Currently, a variety of software is capable of handling a wide range of spatial problems, beginning with approaches for describing spatial objects to quite complex analysis and 3D visualization. However, an increasing number of applications need more advanced tools for representing and analyzing the 3D world. Among all types of systems dealing with spatial information, GIS has proven to be the most sophisticated system that operates with the largest scope of objects (spatial and semantic), relationships and provide means to analyse them.
However, what is the status of 3D GIS? It is the aim of this paper to find the answer by analysing both available software and efforts of researchers. An overview of several software packages and 3D case studies performed in Oracle and Microstation is given to provide knowledge about the 3D functionality offered by commercial systems. The most significant achievements in the 3D research area concerning key issues of 3D GIS, i.e. 3D structuring and 3D topology are summarized to portray the current research status. At the end, some of the issues and problems involved in developing such a system are presented and recommendations on directions for further research are made. The scope of the paper is limited to 3D GIS systems and research in vector domain. Problems of subsurface applications are excluded as well.
The need for 3D information is rapidly increasing. Currently, many human activities make steps toward the third dimension, i.e. urban planning, cadastre, environmental monitoring, telecommunications, public rescue operations, landscape planning, transportation monitoring, real estate market, hydrographical activities, utility management, military applications. Practically, the area of interest grows significantly when 3D functionality is offered on the market.
Moreover, the role of geo-information in all kinds of business processes is getting quite transparent. Terms such as “location-specific information” and “location-based services” become a part of the daily business language to denote the link between the virtual world of information (transactions, events, internet communication) and the real world of information – customers, inventory, shipping and the like. Most business transactions rely on information systems to operate successfully, as the geoinformation (location-specific information) is critical for many of them.
Once the developments in 3D GIS provide a compatible functionality and performance, the spatial information services will evolve into the third dimension. Traditionally, GIS system should be able to maintain information about spatial phenomena and provide means to analyse it and thus gain knowledge of the surrounding world. In general, consensus on the demanded functionality of GIS was achieved years ago. The tasks or the functions of a GIS are specified as follows: 1) data capture, 2) data structuring, 3) data manipulation, 4) data analysis, and 5) data presentation.
Indeed, 3D GIS aims at providing the same functionality as 2D GIS. Unfortunately, such 3D systems are still not available on the market. A variety of different software is employed to maintain the objects of interest and extract the required information.
Due to the deficiencies of all the systems to handle 3D objects, the data are often spread between several systems. For example, one system is used for data storage and another for 3D visualisation. This situation often faces inconsistency problems, which leads to extra time, efforts and money to find the appropriate solution.
This research summarises the current status of 3D GIS development. Firstly, we concentrate on recent achievements reported by vendors. We briefly present our survey on the possibilities of some GIS’s available on the market and analyse two case studies completed on commercial systems. Secondly, we review attempts by researchers toward providing an appropriate structures and operations for 3D spatial analysis and visualisation. Final discussion recommends directions and topics for further research and implementations.
2. 3D GIS in the market
There are few commercial-of-the-shelf (C.O.S.T) systems that can be categorised as systems that attempt to provide a solution for 3D representation and analysis. Four systems are chosen for detailed consideration, because they constitute a large share of the GIS market and provide some 3D data processing functions.
The systems are the 3D Analyst of ArcGIS (see ESRI Inc.), Imagine VirtualGIS (ERDAS Inc., http://www.erdas.com), GeoMedia Terrain (Intergraph Inc., http://www.integraph.com) and PAMAP GIS Topographer (PCIGEOMATICS, http:// www.pcigeomatics.com). Parts of the following text are based on available literature and Webbased product reviews.
2.1 Traditional GIS vendors
ArcGIS 3D Analyst, ESRI: The 3D Analyst (3DA) is one of the modules available in ArcGIS GIS. ArcView is designed to provide stand alone and corporate wide (using client-server network connectivity) integration of spatial data. With 3DA one can manipulate basically 2.5D data such as surface generation, volume computation, draping raster images, terrain inter-visibility from one point to another. The system works mainly with vector data. Raster files can be incorporated into 3DA, but only for improving the display of vector data.
During the last three years, ESRI has further developed the 3D Analyst for the ArcGIS 8.1 environment. ArcGIS consists of the Desktop and Workstation components. The Desktop component is based on personal computer (PC) and Microsoft Windows operating system, while the Workstation component is available for both PC and UNIX platforms. ESRI also introduced a new ArcScene desktop application as part of the 3D Analyst extension to ArcGIS 8.1 .
ArcScene is a stand-alone application that provides all the capabilities similar to 3DA with enhanced 3D visualization, flyby, texture mapping on building facades, 3D symbols, animation and surface analysis for both raster and vector data. Commonly used CAD data formats (e.g. DGN, DXF, DWG) can directly be read and displayed in ArcScene. ArcScene can also access and display both raster and vector data stored on the multiuser geographic database using ESRI Spatial Data Engine (SDE) or data service on the Internet in the distributed environment using ESRI ArcIMS. Although major progress on improving 3D visualization, animation, and data access has been made, full 3D geometry for 3D representation, topological relationships and analysis remain as areas to be addressed.
Imagine VirtualGIS, ERDAS: It is worth mentioning that the Imagine system was originally developed for remote sensing and image processing tasks. The recently provided GIS module is called VirtualGIS and supplies some 3D visual analysis tools. It is a system that has an emphasis on dynamic visualisation and real-time display in the 3D display environment. Interesting 3D visualisation features of the system are: 1) the possibility to include rapidly 3D models in a selected polygon or along a line (e.g. 3D models of trees in a forest area) and 2) the logo layer that can accommodate a 2D image into the 3D scene and stretch it over the entire view as foreground. Besides these and other extensive 3D visualizations, the system also provides fly-through capabilities. As with 3DA this system also centres on 3D visualization with true 3D GIS functions hardly available.
GeoMedia Terrain, Integraph.Inc: GeoMedia Terrain is one of the subsystems that work under the GeoMedia GIS. The system runs under the Windows operating systems. The Terrain system performs three major terrain tasks, namely, terrain analysis, terrain model generations, and flythrough. The navigation tool ActiveFlight provides the three most common navigation modes — free flight, constant above ground elevation and terrain avoidance. It also offers the ability to save and restore viewpoints. In general, the GeoMedia Terrain serves as DTM module for the GeoMedia GIS without true 3D GIS capabilities.
PAMAP GIS Topographer, PCIGeomatics: It runs under Windows operating systems. PAMAP GIS is a raster and vector system considered by many (Limp, 2002) the most integrated system (i.e. points and lines are represented as vectors and areas as raster). Four main GIS modules, i.e. Mapper, Modeller, Networker and Analyser form the core system. Topographer, similarly to other terrain systems, provides terrain analysis and display system with the ability to perform a number of functions including the calculation of slope, aspect, visibility, and the generation of perspective views. As a decision-making tool, Topographer gives answers to visibility and planning issues, such as site location and viewing areas. Besides 3D terrain data, Topographer can process any 3D data, including temperature, rainfall and pollution concentrations. Again, true 3D objects cannot be handled by the system In summary, all the systems revealed little provision of 3D GIS functionality in terms of 3D structuring, 3D manipulation and 3D analysis but most of them can efficiently handle 3D data in the 3D visualization aspect. A fully integrated 3D GIS solution has yet to be offered by general purpose GIS vendors.
The GIS, i.e. integration of semantic and geometric data, and spatial relationships, seems to be the most appropriate system ensuring a large scope of analysis and thus serving many applications and daily activities. Therefore, vendors dealing with either spatial or semantic information attempt to achieve the GIS functionality already for years. CAD vendors (such as Autodesk, Bentley) provide means to link 2D, 3D geometry to semantic data and organise topologically structured layers; DBMS (Oracle, Informix) introduce spatial descriptors to represent geometry data and maintain them together with the semantic data.
A logical consequence of all the attempts is the agreement on the manner for representing, accessing and disseminating spatial information, i.e. the OpenGIS specifications (http://www.opengis.org/). This agreement makes possible efforts of vendors and researches from different fields to be united and streamed to one direction, i.e. development of a functional GIS. The result is the changed role of DBMS. This is to say, it becomes Geo-DBMS, an integrated “container” of semantic and geometric properties of real-world objects attempting to provide functionality to store, retrieve and analyse spatial data.
Here, we will briefly present the representation of geometry in Oracle Spatial (http: //technet.oracle.com/products/spatial/content.html).
The geometric characteristics of spatial objects are defined by the geometric type. Currently, the supported geometric types are 2D but 3D coordinates are accepted. The rules for objects description are very simple and intuitive. Lines and polygons are represented as an ordered set of coordinates (2D or 3D).
Self-intersecting lines are allowed but self-intersecting polygons are not supported. Polygons with holes are maintained as well. Oracle is an object-relational DBMS and the geometric types are defined using exactly the object-oriented approach. They are defined in the mdsys.sdo_geometry object-relational model and contain information about type, dimension, coordinate system, holes of objects, and provide a list with the coordinates.
3. Experiments with Oracle Spatial and GeoGraphics iSpatial
Several CAD/GIS applications (MicroStation, AutoCAD, MapInfo) already embrace OpenGIS specifications. We have selected MicroStation Geographics iSpatial (http://www2.bentley.com/ products/default.cfm) to investigate the 3D functionality offered at present. GeoGraphics iSpatial establishes a connection directly to Oracle Spatial.
The spatial objects within GeoGraphics iSpatial are defined in a hierarchical structure. Project refers to as the root and represents the data for the entire study area. The second level is the category, which groups features with a similar theme (e.g. buildings, rivers). One project can have many categories but a category may belong to only one project. Feature is at the third level and represents one or more objects in the real world (e.g. the bank building, the school building). A feature incorporates all the attribute and geometric data available for a particular real object.
A category may have many features but a feature may belong to only one category. Feature is the basic structural unit in GeoGraphics iSpatial. To be able to distinguish between different spatial objects stored in Oracle Spatial 8i, each object has to be assigned to a feature. Furthermore, edited and newly created objects cannot be posted in the database without attributing predefined features to them. Geometry of the objects is organised in one or more spatial layers.
To investigate the functionality of the two software products in representing, maintaining and visualising 3D spatial objects, we have completed two case studies following two different approaches. In the fist case study, we have the 3D data organised in Oracle Spatial in user-defined relational tables and the task was to access, query and edit them from GeoGraphics. In the second case, the 3D data were available in a DGN file and had to be imported in Oracle Spatial.
4. 3D GIS in the research
The research in 3D GIS is intensive and covers all aspects of acquisition, storage and analysis of real world phenomena. Among all, 3D analysis and other related issues (topological models, frameworks for representing spatial relationships, 3D visualisation) are mostly in the focus of investigations.
Topological model: The topological model is closely related to the representation of spatial relationships, which are the basis of a large group of operations to be performed in GIS, e.g. inclusion, adjacency, equality, direction, intersection, connectivity, and their appropriate description and maintenance is inevitable. Several 3D models have already been reported in the literature. Each of the models has strong and weak points for representing spatial objects.
Formalism for detecting spatial relationships: OpenGIS consortium has adopted two frameworks to detect spatial relationships known as Egenhofer operators and Clementini operators based on the 9-intersection model (see Egenhofer and Herring, 1992, Clementini and Felice, 1994). Although the topology is considered the most appropriate mechanism to describe spatial relationships, the study on other mathematical frameworks continues. Billen et al (2002) propose another framework (i.e. the Dimensional model) for representing spatial relationships, built up in affine space and convexity properties of the constructing elements (named dimensional elements). The Dimensional model allows larger variations in the grouping of spatial relationships compared to the 9- intersection model.
Data Presentation: Advances in the area of computer graphics have made visual media a major ingredient of the current interface in the communication and interaction with computers.
Therefore, research related to the visualisation of real world 3D data is mostly “shifted” to the computer graphics society. Many viewers and browsers as stand-alone applications and plug-ins have been developed to quickly visualise and navigate through 3D models for a variety of applications. New algorithms and implementations are reported daily.
The design criteria, however, are fast rendering techniques based on internal structures rather than utilisation of database representations. TerraExplorer(http://www.skylinesoft.com/corporate/corporate_home.asp), the current leader for visualising large 3D textured data from the real world and the first software with acceptable performance, also requires re-structuring of data.
3D Web applications: The Web has already shown a great potential in improving accessibility to 2D spatial information (raster or vector maps) hosted in different computer systems over the Internet. 3D data were not even transferable over the Web until five years ago. The first attempt to disseminate and explore 3D data, i.e. VRML, appeared to be rather “heavy” for encoding real geodata due to the lack of a successful compression concept.
Despite the drawbacks, the language became a tool for research visualization. Researchers could concentrate on data structuring and analysis and leave the rendering issues to browsers offered freely on Internet. The research on spatial query and 3D visualization utilizing VRML has resulted in a few prototype systems (Coors and Jung 1998, Lindenbeck and Ulmer 1998, Zlatanova 2000). GeoVRML (VRML extended with geo-nodes) and the new standard X3D (http://www.web3d.org/news/x3d/) are among promising opportunities for representing 3D data on the Web. A key goal of the X3D standard is definition of interoperable, light-weight components for 3D Web and broadcast applications beyond VRML 97.
In this research we reported our observations and experience with current 3D GIS developments. The major 3D progress is observed in the area of data presentation. All traditional GIS vendors provide extended tools for 3D navigation, animation and exploration. However, many of these systems still are lacking full 3D geometry. OpenGIS specifications seem to have been adopted rapidly by DBMS, CAD, and GIS developers. In this order of thoughts, the understanding for GIS is changing. Instead of a monolith, desktop, individual system, GIS is becoming an integration of strong database management (ensuring data consistency and user control) and powerful editing and visualisation environments (inheriting advanced computer graphics achievements).
At present, only the first step is made, i.e. the implementations focus mostly on the geometry. 2D topological representations and operations are intensively in the process of implementation. The third dimension with respect to topological issues is still in the hands of the researchers. The case studies clearly showed the benefits of a standardised spatial data structuring as well as revealed the very early stage of the integration. The large number of specialised settings, the application dependent feature-geometry linkage, the limited semantic hierarchy, the spatial operators utilising only X, Y values, are some of the issues that need further improvement.
Although, there are quite a significant number of works devoted to 3D data structuring, the research is concentrated around a few basic ideas, as the level of explicitly described spatial relationships varies. Each suggested data structure exhibits efficiency and deficiency with respect to particular applications and operations to be performed. Still 3D GIS functionality needs to be addressed: 3D buffering, 3D shortest route, 3D inter-visibilities are some of the most appealing for research. Integration of object-oriented approaches with the 3D GIS raises research topics at a database level toward standard object descriptors and operations. 3D visualisation within 3D GIS requires appropriate means to visualise 3D spatial analysis tools to effortlessly explore and navigate through large models in real time. Observations on the demand for 3D City models show user preferences for photo-true texturing. Trading photo-true texture brings up necessities to store parameters for mapping onto the geometry.
Abdul-Rahman, A., 2000. The design and implementation of two and three-dimensional triangular irregular network (TIN) based GIS. PhD thesis, University of Glasgow, UK. 250pp.
Billen, R., Zlatanova, S., Mathonet, P. and Boniver, F., 2002. The Dimensional model: a framework to distinguish spatial relationships. In: Richardson, D. and van Oosterom, P. (Eds.), Advances in Spatial Data Handling, Springer-Verlag, Berlin, pp.285-298.