LAUSR

Abstract Known as an unpredictable natural disaster, earthquake is one of the most devastating natural disasters that causes significant life losses and damages, every year. After an earthquake, quick and accurate buildings damage identification for rescuing can reduce the number of fatalities. In this regard, Remote Sensing (RS) technology is an efficient tool for rapid monitoring of damaged buildings. This paper proposes a novel method, titled segment-by-segment comparison (SBSC), to generate buildings damage map using multi-temporal satellite images. The proposed method begins by extracting image-objects from pre- and post-earthquake images and equalizing them through segmentation intersection. After the extraction of various textural and spectral descriptors on pre- and post-event images, their differences are used as an input feature vector in a classification algorithm. Also, the Genetic Algorithm (GA) is used to find the optimum descriptors in the classification process. The accuracy of the proposed method was tested on two different datasets from different sensors. Comparing the damage maps obtained from the proposed method with the manually extracted damage map, above 92% of the buildings were correctly labelled in both datasets.