AUTOMATED LARGE-SCALE DAMAGE DETECTION ON HISTORIC BUILDINGS IN POST-DISASTER AREAS USING IMAGE SEGMENTATION

This research aims to investigate the application of computer vision and machine learning for the automatic detection of wall collapse damage in historic buildings caused by natural and man-made disasters. Given the complexities involved in inspecting damaged buildings, particularly in post-disaster scenarios, this research aims to establish a foundation for creating an automated assessment process. Our findings demonstrate the successful automatic detection of various shapes of wall collapse on damaged buildings from the Beirut explosion of 2020, as well as from other damaged buildings obtained from the internet, thereby highlighting the transferability of our method. This research paves the way for the development of a more robust machine learning model capable of detecting a broader range of damages, which can significantly enhance the efficiency and accuracy of post-disaster assessment of historic structures. The paper presents a novel approach for damage detection and quantification, which underscores the potential of structural health monitoring in improving disaster response and recovery efforts.