Centre de documentation
scientifique

Clarity experimentations for cartographic generalisation in production / François Lecordix (2006)    

Public Titre : Clarity experimentations for cartographic generalisation in production Type de document : Article/Communication Auteurs : François Lecordix , Auteur ; Jenny Trévisan, Auteur ; Jean-Marc Le Gallic, Auteur ; Loïc Gondol , Auteur Editeur : Paris : Institut Géographique National - IGN (1940-2007) Année de publication : 2006 Conférence : ICA 2006, 10th ICA Workshop, Commission on Map Generalisation and Multiple Representation 25/06/2006 Portland Etats-Unis OA Proceedings Importance : 7 p. Format : 21 x 30 cm Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Termes IGN] bati [Termes IGN] chaîne de production [Termes IGN] Clarity (plateforme de généralisation) [Termes IGN] généralisation cartographique automatisée [Termes IGN] réseau routier [Vedettes matières IGN] Généralisation Résumé : (Documentaliste) Le besoin d'un logiciel basé sur le système multi-agents pour la généralisation automatique est apparu et a réuni plusieurs organismes cartographiques nationaux dans leurs recherches. Clarity a permis de produire des cartes après généralisation des données géographiques. Ce logiciel améliore certains défauts du système multi-agents et va encore évoluer. Numéro de notice : 13684 Affiliation des auteurs : COGIT (1988-2011) Thématique : GEOMATIQUE/INFORMATIQUE Nature : Communication DOI : sans En ligne : https://kartographie.geo.tu-dresden.de/downloads/ica-gen/workshop2006/ICA2006-Le [...] Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=64267

Documents numériques

en open access 13684_ica2006_lecordix.pdf Adobe Acrobat PDF

How to merge optimization and agent-based techniques in a single generalization model? / Julien Gaffuri (2006)    

Public Titre : How to merge optimization and agent-based techniques in a single generalization model? Type de document : Article/Communication Auteurs : Julien Gaffuri , Auteur Editeur : Paris : Institut Géographique National - IGN (1940-2007) Année de publication : 2006 Conférence : ICA 2006, 10th ICA Workshop, Commission on Map Generalisation and Multiple Representation 25/06/2006 Portland Etats-Unis OA Proceedings Importance : 15 p. Format : 21 X 30 Note générale : Bibliographie Langues : Anglais (eng) Descripteur : [Termes IGN] agent (intelligence artificielle) [Termes IGN] généralisation cartographique automatisée [Termes IGN] optimisation (mathématiques) [Termes IGN] système multi-agents [Vedettes matières IGN] Généralisation Résumé : (Auteur) [introduction] Many works in generalization automation concern the conception of generalization models. The role of generalization models is to get a complete framework to perform the complete generalization of a geographic dataset. In most of them, generalization is seen as a constraint satisfaction problem. Constraints are made explicit following (Beard, 1991), and are a translation of the final map requirements. Some constraints concern the legibility of the objects (for example, objects must not be too closed), and force their geometry to change (too closed objects are displaced), while other constraints force to preserve some characteristics of the objects (an object should preserve its position and its shape). Generalization models aim to find a way to manage the satisfaction of these change and preservation constraints. In this paper, we focus especially on two families of generalization models: Optimization- based models, illustrated by the works of Sester (2005), Harrie & Sarjakoski (2002), Højholt (2000), Bader (2001), Burghardt & Meier (1997), and Agent-based models of Duchêne (2004), Ruas (1999), and the AGENT project (Barrault et al., 2001). An important difference between optimization and agent-based models comes from the way the constraints are considered. In the optimization models, the constraints are satisfied altogether in one step, using a global resolution method to find a compromise between them: all the constraints are “elastic” and a balance between them is found. In the agent-based models, constraints are satisfied step by step, by triggering an algorithm to solve an identified cartographic conflict. The constraints are satisfied depending on an importance value. The result is not a compromise between the constraints: the most important constraints are satisfied totally while others, less important, are relaxed. These two families of models have provided very good improvements and are now used in several map series production lines as presented in (Lemarié, 2003; Lecordix, 2005). The purpose of this article is to show that these models have different application fields and combining it would allow to improve the automatic generalization process. We show that optimization-based models are much more adapted to compute “continuous transformations”, such as deformations, while agent-based models are adapted to “discrete transformation”. We introduce the notion of “malleable” and “rigid” objects. Numéro de notice : 13683 Affiliation des auteurs : COGIT (1988-2011) Thématique : GEOMATIQUE/INFORMATIQUE Nature : Communication DOI : sans En ligne : https://kartographie.geo.tu-dresden.de/downloads/ica-gen/workshop2006/ICA2006-Ga [...] Format de la ressource électronique : URL article Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=64266

Documents numériques

en open access 13683_ica2006_gaffuri.pdf Adobe Acrobat PDF