Quantifying Damage in Unreinforced Masonry Walls Using an Image-Based Method

Earthquakes can cause significant damage to masonry structures, leading to potential collapse risks and high repair costs. This study aims to investigate the relationship between visual damage features and the damage state of unreinforced masonry walls subjected to earthquakes using a comprehensive database including ۱۱۱ images of different unreinforced masonry walls at different drift ratios. This research focuses on extracting visual damage features, such as crack length and crushed area, using computer vision tools and then assessing their correlation with the peak experienced drift ratio of the masonry walls. The results of the study demonstrate a meaningful correlation between the extracted visual damage features and the experienced drift ratio of the unreinforced masonry walls. This suggests that the proposed method has the potential to be incorporated into automated post-earthquake damage assessment frameworks, which could facilitate informed decision-making processes regarding repair and reconstruction efforts of unreinforced masonry walls following seismic events.

Keywords:

Visual damage , unreinforced masonry wall , Structural health monitoring , computer vision