Automatic building extraction from high resolution satellite images for map updating: A model based approach

An approach was developed for automatically updating the buildings of an existing vector database from high resolution satellite images using spectral image classification, Digital Elevation Models (DEM) and the model-based extraction techniques. First, the areas that contain buildings are detected using spectral image classification and the normalized Digital Surface Model (nDSM). The classified output provides the shapes and the approximate locations of the buildings. However, those buildings that have similar reflectance values with the other classes were not able to be detected. Therefore, nDSM was generated by subtracting the Digital Terrain Model (DTM) from the Digital Surface Model (DSM). Next, the buildings were differentiated from the trees by using the Normalized Difference Vegetation Index (NDVI). Areas other than the buildings are excluded from further processing. The buildings that exist in the vector database but missing in the image were detected through analyzing the results of the classification and nDSM. Finally, the buildings constructed after the date of the compilation of the existing vector database were extracted through the proposed model-based approach and the vector database was updated with the new building boundaries. The method was implemented in a selected urban area in Ankara, Turkey using the IKONOS pan-sharpened and panchromatic images. The results show that the proposed approach is quite satisfactory for detecting and delineating the buildings from high resolution space images.