Rapid Extraction of Disaster-Affected Areas Based on Affected Objects Classification System for Disaster Mitigation

Abstract—Remote sensing images have been a main information source of the pre-disaster monitoring, disaster prediction and rapid assessments of disasters. However, the existing affected objects classification system, which is designed for the post-disaster survey, hardly meets the requirements of applications in disaster mitigation. In this paper, we proposed the construction scheme of affected objects classification system for disaster mitigation, considering of characteristics of data disasters and affected objects. In the framework of this classification scheme, we specified the nomenclature for floods, which is then used in the rapid extraction of the flooded area in Yuyao. And the flooded area is extracted through change detection. Experiments showed that the classification system can be applied in classifying the affected bbjects directly, which is the basis of rapid assessment with remote sensing images. Keywords-remote sensing information extraction; disaster mitigation；affected bbjects ； classification system ； flood; I