Computer-assisted land cover mapping with SPOT in Indonesia

Abstract The capability of SPOT combined with specifically designed classifiers was investigated for computer-assisted land cover/use mapping in Indonesia. Atmospheric conditions and the small size, complexity and dynamic nature of agro-forest systems confuse spectral analyses. In this context, conventional spectral classifiers are inadequate. This led to the development of two classifiers adapted to the large number of spectral classes that are apparent with highresolution satellite data; i.e. an interactive box classifier and a layered textural/contextual classifier. These were tested on two study areas in Java. Preliminary results suggest that in Indonesia SPOT has the potential of a major data source for land cover mapping down to 1:50000/100000 scale.