You may observed an increased interest in 3D modeling technologies utilization by geographic information systems (GIS) in recent years. 3D models are used to envision future projects (new buildings, communications), to recreate areas from past (ancient cities, castles – see MayaArch3d for example), to help analyze results of various simulation scenarios (traffic, disasters) or to visualize complex geospatial data (places in a city). The use cases are numerous, but in this article, I’d like to focus more on a recently very popular application – 3D modeling of urban areas.
As you probably know, there are multiple mapping providers that offer 3D models of the world or of at least some cities. Probably the most notable 3D experience providers are Nokia/Navteq, Google and Apple. Nokia’s presence in the list may surprise someone, but since 2008 it is the owner of an important mapping provider Navteq, whose maps are used by many large customers. Navteq maps, including the 3D version, are therefore used across industries and are also available online through Nokia Maps service. Google offers its 3D maps in their Google Earth program/browser plugin since 2007 and has also a significant customer base across industries. Apple is a newcomer to the market as it introduced its mapping service (including 3D models) this year in its iOS 6 operating system. You may wonder why Microsoft is missing in the list of providers, but it is not a mistake. They used to offer 3D version of their Bing Maps since 2007, but they discontinued the service in 2010. It is possible, however, that Microsoft will leverage both 2D and 3D mapping data of Nokia/Navteq in the future, due to recent agreements between the firms.
The 3D city models are used in various scenarios. The most important is orientation, where the 3D model enables users to discover places in virtual reality, obtaining spatial awareness even before visiting the place in real. That may be helpful especially in areas where 2D maps are not clear or to attract tourists. 3D models are also starting to be used in Turn-by-turn navigations, where the 3D model may greatly help drivers to understand the route plan in real time (solving situations like driving into a tunnel while the planned route is above ground). 3D city models are sometimes as well used by media to better explain what, where and why is happening. Showing a realistic birds-eye view over a place of interest is much more intuitive to viewers than a map, especially if they don’t know the city well.
The 3D models are becoming more and more important, but how do they do that? How do they create 3D models of entire cities? The answer is not obvious as more approaches were used over time. Google Earth, for example, offered 3D models of several cities since its launch in 2007. These models were hand-made, therefore simplified and expensive to create. Google even provided a tool to create own 3D models that could be submitted to Google Earth directories. Today, a much more sophisticated approach is used, introduced by Nokia/Navteq in 2010 (and now being adopted by Apple and Google). Instead of modeling manually, the models are created fully automatically using data collected from multiple sources: cars equipped with cameras and LIDAR laser sensors, airplanes equipped with high resolution cameras and satellite provided data. Using this method the resulting model is very accurate (recognizes even small objects in the streets, trees) and covered by high-resolution textures. The making-of process is very well described in the video below by Nokia/Navteq.
The urban 3D models are experiencing a boom of interest, both of providers and of customers. The technologies used to create them have evolved to a state where the provided virtual reality resembles the real world very closely, becoming a completely new way to discover our world, in addition to traditional 2D representations.