View in EDS

Quantifying Crack Damage in BFRP-Reinforced Concrete Beams with YOLOv8 and 3D-DIC.

Saved in:
Bibliographic Details
Title: Quantifying Crack Damage in BFRP-Reinforced Concrete Beams with YOLOv8 and 3D-DIC.
Authors: Zeng, Yunqi1 (AUTHOR), Lei, Dong1,2 (AUTHOR) leidong@hhu.edu.cn, Zhou, Kaiyang1,3 (AUTHOR), He, Jintao1,4 (AUTHOR), She, Zesheng1 (AUTHOR), Yu, Yang5 (AUTHOR), Liu, Ling6 (AUTHOR), Yu, Kexin6 (AUTHOR)
Source: Journal of Nondestructive Evaluation. Dec2025, Vol. 44 Issue 4, p1-24. 24p.
Subject Terms: *STRUCTURAL health monitoring, *DIGITAL image correlation, *MATERIAL fatigue, *REINFORCED cement, *IMAGE processing, *OBJECT recognition (Computer vision), *FRACTURE strength
Abstract: This study presents an novel structural health monitoring (SHM) approach by integrating Digital Image Correlation (DIC) with the YOLOv8 instance segmentation model to quantify crack damage evolution in concrete beams subjected to different preloading conditions. Four-point bending tests were conducted on plain concrete, BFRP-reinforced concrete, and preloaded BFRP-reinforced concrete beams. Our method leverages the model's pixel-level segmentation capabilities to provide a more granular and continuous tracking of damage progression. A novel Weighted Damage Index (WDI) was developed to quantify the extent and progression of cracking based on the spatial and probabilistic features extracted by the model. The WDI demonstrated a clear correlation with mechanical degradation and effectively characterized three distinct stages of damage: elastic, stable, and unstable. As an interpretable and scalable visual damage metric, WDI shows strong potential for computer-assisted or semi-automated SHM applications, offering a cost-efficient tool to support early warning, maintenance prioritization, and reinforcement strategy optimization. These findings provide a new perspective on integrating vision-based techniques into intelligent infrastructure monitoring. [ABSTRACT FROM AUTHOR]
Database: Academic Search Index
Description
Abstract:This study presents an novel structural health monitoring (SHM) approach by integrating Digital Image Correlation (DIC) with the YOLOv8 instance segmentation model to quantify crack damage evolution in concrete beams subjected to different preloading conditions. Four-point bending tests were conducted on plain concrete, BFRP-reinforced concrete, and preloaded BFRP-reinforced concrete beams. Our method leverages the model's pixel-level segmentation capabilities to provide a more granular and continuous tracking of damage progression. A novel Weighted Damage Index (WDI) was developed to quantify the extent and progression of cracking based on the spatial and probabilistic features extracted by the model. The WDI demonstrated a clear correlation with mechanical degradation and effectively characterized three distinct stages of damage: elastic, stable, and unstable. As an interpretable and scalable visual damage metric, WDI shows strong potential for computer-assisted or semi-automated SHM applications, offering a cost-efficient tool to support early warning, maintenance prioritization, and reinforcement strategy optimization. These findings provide a new perspective on integrating vision-based techniques into intelligent infrastructure monitoring. [ABSTRACT FROM AUTHOR]
ISSN:01959298
DOI:10.1007/s10921-025-01277-8