Annual report for Synthetic Environment Modeling 2012

Authors:

  • Gustav Tolt
  • Peter Follo
  • Johan Hedström
  • Torbjörn Härje

Publish date: 2012-12-28

Report number: FOI-R--3553--SE

Pages: 20

Written in: Swedish

Keywords:

  • Synthetic Environment Models
  • 3D
  • CityGML
  • semantics

Abstract

This report summarizes the activities 2012 in the Synthetic Environment Modeling project. The project addresses questions related to automatic handling of synthetic environment models and methods for creating such models from different types of input data. The project focuses especially on semantically enriched models, e.g. CityGML, and how these can improve the possibilities of using terrain information in different applications. In 2012, the work within the project can roughly be divided into three parts. One is about automated couplings between CityGML and a relational database. The purpose is to enable dynamic and customized access to such models instead of having to treat them as (large) files, which is normally the case. The initial work on this has resulted in a paper for the conference Geoprocessing 2013 that describes the different stages in the process and presents the results from experimental trials with a number of candidate frameworks for this. The second part concerns methods for creating CityGML models. In cooperation with the project Surveillance Systems, a CityGML model of an interior of a building has been created. From 3D data acquired with a portable stereo camera system, planar surfaces can be extracted and classified as wall, roof and floor and finally be represented as a CityGML model. The third part concerns integration and further development of a plug-in for importing CityGML in a software environment, which opens the doors for analysis and thematic visualization of such models in a range of applications. Moreover, a number of contacts have been taken with other projects, which has amongst others resulted in plans for a deeper cooperation with the project Synthetic Actors in 2013. The purpose is to achieve synergies between the two projects and focuses on using terrain properties to make simulated agents move more realistically.