A computer vision application for automatic detection of defects in urban asphalt pavement in Brazil
Resumo
This work aims to develop an automated system for detecting damage in asphalt pavements by leveraging computer vision and deep learning techniques. It addresses key challenges such as the accurate identification and classification of various types of pavement distress. As a central contribution, we present a publicly available dataset of high-resolution digital images of urban pavements, structured in accordance with Brazilian standards and accompanied by per-image annotations curated by a technical committee. Based on this dataset, the study also presents initial experiments that implement a methodology for the automated diagnosis of road conditions. The model trained with YOLOv5 achieved an mAP@50 of 84.4%. The proposed approach is intended to support public administration in decision-making processes related to road maintenance and intervention planning, ultimately contributing to improvements in the quality, efficiency, and safety of urban road infrastructure.
Palavras-chave:
asphalt pavement defect, computer vision, deep learning, YOLO
Referências
DNIT, “Levantamento visual contínuo para avaliação da superfície de pavimentos flexíveis e semi-rígidos procedimento,” 2003. [Online]. Available: [link]
——, “Defeitos nos pavimentos flexíveis e semi-rígidos terminologia,” 2003. [Online]. Available: [link]
L. Fan, D. Wang, J. Wang, Y. Li, Y. Cao, Y. Liu, X. Chen, and Y. Wang, “Pavement defect detection with deep learning: A comprehensive survey,” IEEE Transactions on Intelligent Vehicles, vol. 9, no. 3, pp. 4292–4311, 2024.
Jo, Ryu, “Pothole detection system using a blackbox camera,” Sensors, 2015. [Online]. Available: [link]
A. Mohan and S. Poobal, “Crack detection using image processing: A critical review and analysis,” Alexandria Engineering Journal, vol. 57, no. 2, pp. 787–798, 2018.
L. Zhang, F. Yang, Y. Daniel Zhang, and Y. J. Zhu, “Road crack detection using deep convolutional neural network,” in 2016 IEEE International Conference on Image Processing (ICIP), 2016, pp. 3708–3712.
H. Maeda, Y. Sekimoto, and T. Seto, “Lightweight road manager: smartphone-based automatic determination of road damage status by deep neural network,” in Proceedings of the 5th ACM SIGSPATIAL International Workshop on Mobile Geographic Information Systems, ser. MobiGIS ’16. New York, NY, USA: Association for Computing Machinery, 2016, p. 37–45. [Online]. DOI: 10.1145/3004725.3004729
W. Yang, X. Zhang et al., “Enhanced crack detection using improved convolutional neural network with attention modules,” Journal of Infrastructure Systems, vol. 26, no. 3, p. 04020025, 2020. [Online]. DOI: 10.1061/(ASCE)IS.1943-555X.0000578
Q. Li, D. Zhang, Y. Chen, M. Cao, L. He, and B. Zhang, “An efficient and reliable coarse-to-fine approach for asphalt pavement crack detection,” Image and Vision Computing, vol. 57, pp. 130–146, 2022.
P. Jiang, D. Ergu, F. Liu, Y. Cai, and B. Ma, “A review of yolo algorithm developments,” Procedia Computer Science, vol. 199, pp. 1066–1073, 2022.
K. Guo, C. He, M. Yang, and S. Wang, “A pavement distresses identification method optimized for yolov5s,” Scientific Reports, vol. 12, p. 3542, 2022.
H. Li, H. Zhang, H. Zhu, K. Gao, H. Liang, and J. Yang, “Automatic crack detection on concrete and asphalt surfaces using semantic segmentation network with hierarchical transformer,” Engineering Structures, vol. 307, p. 117903, 2024.
M. A. Borges and F. A. Spanhol, “BR-AsPavDam – brazilian asphalt pavement damage dataset,” 2025. [Online]. DOI: 10.5281/zenodo.16291115
Ultralytics, “Ultralytics YOLOv5: Real-time object detection,” [link], 2023, commit f4ff536, accessed 25 Jul 2025.
J. Redmon, S. Divvala, R. Girshick, and A. Farhadi, “You Only Look Once: Unified, real-time object detection,” in Proc. 30th IEEE Conf. on Computer Vision and Pattern Recognition, 2016, pp. 779–788.
D. Arya, H. Maeda, S. K. Ghosh, D. Toshniwal, H. Omata, T. Kashiyama, A. Mraz, and Y. Sekimoto, “RDD2020: An image dataset for smartphone-based road damage detection and classification,” 2021. [Online]. DOI: 10.17632/5ty2wb6gvg.1
——, “Defeitos nos pavimentos flexíveis e semi-rígidos terminologia,” 2003. [Online]. Available: [link]
L. Fan, D. Wang, J. Wang, Y. Li, Y. Cao, Y. Liu, X. Chen, and Y. Wang, “Pavement defect detection with deep learning: A comprehensive survey,” IEEE Transactions on Intelligent Vehicles, vol. 9, no. 3, pp. 4292–4311, 2024.
Jo, Ryu, “Pothole detection system using a blackbox camera,” Sensors, 2015. [Online]. Available: [link]
A. Mohan and S. Poobal, “Crack detection using image processing: A critical review and analysis,” Alexandria Engineering Journal, vol. 57, no. 2, pp. 787–798, 2018.
L. Zhang, F. Yang, Y. Daniel Zhang, and Y. J. Zhu, “Road crack detection using deep convolutional neural network,” in 2016 IEEE International Conference on Image Processing (ICIP), 2016, pp. 3708–3712.
H. Maeda, Y. Sekimoto, and T. Seto, “Lightweight road manager: smartphone-based automatic determination of road damage status by deep neural network,” in Proceedings of the 5th ACM SIGSPATIAL International Workshop on Mobile Geographic Information Systems, ser. MobiGIS ’16. New York, NY, USA: Association for Computing Machinery, 2016, p. 37–45. [Online]. DOI: 10.1145/3004725.3004729
W. Yang, X. Zhang et al., “Enhanced crack detection using improved convolutional neural network with attention modules,” Journal of Infrastructure Systems, vol. 26, no. 3, p. 04020025, 2020. [Online]. DOI: 10.1061/(ASCE)IS.1943-555X.0000578
Q. Li, D. Zhang, Y. Chen, M. Cao, L. He, and B. Zhang, “An efficient and reliable coarse-to-fine approach for asphalt pavement crack detection,” Image and Vision Computing, vol. 57, pp. 130–146, 2022.
P. Jiang, D. Ergu, F. Liu, Y. Cai, and B. Ma, “A review of yolo algorithm developments,” Procedia Computer Science, vol. 199, pp. 1066–1073, 2022.
K. Guo, C. He, M. Yang, and S. Wang, “A pavement distresses identification method optimized for yolov5s,” Scientific Reports, vol. 12, p. 3542, 2022.
H. Li, H. Zhang, H. Zhu, K. Gao, H. Liang, and J. Yang, “Automatic crack detection on concrete and asphalt surfaces using semantic segmentation network with hierarchical transformer,” Engineering Structures, vol. 307, p. 117903, 2024.
M. A. Borges and F. A. Spanhol, “BR-AsPavDam – brazilian asphalt pavement damage dataset,” 2025. [Online]. DOI: 10.5281/zenodo.16291115
Ultralytics, “Ultralytics YOLOv5: Real-time object detection,” [link], 2023, commit f4ff536, accessed 25 Jul 2025.
J. Redmon, S. Divvala, R. Girshick, and A. Farhadi, “You Only Look Once: Unified, real-time object detection,” in Proc. 30th IEEE Conf. on Computer Vision and Pattern Recognition, 2016, pp. 779–788.
D. Arya, H. Maeda, S. K. Ghosh, D. Toshniwal, H. Omata, T. Kashiyama, A. Mraz, and Y. Sekimoto, “RDD2020: An image dataset for smartphone-based road damage detection and classification,” 2021. [Online]. DOI: 10.17632/5ty2wb6gvg.1
Publicado
24/11/2025
Como Citar
BORGES, Marcos Augusto; SPANHOL, Fabio Alexandre.
A computer vision application for automatic detection of defects in urban asphalt pavement in Brazil. In: WORKSHOP LATINOAMERICANO DE DEPENDABILIDADE E SEGURANÇA EM SISTEMAS VEICULARES (SSV), 2. , 2025, Campinas/SP.
Anais [...].
Porto Alegre: Sociedade Brasileira de Computação,
2025
.
p. 21-24.
