Automatic detection of damaged buildings after the earthquake with current cadastral maps and 3D point cloud produced from images supplied by unmanned aircraft after the earthquake

In the study, it is aimed to automatically detect the damaged or destroyed buildings after the earthquake as a result of using the photogrammetric point cloud produced from the UAV images and the cadastral map of the region. Two different methods based on the principle that damaged or collapsed buildings will change geometrically have been prepared in the MATLAB interface. In the first method produced, the destroyed buildings were determined with the height-based algorithm. However, in some cases, it was observed that the height of the buildings did not change even though they were tilted. In such cases, thanks to the 3D buffer thrown around the roof, the buildings with damaged vertical structure were detected automatically. With the developed methods, an accuracy rate of 96.15% has been achieved in the detection of damaged or collapsed buildings after the earthquake. As a result of processing the data obtained after the earthquake, it is thought that the damaged and collapsed buildings will be detected in less than a day, and the rescue efforts will accelerate. In addition, the fact that it can prevent the loss of life caused by late interventions increases the importance of the proposed study.