Population estimation in urban areas based on 'mixed/cross' stereo models

In the project X3D4Pop we investigate if "mixed" satellite image pairs acquired on two different dates can be used to calculate meaningful building heights in urban areas. Based on These results, the dependency of urban population estimation models on the availability and quality of 3D data is tested. Study areas are Port-au-Prince and Salzburg. The 3D-models are validated against LiDAR-derived elevation models, population numbers are compared with rastered population data from Statistik Austria. The project explores first steps towards an urban population estimation Service by remote sensing, for the humanitarian community.