This involves, to begin with, the intelligent use of terrestrial laser scanners which project laser beams onto buildings from the ground and make recordings of them. These have been matched with a databank containing the most important characteristics of a building, such as doors, walls and window bays allowing an application to compare the measurement data with the known characteristics of buildings and thus attempting to identify what each element is. Two-dimensional topographic information from the land registry was also thrown into the mix to create three-dimensional models. It should also be noted that the databank 'learns' more each time information is entered manually.
Sander Oude Elberink on the hand made use of plane laser data together with two-dimensional topographic information from the land registry in his research. This method is more accurate for images from above because of the nature of the data, while Pu's method is more accurate for images from street level. However these two models can be easily combined producing a very interesting take in solving a problem that has persisted for awhile now.