Centre de documentation
scientifique

Regelbasierte Interpretation unstrukturierter Vektorkarten / M. Weindorf (2002)

Public Titre : Regelbasierte Interpretation unstrukturierter Vektorkarten Titre original : [Interprétation basée sur des règles de cartes vectorielles non structurées] Type de document : Thèse/HDR Auteurs : M. Weindorf, Auteur Editeur : Munich : Bayerische Akademie der Wissenschaften Année de publication : 2002 Collection : DGK - C Sous-collection : Dissertationen num. 555 Importance : 74 p. Format : 21 x 30 cm ISBN/ISSN/EAN : 978-3-7696-9594-6 Note générale : Bibliographie Langues : Allemand (ger) Descripteur : [Vedettes matières IGN] Applications SIG [Termes IGN] aide à la décision [Termes IGN] base de données localisées [Termes IGN] cartographie numérique [Termes IGN] intelligence artificielle [Termes IGN] interprétation automatique [Termes IGN] moteur d'inférence [Termes IGN] représentation graphique [Termes IGN] système d'information géographique [Termes IGN] vectorisation Résumé : (Auteur) Maps provide a rich set of geographic information. Formerly produced in analogue way to support human users in their tasks concerning spatial decisions, nowadays digital information systems in general and geographical information systems (GIS) are up to take over this role. Today not only the human user is addressed, but the interoperability of information systems and the support of automatic selection of information from different sources is a major task in research and industry. This work presents a concept for the automated interpretation of digital maps represented as low-level, a priori unstructured vector data. A huge amount of digital map data is stored in CAD based systems which only support a geometry focused and layer-oriented concept of data modelling. To use and import this legacy data into modern schema based geoinformation systems, the geometry inherent object descriptions have to be inferred out of these graphics. Nowadays data exchange is often still made on a graphical level of representation, in the way that for every combination of data source and data sink a specific mapping (at a syntactic level) has to be defined, to translate the several data formats into other ones. Problems arise if the expressive power of these data formats differs in the way that the source format is less powerful than the target format (e.g. there's no representation of topology available, or attribute data can't be handled). The idea proposed in this work is based on the assumption that the graphic itself may transport most of the information within a map. Therefore a knowledge-based system for map interpretation was developed, which is based on a backward-chaining rule system. The objects, which have to be extracted out of the maps, are described in their structure within a specific rule base, a corresponding interpreter (inference engine) derives objects out of that descriptions, which may be stored in a database oriented geoinformation system for further use. The focus of this work lies in the definition of structure measures respective rules, which are especially suited to infer application specific objects from graphical representations. An algorithm is provided to derive topological structured maps from a priori unstructured vector data. This map data is transformed into a relational structure, which serves as a basis for the subsequent interpretation process. This interpretation relies mainly on structural and contextual information, whereas numerical classification methods may be avoided to a large extend. Map objects are defined through their parts, spatial relations between these parts and the relations to other objects (contextual information). The object recognition is formulated as a search for subgraphisomorphisms within the map graph. Numéro de notice : 15026 Affiliation des auteurs : non IGN Thématique : GEOMATIQUE Nature : Thèse étrangère Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=55044

Exemplaires(2)

Code-barres	Cote	Support	Localisation	Section	Disponibilité
15026-02	37.40	Livre	Centre de documentation	Géomatique	Disponible
15026-01	37.40	Livre	Centre de documentation	Géomatique	Disponible