Accès ouvert · CC BY

Deep learning–based pavement crack classification and road condition assessment for intelligent transportation infrastructure

Article scientifique 2026 Sciences de l'ingénieur et technologies Anglais

Résumé

Introduction Efficient maintenance of road infrastructure is essential for ensuring traffic safety, preserving pavement serviceability, and supporting the long-term sustainability of transportation networks. However, conventional pavement inspection mainly relies on manual visual assessment, which is time-consuming, costly, and subjective. This study proposes a deep learning–based framework for automated pavement crack classification and road condition assessment. Methods The proposed framework uses real-world pavement images collected through the Multifunctional Pavement Assessment System (SMAC). Transfer learning was applied using ResNet50 and EfficientNet-B0 architectures to classify six pavement surface categories. Data augmentation and weighted cross-entropy loss were employed to improve robustness and address class imbalance. Model performance was evaluated using accuracy, precision, recall, F1-score, and confusion matrix analysis. A Road Condition Index (RCI) was then formulated from the classification results to support maintenance prioritization. Results EfficientNet-B0 achieved an overall classification accuracy of 89.8%, outperforming ResNet50 (73.3%). It obtained F1-scores of 0.92 for alligator and longitudinal cracks, 0.93 for transverse cracks, and 1.00 for non-cracked pavement surfaces. The proposed RCI successfully differentiated pavement degradation levels and supported maintenance prioritization. Discussion The proposed framework provides an intelligent, interpretable, and practical solution for pavement crack classification, road condition assessment, and maintenance-oriented decision support. These results demonstrate the potential of deep learning for scalable pavement monitoring and more efficient infrastructure management within intelligent transportation systems.

Citer ce document

Chouiekh, C., Hammouch, W., Yahyaouy, A., & Sabri, M. A. (2026). Deep learning–based pavement crack classification and road condition assessment for intelligent transportation infrastructure. Frontiers in Signal Processing. https://doi.org/10.3389/frsip.2026.1901600

Exporter : BibTeX · RIS (Zotero, Mendeley, EndNote)

Accès au document

Texte intégral archivé sur SPHAERO, en accès ouvert : lisible gratuitement en ligne après connexion. Le document reste aussi disponible à sa source. Se connecter

Voir l'article sur le site de la revue

Licence et provenance

Licence : CC BY

Notice moissonnée depuis OpenAlex le 26/10/2026. Le document reste hébergé par sa source.
Voir le document à la source →

Statistiques

Consultations (hors robots) : 14

Téléchargements (hors robots) : 0

Méthode de comptage