AUTOMATIC DETECTION AND CLASSIFICATION OF DAMAGED BUILDINGS , USING HIGH RESOLUTION SATELLITE IMAGERY AND VECTOR DATA

Receiving rapid, accurate and comprehensive knowledge about the conditions of damaged buildings after earthquake strike and other natural hazards is the basis of many related activities such as rescue, relief and reconstruction. Recently, commercial high-resolution satellite imagery such as IKONOS and QuickBird is becoming more powerful data resource for disaster management. In this paper, a method for automatic detection and classification of damaged buildings using integration of high-resolution satellite imageries and vector map is proposed. In this method, after extracting buildings position from vector map, they are located in the pre-event and post-event satellite images. By measuring and comparing different textural features for extracted buildings in both images, buildings conditions are evaluated through a Fuzzy Inference System. Overall classification accuracy of 74% and kappa coefficient of 0.63 were acquired. Results of the proposed method, indicates the capability of this method for automatic determination of damaged buildings from high-resolution satellite imageries.