A Study on Building Extraction Within Flood and Landslide Prone Areas Utilizing Spatial Information of Buildings

Abstract

The purpose of this study was to extract buildings located within flood and landslide hazard areas using spatial information of domestic buildings. By identifying the characteristics of buildings in disaster-prone regions, this research aims to contribute to the identification of buildings requiring urgent preparation. This study employed a combination of building ledger data and land characteristic information to generate building spatial data. Building properties were defined based on the number of basement floors and the years that elapsed since the use approval date of the buildings. The resulting building spatial data were then categorized into seven distinct clusters based on their properties. To analyze the differences, ANOVA and posthoc tests were conducted to examine the average variances in flood hours and landslide hazard rates among the seven clusters. Furthermore, this study identified the building properties located within flood and landslide hazard areas. By implementing the proposed method, central and local governments can more efficiently and effectively prevent accidents in buildings during disasters through proactive measures.