Automated Dem Extraction of Urban Areas from Multi Resolution Satellite Images

This paper reports experiments carried out for automated extraction of a digital elevation model (DEM) from satellite images. In particular, this paper focuses on DEM extraction from dense urban areas where there are occlusions, shadow and height discontinuities. These properties have been known to create many problems in stereo matching and for automated DEM extraction. Using satellite images at various resolution, a DEM extraction algorithm was applied and the output DEM quality, on urban areas was studied. We tested SPOT images at 10m resolution, EOC images at 6.6m and IKONOS images at 1m resolution. These images include the same urban area in Daejeon city. Because IKONOS images do not provide the ancillary data, we use Gupta and Hartley’s(1997) DLT model in camera modeling procedure. We carried out accuracy assessment using USGS DTED for SPOT, EOC DEM and 23 check points for IKONOS. It showed that SPOT DEM had about 38m RMS error, EOC DEM 12m RMS error and IKONOS DEM 6.5 m RMS error. IKONOS DEM had more error than SPOT, EOC DEM with respect to image resolution. However with IKONOS images we could make DEMs in complex and dense urban areas. These experiments show that our algorithm can generate urban DEM from 1m resolution and that, however, we need to improve the algorithm to minimize effects of occlusion, building shadows, and height discontinuity on DEMs.