Evaluation of the damages caused by seismic events: First tests on supporting traditional multispectral classification with DSM

Seismic damages, as a roof entirely collapsed on the ground, are very difficult to be found using only multispectral classification algorithms. The availability of high resolution stereopairs from satellite disclose new possible fields of application to estimate changes and transformations of areas following catastrophic events. Combining both techniques it is obviously possible only when stereoscopic and multispectral images are available. In this case, as for all monitoring studies, it is necessary to compare the present situation to the pre-seismic one. The pre-seismic situation can be advantageously studied by classic photogrammetric techniques based on aerial frames, that are available in archives managed by photogrammetric companies and local government agencies. But it is also possible to extract the pre-seismic morphology from digital maps, containing the three-dimensional characteristics of the buildings. The present research tries to: a) improve the digital surface model extracted from Ikonos satellite images covering an area of central Italy (Foligno, Umbria), through a pre-treatment of images and a manual editing b) study the best DSM models to improve the detection of height difference, mainly in urban areas, and evaluate the results of the classification of land cover as further data to detect changes in building shape. DSM obtained by three-dimensional maps have been compared with DSM extracted directly from aerial stereo-pairs using different approaches. In the area under study a seismic event happened in September of the '97 causing relevant damages to different urbanized centres of the area.