REA Press Null REA Press3042-02023042-0202 REA Press https://doi.org/10.48314/ijrceai.v3i2.58 Research Article Deep learning, Convolutional neural networks, U-net, Crack detection, Post-earthquake damage assessment, Reinforced concrete structures, Structural health monitoring, Image segmentation Automated Post-Earthquake Damage Assessment in Reinforced Concrete Structures Using U-Net-Based Convolutional Neural NetworksAutomated Post-Earthquake Damage Assessment in Reinforced Concrete Structures Using U-Net-Based Convolutional Neural Networks Mirabedini Shirin Department of Computer Engineering, Payame Noor University, PO Box 19395-3697 Tehran, I.R of Iran. Nourani Seyedeh Fatemeh Department of Computer Engineering, Payame Noor University, PO Box 19395-3697 Tehran, I.R of Iran. 06 2026 20 06 2026 3 2 © 2026 REA Press 2026

This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

 Automated Post-Earthquake Damage Assessment in Reinforced Concrete Structures Using U-Net-Based Convolutional Neural Networks

Rapid and accurate assessment of damage to structures following an earthquake is critical for disaster management, emergency response, and rehabilitation planning. Traditional visual inspection methods are time-consuming, costly, and inherently subjective. This research presents an automated approach based on deep learning for identifying and quantifying cracks in Reinforced Concrete (RC) structures after seismic events. The proposed model utilizes a U-Net architecture, a type of Convolutional Neural Network (CNN), for pixel-level crack segmentation. It is trained and evaluated on a dataset of real-world images from the recent Kahramanmaraş earthquake (Türkiye, 2023). Results demonstrate that the model effectively identifies various crack patterns with high accuracy, achieving a strong Intersection over Union (IoU) score of 0.745. This approach offers an efficient, objective, and scalable tool for structural engineers to assess integrity and prioritize repairs. The integration of an Efficient Channel Attention (ECA) mechanism further enhances feature extraction, leading to improved segmentation performance, particularly for subtle and narrow cracks.

Null