Mona Peyk herfeh - Islamic Azad University Oloum TahghighatQazvin,
Asadollah Shahbahrami - University of Guilan Faculty of engineering Rasht,
Farshad Parhizkar Miandehi - Islamic Azad University Electronic and Computer Faculty Zanjan,

when an earthquake happens, the image-basedtechniques are influential tools for detection and classification ofdamaged buildings. Obtaining precise and exhaustiveinformation about the condition and state of damaged buildingsafter an earthquake is basis of disaster management. Today’susing satellite imageries such Quickbird is becoming moresignificant data for disaster management. In this paper, a methodfor detecting and classifying of damaged buildings using satelliteimageries and digital map is proposed. In this method afterextracting buildings position from digital map, they are located inthe pre-event and post-event images of Bam earthquake. Aftergenerating features, genetic algorithm applied for obtainingoptimal features. For classification, Adaptive boosting is used andcompared with neural networks. Experimental results show thattotal accuracy of adaptive boosting for detecting and classifyingof collapsed buildings is about 84 percent

Earthquake; Collapese detection; Classification Adaptive boosting