GIS systems are an irreplaceable tool for “decision supporting”. In fact, one of the most important characteristic of a GIS is its ability to offer users and planners proper tools in order to question the same system, that is to make some query of different kind and retrieve useful data for any following decision. This is because the GIS reply can “make explicit” the informative contents, already existing in stored data but in a implicit way, because of the incorrelation among them. While concerning the spatial query, these querying tools are the overlay functions and union functions between layer and proximity, in reference to alphanumeric data GIS softwares use the SQL language to make the queries. Retrieved data allow a management and an advanced control of territory. Nevertheless, there are some conditions in which it need to have informations that are unreachable with querying tools above analyzed: for example, to evaluate correlation between cartographic data and non-deterministic variables (i.e. data on environmental variations, social and economical variables, etc...). Because of these requirements, during last years some Artificial Intelligence (AI) based algorithms are implemented and embedded in GIS softwares, in order to allow the automatic or self-controlled creation from raster layers of thematic maps with equivalence classes hardly identifiable by classic methods (such as spectral analysis); other AI-based solutions are applied to an “on the fly” overlay between raster and vectorial layers. The aim of our work is to illustrate a further possibility for usage of intelligent algorithms with GIS; in fact, we want to show how it is possible to implement new spatial querying operators (Overlay A.I., Union A.I., Proximity A.I.) by means of neural networks, with a redefinition of classical ones, and the possibility to obtain the “fuzzy” results above described in a typical application of heritage conservation and mapping planning sciences.
Application of mapping plan with a non-deterministic algorithm for gis querying / Barrile, Vincenzo; F., Cotroneo; S., Tringali. - In: THE INTERNATIONAL ARCHIVES OF THE PHOTOGRAMMETRY, REMOTE SENSING AND SPATIAL INFORMATION SCIENCES. - ISSN 1682-1777. - 20:(2005), pp. 675-678. (Intervento presentato al convegno The CIPA International Archives for Documentation of cultural Heritage tenutosi a Torino (Italy) nel September 2005).
Application of mapping plan with a non-deterministic algorithm for gis querying
BARRILE, Vincenzo;
2005-01-01
Abstract
GIS systems are an irreplaceable tool for “decision supporting”. In fact, one of the most important characteristic of a GIS is its ability to offer users and planners proper tools in order to question the same system, that is to make some query of different kind and retrieve useful data for any following decision. This is because the GIS reply can “make explicit” the informative contents, already existing in stored data but in a implicit way, because of the incorrelation among them. While concerning the spatial query, these querying tools are the overlay functions and union functions between layer and proximity, in reference to alphanumeric data GIS softwares use the SQL language to make the queries. Retrieved data allow a management and an advanced control of territory. Nevertheless, there are some conditions in which it need to have informations that are unreachable with querying tools above analyzed: for example, to evaluate correlation between cartographic data and non-deterministic variables (i.e. data on environmental variations, social and economical variables, etc...). Because of these requirements, during last years some Artificial Intelligence (AI) based algorithms are implemented and embedded in GIS softwares, in order to allow the automatic or self-controlled creation from raster layers of thematic maps with equivalence classes hardly identifiable by classic methods (such as spectral analysis); other AI-based solutions are applied to an “on the fly” overlay between raster and vectorial layers. The aim of our work is to illustrate a further possibility for usage of intelligent algorithms with GIS; in fact, we want to show how it is possible to implement new spatial querying operators (Overlay A.I., Union A.I., Proximity A.I.) by means of neural networks, with a redefinition of classical ones, and the possibility to obtain the “fuzzy” results above described in a typical application of heritage conservation and mapping planning sciences.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
