Deep Learning-based Building Footprint Mapping using High Resolution SAR Data
Titel Konferenzpublikation:
IGARSS 2024 - 2024 IEEE International Geoscience and Remote Sensing Symposium
Konferenztitel:
IEEE International Geoscience and Remote Sensing Symposium (44., 2024, Athens, Greece)
Tagungsort:
Athens, Greece
Jahr der Konferenz:
2024
Datum Beginn der Konferenz:
07.07.2024
Datum Ende der Konferenz:
12.07.2024
Verlag:
IEEE
Jahr:
2024
Seiten von - bis:
11168-11171
Sprache:
Englisch
Stichwörter:
Building Footprints ; Synthetic Aperture Radar ; Machine Learning ; Deep Learning
Abstract:
Investigating the synergy of deep learning and high-resolution Synthetic Aperture Radar (SAR) data, this paper focuses on building footprint extraction -- a domain traditionally dominated by optical imagery. The proposed method involves projecting SAR data onto a digital terrain model and utilizing a modified U-Net for segmenting the building outlines in a common projected map system. An extensive data set consisting of TerraSAR-X images and OpenStreetMap building footprints was created to train the model. With the very promising results, the study positions SAR as a reliable alternative for accurate building footprint mapping, with implications for time-critical disaster management and urban monitoring. «
Investigating the synergy of deep learning and high-resolution Synthetic Aperture Radar (SAR) data, this paper focuses on building footprint extraction -- a domain traditionally dominated by optical imagery. The proposed method involves projecting SAR data onto a digital terrain model and utilizing a modified U-Net for segmenting the building outlines in a common projected map system. An extensive data set consisting of TerraSAR-X images and OpenStreetMap building footprints was created to train... »
ISBN:
979-8-3503-6032-5
Fakultät:
Fakultät für Luft- und Raumfahrttechnik
Institut:
LRT 9 - Institut für Raumfahrttechnik und Weltraumnutzung
Professur:
Schmitt, Michael
Projekt:
SUSO - Scene Understanding by SAR-Optical Data Fusion