Eco-geographic classification based on time series of HRPT [High Resolution Picture Transmission] NDVI [Normalized Difference Vegetation Index] Images
1998
Griguolo, S. | Santacroce, P. (DAEST, Venezia (Italy))
NDVI (Normalized Difference Vegetation Index) GAC (Global Area Coverage) images from the AVHRR (Advanced Very High Resolution Radar) sensor are quite a common input to monitor vegetation in vast areas in Africa. Both monthly and decade images are regularly produced and distributed, and the evolution of phenomena of interest can be analysed from a time series of such images. An analytical method, the core of which is a Principal Component Analysis followed by a non-hierarchical Clustering procedure, has recently been proposed and implemented on Personal Computer. The software, named ADDAPIX (Analisi Dei DAti su PIXel), is distributed by FAO. Pixels are grouped according to the similarity of their time patterns, thus creating a convenient number of classes. The time profile of any pixel can then be displayed in order to investigate local peculiarities. This paper deals briefly with problems and advantages related to the use of the same method for eco-geographic classifications on high-resolution (HRTP) data, which are getting more and more easily available. The results of an analysis carried out in Kenya and based on high-resolution NDVI images are reported in some detail.
Mostrar más [+] Menos [-]Palabras clave de AGROVOC
Información bibliográfica
Este registro bibliográfico ha sido proporcionado por University of Liège