CORINE Land Cover : Automatic or semi-automatic updates
The CORINE Land Cover Map shows both land cover and land use and is based on visual interpretation of satellite photographs, supported by existing maps and aerial photographs. This article introduces some of the problems involved in developing a method of automatically updating the map. The method requires considerable computing power, since two sets of satellite photographs have to be classified and simultaneously compared to CORINE. All changes observed then have to be presented to an image interpreter, who makes the final decisions. To perform the calculations, special so-called neural networks with parallel processing are to be developed.
Bibliographic Reference: Article: Tidsskrift for Dansk Kartografisk Selskab
Record Number: 199510864 / Last updated on: 1995-08-03
Original language: da
Available languages: da