Rule-Based Modeling for 3D GIS

Sung-Soo Kim, Seong Ho Lee, Jong Hyun Park, Young-Kyu Yang

Abstract: One popular approach for modeling 3D geo-features is to create 3D CAD model manually. In particular, manual 3D geo-feature modeling approach provides a high quality of visualization for 3D GIS features. However, this approach can be time consuming to generate 3D models being created manually and requires high costs. We introduce a novel modeling concept; so-called, Rule-based modeling (RBD) that can model 3D geo-features using 2D profiles and 3D feature attributes. By using rule-based engine and model library, we are able to create 3D geo-features automatically. The RBD component combines a synthetic modeler with efficient LOD (level-of-detail) control for real-time rendering. In addition to, describing the details of our component, we present the comparison result between manual modeling and proposed rule-based modeling in terms of the following criteria: input data domain, interoperability, reusability and modeling cost.