Deriving leafless trees and urban structures from the DLR 3K airborne camera system for the city of Braunschweig

Thomas Krauß

doi:10.1117/12.2534440

Translator Disclaimer

27 June 2019 Deriving leafless trees and urban structures from the DLR 3K airborne camera system for the city of Braunschweig

Thomas Krauß

Author Affiliations +

Proceedings Volume 11174, Seventh International Conference on Remote Sensing and Geoinformation of the Environment (RSCy2019); 1117416 (2019) https://doi.org/10.1117/12.2534440
Event: Seventh International Conference on Remote Sensing and Geoinformation of the Environment (RSCy2019), 2019, Paphos, Cyprus

Abstract

In this work we present a method for detecting leafless trees in digital elevation models derived from high resolution aerial imagery. No spectral classification as usually conducted is needed for our approach. This makes the method applicable for stereo imagery using only panchromatic or RGB imagery or even acquisitions in early spring showing trees without foliage. The method is based on statistics of densely calculated aspect and slope angles in the DSM. Using a segmentation based on the watershed transformation gives areas large enough for reasonable statistics but also limited to individual tree or building objects. Removing the detected trees from the DSM allows the calculation of a building mask which can in turn be used for the urban modeling of buildings. The detected trees will be modeled in parallel and the models joined together afterwards. The development of the method and the evaluation is based on images acquired with the 3K camera system developed by DLR in early spring 2015 over the german city of Braunschweig. After description of the algorithm the selection of the main classification parameter is shown together with a manual quality assessment.

Citation Download Citation

Thomas Krauß "Deriving leafless trees and urban structures from the DLR 3K airborne camera system for the city of Braunschweig", Proc. SPIE 11174, Seventh International Conference on Remote Sensing and Geoinformation of the Environment (RSCy2019), 1117416 (27 June 2019); https://doi.org/10.1117/12.2534440

ACCESS THE FULL ARTICLE