Application of LiDAR Point-Cloud for Damage Map Generation in Urban Area

Rapid detection of damaged buildings after an earthquake has been a critical issue indisaster management. Many data sources such as LiDAR data and aerial/satellite imagescan be used for obtaining information from the affected area for mapping destruction. LIDAR, which stands for Light Detection and Ranging, is a remote sensing method that usespulsed laser to measure precise, three-dimensional information about the shape of the Earthand its surface characteristics. This three-dimensional information is usually a dense, detailrich group of elevation points, called a point cloud. The height variations of a building can be an important parameter in the diagnosis of destroyed buildings. In this case, regular roof with small changes in height points indicatesthat building is not destroyed. Otherwise, non-homogeneity in height of the points proofs the presence of debris and demolition of the building. In this study, a new method for detection of damaged building in an urban area using post-event LiDAR point-cloud is proposed. Inthe proposed method, first, post-event LiDAR point-cloud and a pre-event vector map in a pre-processing step are registered. Then, corresponding points of a buildings are extractedfrom LiDAR data based on registration information. Finally, based on the height variation of the extracted points the candidate building is labeled as intact or damaged building. The proposed method was evaluated using LiDAR data of Port au. Prince, Haiti, which washeavily hit by an earthquake in January 2010. The obtained results proved the high ability of this method for generation of building damage map in an urban area after an earthquake.

کلیدواژه ها:

Disaster Management ، Damage Map ، LiDAR Point Cloud