DEM Extraction in Urban Areas Using High-Resolution TerraSAR-X Imagery |
| |
Authors: | Umut Gunes Sefercik Naci Yastikli Iulia Dana |
| |
Affiliation: | 1. Department of Geomatics Engineering, Bulent Ecevit University, 67100, Zonguldak, Turkey 2. Department of Geomatics Engineering, Yildiz Technical University, Davutpasa Campus, 34210, Istanbul, Turkey 3. Romanian Space Agency, Bucharest, Romania
|
| |
Abstract: | Three-dimensional (3D) spatial information is crucial for improving the quality of human life through urban planning and management, and it is widely utilized due to its rapid, periodic and inexpensive acquisition. In this context, extraction of digital surface and elevation models (DSM and DEM) is a significant research topic for space-borne optical and synthetic aperture radar (SAR) remote sensing. The DSMs include visible features on the earth’s surface such as vegetation, forest and elevated man-made objects, while DEMs contain only the bare ground. In this paper, using TerraSAR-X (TSX) high resolution Spotlight (HS) images, high-resolution interferometric DEM generation in a part of Istanbul urban area is aimed. This is not an easy task because of SAR imaging problems in complex geometry of urban settlements. The interferometric processing steps for DSM generation were discussed including critical parameters and thresholds to improve the quality of the final product and a 3 m gridded DSM was generated. The DSM-DEM conversion was performed by filtering and the quality of generated DEM was verified against a reference DEM from stereo photogrammetry with 3 m original grid spacing. The achieved root mean square error of height differences (RMSZ) varies from 7.09 to 8.11 m, depending on the terrain slope. The differential DEM, illustrates the height differences between generated DEM and the reference DEM, was generated to show the correlation between height differences and the coherence map. Finally, a perspective view of test area was created draping extracted DEM and a high-resolution IKONOS panchromatic image. |
| |
Keywords: | |
本文献已被 SpringerLink 等数据库收录! |
|