Computer Vision -Aided Damage Evaluation of Reinforced Concrete Columns

This study proposes an automated damage assessment method that utilizes visual damage features of Reinforced Concrete (RC) columns. In this research, a comprehensive database comprising ۱۰۰ images of damaged RC columns tested under cyclic quasi-static loading is collected at different drift ratios. Then, image processing techniques are used to extract visual damage features from the database, including crack length and crushed area. These visual damage features are then used to derive a predictive equation for the maximum-experienced drift ratio of the specimens through symbolic regression analysis. The results of this study demonstrate a strong correlation between the damage state of the RC columns and the extracted visual damage features, with a meaningful correlation coefficient of ۰.۸۴. This finding suggests that the proposed method has great potential for quickly and accurately assessing the damage to RC columns, reducing the need for manual inspections and improving the safety and efficiency of the overall damage assessment process. By leveraging visual damage features, the method can contribute to more informed decision-making regarding repair and rehabilitation efforts, enhancing the overall response and recovery strategies in the aftermath of seismic events.

کلیدواژه ها:

Reinforced concrete columns ، Surface damage ، Predictive equation ، Symbolic regression