of natural resources.Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.