A cost-effective strategy for multi-scale photo-realistic building modeling and web-based 3-D GIS applications in real estate

Web-based 3-D GIS may be the most appropriate tool for decision makers in land management and development. It provides not only the basic GIS functions, but also visually realistic landscape and architectural detail. It also gives the user an immersive 3-D virtual reality environment through the Internet that is rather different from that obtained merely through text, pictures, or videos. However, in terms of high accuracy and level-of-detail (LOD), the generation of a fully photo-realistic city model is labor intensive and time consuming. At the same time, from the aspect of computer graphics, the result is simply a geometric model without thematic information. Thus, the objective of this study is to propose a cost-effective multi-scale building modeling strategy based on the 2-D GIS building footprint that has rich attributes and to realize its application in the real estate market through a web-based 3-D GIS platform. Generally, the data volume needed for a photo-realistic city model is huge, thus for the purpose of increasing Internet data streaming efficiency and reducing the building modeling cost, a multiple-scale building modeling strategy, including block modeling, generic texture modeling, photo-realistic economic modeling, and photo-realistic detailed modeling is proposed. Since 2-D building boundary polygons are popularly used and well attributed, e.g., as to number of stories, address, type, material, etc., we are able to construct the photo-realistic city model based on this. Meanwhile, the conventional 2-D spatial analysis can be maintained and extended to 3-D GIS in the proposed scheme. For real estate applications, a location query system for selecting the optimum living environment is established. Some geospatial query and analysis functionalities are realized, such as address and road-junction positioning and terrain profile analysis. An experimental study area of 11km² in size is used to demonstrate that the proposed multi-scale building modeling strategy and its integration into a web-based 3-D GIS platform is both efficient and cost-effective.