3D GIS, where are we standing? Jantien Stoter and Siyka Zlatanova Section GIS technology, Delft University of Technology, The Netherlands {j.e.stoter|s.zlatanova}@citg.tudelft.nl 1. Introduction Since early ’90 GIS has become a sophisticated system for maintaining and analysing spatial and thematic information on spatial objects. The need for 3D information is rapidly increasing. 2D GIS analysis has shown its limitations in some situations, e.g. noise prediction models (noise spreads out in three dimensions) (Kluijver and Stoter, 2003), water flood models, air pollution models, geological models (Van Wees, 2002), real estate market (CEUS, 2003; Stoter and Ploeger, 2003). Other disciplines that have met the need for 3D geo-information are: 3D urban planning, environmental monitoring, telecommunications, public rescue operations, landscape planning. The breakthrough of 3D GIS seems to go slow. The developments in the area of 3D GIS are pushed by a growing need for 3D information from one side and new technologies on the other side. In (Zlatanova et al., 2002) the side of new technologies were addressed by discussing the current status of 3D GIS considering developments reported by vendors and researchers. In VI Matrix (a Dutch magazine for real estate) the developments of 3D GIS in the Netherlands were described (Oosterom et al., 2002 (1); Oosterom et al., 2002 (2)). This paper continues on these discussions and gives an extended overview on the status of 3D GIS in both research and practise. We start with a description of technology developments in 3D GIS. Then we address the main complexities of 3D GIS: organisation of 3D data, 3D object reconstruction, and visualisation and navigation in 3D environments. 3D analysing and 3D editing can be added to this list. More on 3D analysing can be found in (Zlatanova and Stoter, 2003). An example of the implementation of a specific 3D analysis, 3D buffering, is described in (De Vries, 2001). More on 3D editing can be found in (Stoter and Zlatanova, 2003). We end this paper with a discussion on where to go for a serious breakthrough of 3D GIS. 2. Technology progress An important development is improvement of 3D data collection techniques (aerial and close range photogrammetry, airborne or ground based laser scanning, surveying and GPS). Sensors are faster and more accurate than before. Other new techniques that push 3D GIS developments are hardware developments: processors, memory and disk space devices have become more efficient in processing large data sets, especially graphics cards also used by gaming industry. Furthermore elaborated tools to display and interact with 3D data are evolving. GIS software-tools have also made a significant movement towards 3D GIS. Zlatanova et al., 2002 present a survey on mainstream GIS software: ArcGIS (Esri, 2003), Imagine VirtualGIS (Erdas, 2003), PAMAP GIS Topographer (PCIGeomatics, 2003) and GeoMedia Terrain (Integraph, 2003). Zlatanova et al., 2002 conclude that major progress in 3D GIS has been made on improving 3D visualisation and animation. However, 3D functionality is still lacking such as generating and handling (querying) 3D geo-objects, 3D structuring, 3D manipulation and 3D analyses (3D overlay, 3D buffering, 3D shortest route). This is caused by the specific character of 3D data compared to 2D. Bottlenecks are still organization of 3D data, 3D object reconstruction, and representation and navigation through large 3D models. These three issues are addressed in sections 2.1 to 2.3. 2.1 Organisation of 3D data 3D representations For modelling 3D objects, several 3D abstractions are possible (Mäntylä, 1988). Constructive Solid Geometry (CSG) is an approach for modelling 3D objects by solids. Solid modelling has its origin in CAD. The basic primitives in constructive modelling are spheres, cubes, and cylinders and they are used with varying parameters. Set operations are applied to the basic primitives to construct 3D bodies, such as union, intersect and difference. The advantages of CSG that they are good in computer-aided manufacturing: a brick with a hole drilled through it is represented as “just that”. The disadvantages for real world modelling are that the objects and their relationships might become very complex.