Automated detection of buildings on aero images

Abstract

One of the challenging problems in photogrammetry is extracting of three-dimensional objects from aero images, in particular, extraction of different kinds of buildings. All methods that provide satisfactory results are rather time consuming and process data quite long. In the paper we propose a method that detects areas on aero images that might contain a building behind them. Our method allows reducing the amount of data which should be processed by more complex algorithms. This leads to reducing of the total time spent on extraction process.