Data for: Fuzzy approaches provide improved spatial detection of coastal dune EU habitats
Description
Habitat mapping on coastal dunes requires objective and reliable tools, which are still lacking in Europe. Remote sensing techniques are promising, but their effectiveness for mapping coastal dune habitats from satellites remains unclear. In this study, we compare crisp and fuzzy image classification approaches using WorldView-3 imagery to map coastal dune habitats in two Natural Parks of Tuscany (Italy). The dataset includes: 1) shapefile containing the field-sampled points with geographic coordinates (ESPG: 4326) and assigned habitat types (Annex I code, cluster, EUNIS habitat) 2) R script used to perform the image classifications (crisp Random Forests, fuzzy Random Forests, Spectral Angle Mapper, Multiple Endmember Spectral Mixture Analysis)