Automatic Crack Segmentation in Asphalt Pavement Using U-Net Convolutional Neural Networks
DOI:
https://doi.org/10.62544/ucomscientia.v3i2.59Keywords:
Redes neuronales convolucionales, visión artificial, pavimento asfáltico, segmentación de imágenes, mantenimiento vialAbstract
This article presents the development of an automatic crack segmentation system for asphalt pavement using convolutional neural networks, specifically the U-Net architecture. A proprietary dataset was built with 847 images captured in the city of Coronel Oviedo, Paraguay, of which 505 contained cracks. The images were manually labeled and processed at a resolution of 256x256 pixels. The model was trained for 500 epochs using the binary crossentropy loss function and the Adam optimizer. Outstanding performance metrics were obtained, with an F1 Score of 0.9956 and an Intersection over Union (IoU) index of 0.9913. Additionally, post-segmentation processing was integrated for crack width quantification. In summary, these results demonstrate the high precision and robustness of the developed system.
References
Abadi, M., Agarwal, A., Barham, P., Brevdo, E., Chen, Z., Citro, C., Corrado, G. S., Davis, A., Dean, J., Devin, M., Ghemawat, S., Goodfellow, I., Harp, A., Irving, G., Isard, M., Jia, Y., Jozefowicz, R., Kaiser, L., Kudlur, M., ... Zheng, X. (2016).Tensorflow: Large-scale machine learning on heterogeneous distributed systems. arXiv. https://doi.org/10.48550/ARXIV.1603.04467
Ali, L., AlJassmi, H., Swavaf, M., Khan, W. y Alnajjar, F. (2024). Rs-net: Arquitectura de red U-Net Sharp residual para la segmentación y evaluación de la severidad de grietas en pavimentos. Journal of Big Data,11(1), 116. https://doi.org/10.1186/s40537-024-00981-y
Alzamora, P. (2023). Segmentación semántica de imágenes con Deep Learning. Data Machine Learning Visualization. https://blog.damavis.com/segmentacion-semantica-de-imagenes-con-deep-learning
Astute Analytica. (2024). Road Maintenance Market to Drive Fast to Reach Valuation of USD 23.39 Billion By 2032. GlobeNewsWire: https://www.globenewswire.com/news-release/2024/05/07/2876916/0/en/Road-Maintenance-Market-to-Drive-Fast-to-Reach-Valuation-of-USD-23-39-Billion-By-2032-Astute-Analytica.html
Bradski, G. (2000). The OpenCV Library. Dr. Dobb's Journal of Software Tools. https://drdobbs.com/open-source/the-opencv-library/184404319
Cao, H., Wang, Y., Chen, J., Jiang, D., Zhang, X., Tian, Q. y Wang, M. (2023). Swin-unet: Transformador puro tipo Unet para la segmentación de imágenes médicas. En L. Karlinsky, T. Michaeli y K. Nishino (Eds.),Visión por Computador –Talleres ECCV 2022(pp. 205-218). Springer Nature Suiza. https://doi.org/10.1007/978-3-031-25066-8_9
Chen, L.-C., Zhu, Y., Papandreou, G., Schroff, F. y Adam, H. (2018). Codificador-decodificador con convolución separable de alto rendimiento para la segmentación semántica de imágenes. En V. Ferrari, M. Hebert, C. Sminchisescu y Y. Weiss (Eds.),Visión por Computador –ECCV 2018(pp. 833-851). Springer International Publishing. https://doi.org/10.1007/978-3-030-01234-2_49
Chen, S., Feng, Z., Xiao, G., Chen, X., Gao, C., Zhao, M. y Yu, H. (2024). Detección de grietas en pavimento basada en el modelo swin-unet mejorado.Buildings,14(5), 1442. https://doi.org/10.3390/buildings14051442
Clark, J. (2024). Pillow (PIL Fork) Documentation: Version 10.4.0. https://pillow.readthedocs.io/en/stable/
Cruz, G., Riobó, A., Pfeifer, M. y Duarte, D. (2024). IA desde los cimientos: Desafíos y oportunidades en el contexto de América Latina y el Caribe. Banco Interamericano de Desarrollo. https://doi.org/10.18235/0013275DataScientest. (2021). Convolutional Neural Network: definición y funcionamiento. https://datascientest.com/es/convolutional-neural-network-es
Di Benedetto, A., Fiani, M. y Gujski, L. M. (2023). Arquitectura de CNN basada en U-net para la segmentación de grietas en carreteras.Infraestructuras,8(5), 90. https://doi.org/10.3390/infrastructures8050090
Dong, C., Li, L., Yan, J., Zhang, Z., Pan, H. y Catbas, F. N. (2021). Segmentación de grietas por fatiga a nivel de píxel en imágenes a gran escala de estructuras de acero mediante una red codificador-decodificador.Sensors,21(12), 4135. https://doi.org/10.3390/s21124135
Federal Highway Administration. (2014). Distress Identification Manual for the Long-Term Pavement Performance Program. (5th ed.). U.S. Department of Transportation. https://www.fhwa.dot.gov/publications/research/infrastructure/pavements/ltpp/13092/13092.pdf
Geeksforgeeks. (2025). Image Edge Detection Operators in Digital Image Processing. https://www.geeksforgeeks.org/image-edge-detection-operators-in-digital-image-processing/
Global Infrastructure Hub. (2022). AI and deep learning for identifying pavement failures in Latin American and the Caribbean. World Bank's Public-Private Infrastructure Advisory Facility (PPIAF). https://infratech.gihub.org/infratech-case-studies/ai-and-deep-learning-for-identifying-pavement-failures-in-latin-american-and-the-caribbean/
Goodfellow, I., Bengio, Y., & Courville, A. (2016). Deep Learning. http://www.deeplearningbook.org
Guevara, T. (2025). La inteligencia artificial avanza desigual en Latinoamérica, buscan llevarla al sector productivo. Voz de América (VoA). https://www.vozdeamerica.com/a/ia-avanza-desigual-latinoamerica/8010993.html
Harris, C. R., Millman, K. J., van der Walt, S. J., Gommers, R., Virtanen, P., Cournapeau, D., Wieser, E., Taylor, J., Berg, S., Smith, Nueva Jersey, Kern, R., Picus, M., Hoyer, S., van Kerkwijk, M. H., Brett, M., Haldane, A., del Río, J. F., Wiebe, M., Peterson, P.,... Oliphant, T. E. (2020). Programación de matrices con NumPy.Naturaleza,585(7825), 357-362. https://doi.org/10.1038/s41586-020-2649-2
He, K., Zhang, X., Ren, S., & Sun, J. (2016). Deep residual learning for image recognition.2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 770-778. https://doi.org/10.1109/CVPR.2016.90
Kaveh, H., & Alhajj, R. (2024). Recent advances in crack detection technologies for structures: A survey of 2022-2023 literature. Frontiers in Built Environment,10, 1321634. https://doi.org/10.3389/fbuil.2024.1321634
IBM. (2024). ¿Qué es la segmentación de imágenes?. https://www.ibm.com/es-es/topics/image-segmentation
Kingma, D. P., & Ba, J. (2014).Adam: A method for stochastic optimization. arXiv. https://doi.org/10.48550/ARXIV.1412.6980
Lau, S. L. H., Chong, E. K. P., Yang, X., & Wang, X. (2020). Automated pavement crack segmentation using u-net-based convolutional neural network.IEEE Access,8, 114892-114899. https://doi.org/10.1109/ACCESS.2020.3003638
Long, J., Shelhamer, E., & Darrell, T. (2015). Fully convolutional networks for semantic segmentation.2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 3431-3440. https://doi.org/10.1109/CVPR.2015.7298965
Madroñero Urcuqui, J. D., y Valencia López, Y. C. (2019). Metodología para la identificación automática del deterioro en pavimento flexible, por medio de fotografías aéreas tomadas desde vehículos no tripulados. Universidad del Valle. https://hdl.handle.net/10893/15476
Marín-Acevedo, E. A. (2019). Detección automática de grietas y fisuras en pavimento por medio de fotos y redes neuronales convolucionales. Universidad Católica de Oriente. https://hdl.handle.net/20.500.12516/419
Ministerio de Obras Públicas y Comunicaciones. (2019). Manual de Carreteras del Paraguay.https://apcarreteras.org.py/manual-de-carreteras-del-paraguay-rev-2019/
Nyathi, M. A., Bai, J., & Wilson, I. D. (2024). Deep learning for concrete crack detection and measurement.Metrology,4(1), 66-81. https://doi.org/10.3390/metrology4010005
Peng, Y., Chen, D. Z., & Sonka, M. (2025). U-net v2: Rethinking the skip connections of u-net for medical image segmentation.2025 IEEE 22nd International Symposium on Biomedical Imaging (ISBI), 1-5. https://doi.org/10.1109/ISBI60581.2025.10980742
Pfeifer, M., Fernandez, S., y Riobo, A. (2023). Pavimenta2: infraestructura digital al servicio de los activos viales.(BID) moviliblog. https://blogs.iadb.org/transporte/es/pavimenta2-infraestructura-digital-al-servicio-de-los-activos-viales/Riobo, A., Pfeifer, M., y Calle Jordá, A. (2024). VíaSegura: lecciones aprendidas e inteligencia artificial al servicio de la seguridad vial.(BID) Moviliblog. https://blogs.iadb.org/transporte/es/viasegura-lecciones-aprendidas-e-inteligencia-artificial-al-servicio-de-la-seguridad-vial
Ronneberger, O., Fischer, P., & Brox, T. (2015). U-net: Convolutional networks for biomedical image segmentation. En N. Navab, J. Hornegger, W. M. Wells, & A. F. Frangi (Eds.),Medical Image Computing and Computer-Assisted Intervention –MICCAI 2015(Vol. 9351, pp. 234-241). Springer International Publishing. https://doi.org/10.1007/978-3-319-24574-4_28
Sabouri, M., & Sepidbar, A. (2023). SUT-Crack. https://doi.org/10.17632/gsbmknrhkv.6
Soto Olguín, L. y Ramírez Villanueva, F. (2024). Segmentación de grietas superficiales en pavimento asfáltico utilizando técnicas de visión artificial. [Tesis de grado]. Universidad Nacional de Caaguazú. https://publicaciones.fctunca.edu.py/handle/123456789/128
Srivastava, N., Hinton, G., Krizhevsky, A., Sutskever, I., & Salakhutdinov, R. (2014). Dropout: A Simple Way to Prevent Neural Networks from Overfitting. Journal of Machine Learning Research, 15(1), 1929–1958. https://dl.acm.org/doi/abs/10.5555/2627435.2670313
Szeliski, R. (2022).Computer vision: Algorithms and applications. Springer International Publishing. https://doi.org/10.1007/978-3-030-34372-9
Tan, M., & Le, Q. (2019). EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks.Proceedings of the 36th International Conference on Machine Learning. https://proceedings.mlr.press/v97/tan19a.html
Tang, Y., Qian, Y., & Yang, E. (2022). Weakly supervised convolutional neural network for pavement crack segmentation.Intelligent Transportation Infrastructure,1, liac013. https://doi.org/10.1093/iti/liac013
Tello Cifuentes, L., Marulanda, J., y Thomson, P. (2021). detección de grietas en el pavimento usando técnicas de procesamiento de imágenes y redes neuronales artificiales. Encuentro Internacional De Educación En Ingeniería. http://dx.doi.org/10.26507/ponencia.1565
Yu, Y., Xia, W., Zhao, Z., & He, B. (2024). A lightweight and high-accuracy model for pavement crack segmentation. Applied Sciences,14(24), 11632. https://doi.org/10.3390/app142411632
Zhang, Z., He, Y., Hu, D., Jin, Q., Zhou, M., Liu, Z., Chen, H., Wang, H., & Xiang, X. (2025). Algorithm for pixel-level concrete pavement crack segmentation based on an improved U-Net model.Scientific Reports,15(1), 6553. https://doi.org/10.1038/s41598-025-91352-x
Zhang, Y., & Zhang , L. (2024). Detection of Pavement Cracks by Deep Learning Models of Transformer and UNet. IEEE Transactions on Intelligent Transportation Systems, 25(11). https://doi.org/10.1109/TITS.2024.3420763
Zhang, Q., Chen , S., Wu , Y., Ji , Z., Yan , F., Huang , S., & Liu , Y. (2024a). Improved U-net network asphalt pavement crack detection method. PLoS ONE, 19(5). https://doi.org/10.1371/journal.pone.0300679
Zhang, J., Sun, S., Song, W., Li, Y., & Teng, Q. (2024b). A novel convolutional neural network for enhancing the continuity of pavement crack detection.Scientific Reports,14(1), 30376. https://doi.org/10.1038/s41598-024-81119-1Zhang, J., Xia, H., Li, P., Zhang, K., Hong, W., & Guo, R. (2024c). A pavement crack detection method via deep learning and a binocular-vision-based unmanned aerial vehicle.Applied Sciences,14(5), 1778. https://doi.org/10.3390/app14051778
Zhao, H., Shi, J., Qi, X., Wang, X., & Jia, J. (2017). Pyramid Scene Parsing Network. IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Honolulu, USA. https://doi.org/https://doi.org/10.1109/CVPR.2017.660
Zhou, Z., Rahman Siddiquee, M. M., Tajbakhsh, N., & Liang, J. (2018). Unet++: A nested u-net architecture for medical image segmentation. En D. Stoyanov, Z. Taylor, G. Carneiro, T. Syeda-Mahmood, A. Martel, L. Maier-Hein, J. M. R. S. Tavares, A. Bradley, J. P. Papa, V. Belagiannis, J. C. Nascimento, Z. Lu, S. Conjeti, M. Moradi, H. Greenspan, & A. Madabhushi (Eds.),Deep Learning in Medical Image Analysis and Multimodal Learning for Clinical Decision Support(Vol. 11045, pp. 3-11). Springer International Publishing. https://doi.org/10.1007/978-3-030-00889-5
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2025 Luz Rashell Soto Olguín, Fredy Gabriel Ramírez Villanueva, Héctor Ramiro Estigarribia Barreto
This work is licensed under a Creative Commons Attribution 4.0 International License.
La Revista Científica UCOM Scientia se distribuye bajo una Licencia Atribución 4.0 Internacional (CC BY 4.0) https://creativecommons.org/licenses/by/4.0/deed.es
