Resumen

PASCUAL RAMIREZ, Fermín et al. Generic object classificator in AVHRR images. Terra Latinoam [online]. 2010, vol.28, n.1, pp.01-13. ISSN 2395-8030.

This paper describes an algorithm for generic classification of objects contained in advanced very high resolution radiometer (AVHRR) sensor images, based on the spectral signature of the generic objects (soil, soil-vegetation mixture, bodies of water, clouds, etc.). Due to the particularities of the available bands in the AVHRR sensor, we present a specific algorithm using the 3a band (B3a) and another using the 3b band (B3b) of this sensor. The algorithms are based on decision trees that do not require training fields to carry out the classification. The algorithms have been tested using international spectral data bases with satisfactory results in all the cases, obtaining classification precision above ninety percent.

Palabras llave : generic objects classification; AVHRR; spectral signatures; remote sensing.