Post-earthquake Building Damage Detection Using Deep Learning

Since the creation of Earth, Earthquakes have been one of the most calamitous natural disasters that mankind is being affected with. Accurate evaluation of buildings damaged by the earthquake has been a problem for authorities which makes the evacuation of humans from affected areas tough. This paper discusses the techniques which assess the pre and post earthquake-affected areas and indicates the damage inflicted areas. It uses deep learning techniques to identify the buildings in the images, both pre and post calamitic through which damage inflicted from calamity can be inferred. The training objective and overall training pipeline are defined to achieve accurate detection of buildings.