Introduction of the concept of geometric patrons and application to b buildings with regard to their cartographic generalisation [Introduction du concept de patrons géométriques et application aux bâtiments afin de faciliter leur généralisation cartographique à la volée]
Article
Overview
Research
Additional Document Info
View All
Overview
abstract
Despite significant progress made over the last few years, automatic cartographic generalization remains a difficult task, that more often than not, requires human intervention. That is why current solutions based on the use of algorithms alone cannot meet the requirements of new applications such as Web-based cartography, which requires on-the-fly generalization. This paper presents the concept of geometric pattern-shapes common to several cartographic objects - and proposes their use as part of a cartographic generalization process. Thus, instead of generalizing a cartographic object (e.g., a building) using generalization algorithms to ensure good map readability, a geometric pattern is used to replace the object. For this purpose, the geometric pattern is adjusted to the object%27s geometry using very simple operations such as rotation and displacement. A single pattern can be used for several hundred occurrences of objects, and only a minimal number of parameters distinguishing each pattern-object relation are required. Early tests show that geometric patterns perform four to twenty-seven times better than generalization approaches using algorithms.
Despite significant progress made over the last few years, automatic cartographic generalization remains a difficult task, that more often than not, requires human intervention. That is why current solutions based on the use of algorithms alone cannot meet the requirements of new applications such as Web-based cartography, which requires on-the-fly generalization. This paper presents the concept of geometric pattern-shapes common to several cartographic objects - and proposes their use as part of a cartographic generalization process. Thus, instead of generalizing a cartographic object (e.g., a building) using generalization algorithms to ensure good map readability, a geometric pattern is used to replace the object. For this purpose, the geometric pattern is adjusted to the object's geometry using very simple operations such as rotation and displacement. A single pattern can be used for several hundred occurrences of objects, and only a minimal number of parameters distinguishing each pattern-object relation are required. Early tests show that geometric patterns perform four to twenty-seven times better than generalization approaches using algorithms.
