A growing number of local Australian councils have already started implementing their SmartCity vision, which includes the creation of a 3D city model as a foundation dataset. Objects such as buildings and trees can now be classified from the 3D mesh and be individually selected, highlighted and linked to a geodatabase directly in the 3D mesh.
3D city models can be produced using several 3D capture methods. The most common technique consists of capturing LiDAR data as well as ground or/and aerial imagery and extracting 3D buildings to the CityGML standard.
A more recent technique involves creating 3D city models from aerial and ground imagery using oblique photogrammetry methods to produce realistic 3D meshes.
The market is currently dominated by CityGML 3D city models. However the increase in computing power and 3D format support by all the major geospatial software solution providers is encouraging innovative cities to adopt a reality mesh 3D city model as their base dataset.
Smart Cities such as Helsinki have adopted both types of 3D models to gain access to the best of both worlds, but not all cities can afford this luxury and have to make a choice. Both technologies have pros and cons and the choice will ultimately depend on budgets, applications and software capabilities of the end users.
One of the main barriers for adopting reality mesh city models is the fact that all objects are part of a continuous mesh and don’t carry attribute information.
In this example, the City of Unley in South Australia has acquired a high resolution 3D city reality mesh over key locations in addition to a standard resolution 3D city reality mesh.
The 3D mesh has been classified into three classes (building, tree, other) and an underlying database has been linked to each individual classified object. Using the interactive 3D GIS solution Skyline TerraExplorer pro or Cesium, users can now interact and query the integrated classified 3D data.
This capability opens up a whole new range of opportunities to support SmarCities endeavors.