AI-Based Detection of Micro-Cracks in Concrete Using Drone Imagery

DOI: https://doi.org/10.63345/ijrmeet.org.v10.i11.1

Lucky Jha

 Ghaziabad, Uttar Pradesh 201009

luckyjha200405@gmail.com

Abstract

The integrity of concrete infrastructure is critical to public safety and long‑term asset management. Micro‑cracks, often invisible to the naked eye, can propagate under cyclic loading and environmental exposure, leading to significant structural deterioration. Recent advances in unmanned aerial vehicles (UAVs) equipped with high‑resolution imaging sensors and artificial intelligence (AI) have enabled automated, non‑contact inspection of large concrete surfaces. This manuscript presents an AI‑based framework for detecting micro‑cracks in concrete using drone imagery, employing convolutional neural networks (CNNs) trained on annotated aerial datasets collected in 2022. The methodology integrates image preprocessing, data augmentation, and model optimization to achieve high detection accuracy while minimizing false positives. Statistical analysis of model performance—including precision, recall, F1‑score, and processing time—demonstrates the feasibility of deploying the system for field inspections. A comparative evaluation against traditional edge‑detection techniques highlights a 28% improvement in detection rate. The results suggest that the proposed approach can significantly enhance the efficiency and reliability of concrete condition assessments. Keywords: drone imagery, micro‑crack detection, convolutional neural networks, concrete inspection, UAV-based monitoring.

Keywords

Drone imagery; micro‑crack detection; convolutional neural network; concrete inspection; UAV monitoring

References

• https://www.google.co.in/url?sa=i&url=https%3A%2F%2Fzilliz.com%2Fglossary%2Fconvolutional-neural-network&psig=AOvVaw2rY8-hCoCYo3Ct1SmzJej_&ust=1745216805235000&source=images&cd=vfe&opi=89978449&ved=0CBQQjRxqFwoTCOC147P95YwDFQAAAAAdAAAAABAE
• Cha, Y.-J., Choi, W., & Büyüköztürk, O. (2017). Deep learning‐based crack damage detection using convolutional neural networks. Computer‐Aided Civil and Infrastructure Engineering, 32(5), 361–378.
• Guo, P., Georgakis, C., & Collins, C. (2021). UAV‐based thermal and RGB image fusion for non‐destructive evaluation of concrete bridges. Automation in Construction, 130, 103852.
• He, K., Zhang, X., Ren, S., & Sun, J. (2016). Deep residual learning for image recognition. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (pp. 770–778).
• Khan, M. A., & Niazi, A. U. (2022). Drone‐based inspection and analysis of concrete dams using AI segmentation. Journal of Infrastructure Systems, 28(2), 04021065.
• Li, Z., & Sun, H. (2020). Data augmentation for deep learning‐based defect detection in civil engineering. Journal of Computing in Civil Engineering, 34(3), 04020014.
• Madhusudhana, C., & Kumar, V. (2018). Automatic detection of cracks on pavement surfaces using Canny edge detection and morphological operations. Journal of Transportation Engineering, 144(12), 05018013.
• Machado, J. C., de Andrade, M. A., & Ferrer, S. (2021). Deep learning‐based approach for early detection of micro‐cracks in concrete. Structural Health Monitoring, 20(2), 905–918.
• Nguyen, L., Le, T., & Cheng, H. (2021). Transfer learning for crack detection in concrete structures. Automation in Construction, 125, 103594.
• Oliveira, H., & Correia, P. L. (2014). Automatic road crack detection and characterization. IEEE Transactions on Intelligent Transportation Systems, 14(1), 155–168.
• Park, J., Cho, H., & Lee, S. (2019). Real‐time crack detection in concrete surfaces using edge computing on UAV platforms. Journal of Field Robotics, 36(5), 972–989.
• Peng, X., & Zhang, H. (2021). Fusion of multisensor data for enhanced crack detection in concrete structures. IEEE Sensors Journal, 21(4), 4567–4578.
• Rathore, S., & Ahmed, N. (2022). Lightweight CNN architectures for on‐board UAV crack detection. Aerospace, 9(5), 210.
• Ronneberger, O., Fischer, P., & Brox, T. (2015). U‐Net: Convolutional networks for biomedical image segmentation. In Medical Image Computing and Computer‐Assisted Intervention (pp. 234–241). Springer.
• Shi, H., Ma, J., & Jiang, M. (2020). Enhanced crack detection in 3D point clouds of concrete surfaces. Remote Sensing, 12(11), 1773.
• Song, Q., & Adeli, H. (2019). Novel image processing and AI techniques for concrete surface crack detection. Construction and Building Materials, 216, 223–235.
• Tang, P., Huang, R., & Shen, Y. (2021). Semantic segmentation for pavement distress detection using UAV imagery. Remote Sensing Letters, 12(6), 511–520.
• Wang, S., & Xu, C. (2018). GooseNet: A lightweight network for micro‐crack detection. IEEE Access, 6, 62425–62435.
• Xu, Y., & Li, B. (2022). A review of UAV applications in structural health monitoring of concrete structures. Journal of Civil Structural Health Monitoring, 12(3), 477–496.
• Zhang, L., Yang, F., Zhang, Y., & Zhu, Y. (2019). Road crack detection based on deep convolutional neural network. IEEE Transactions on Intelligent Transportation Systems, 20(9), 3337–3349.
• Zhou, H., & Liu, X. (2020). Multi‐scale feature fusion network for micro‐crack detection. IEEE Transactions on Industrial Electronics, 67(11), 10295–10304.