A multi-scale object-bases approach to mapping coastal Natura 2000 habitat types using high spatial resolution airborne imagery and LIDAR data

This study was conducted to provide a method for a remote sensing-based inventory of the coastal Natura 2000 habitats. A multilevel object-based approach was used. The habitat types were classified with a hierarchical rule-based key derived from the EUNIS habitat classification system. Data used for this classification were high-resolution true color images, high-resolution false color images, and DTM- and LIDAR-derived vegetation heights. The model developed was tested in the dynamic coastal zone of the island of Ameland