Two-level model for land degradation mapping on multispectral satellite imagery

Abstract

Two-level model for land degradation mapping on multispectral satellite imagery of low and medium spatial resolution is shown. First level model applies several different thematic classifications of source multispectral images, e.g. vegetation change, soil erosion etc. Second level gives data fusion of specific thematic classifications of the first level into final thematic map to improve accuracy and reliability of wing to joint interpretation. Proposed model will be useful for information support systems to provide land management.