Mohamed, Mohamed, Esmat, Adel, Hamdy, Ahmed, El Sheshtawy, Amr. (1405). Assessment of Deep Learning Models for Pavement Distress Detection in High-Resolution UAV Imagery. هنرهای کاربردی, 14(3), -. doi: 10.22075/jrce.2025.2415
Mohamed Eid abdelshakour Mohamed; Adel A. Esmat; Ahmed M. Hamdy; Amr M. El Sheshtawy. "Assessment of Deep Learning Models for Pavement Distress Detection in High-Resolution UAV Imagery". هنرهای کاربردی, 14, 3, 1405, -. doi: 10.22075/jrce.2025.2415
Mohamed, Mohamed, Esmat, Adel, Hamdy, Ahmed, El Sheshtawy, Amr. (1405). 'Assessment of Deep Learning Models for Pavement Distress Detection in High-Resolution UAV Imagery', هنرهای کاربردی, 14(3), pp. -. doi: 10.22075/jrce.2025.2415
Mohamed, Mohamed, Esmat, Adel, Hamdy, Ahmed, El Sheshtawy, Amr. Assessment of Deep Learning Models for Pavement Distress Detection in High-Resolution UAV Imagery. هنرهای کاربردی, 1405; 14(3): -. doi: 10.22075/jrce.2025.2415

Assessment of Deep Learning Models for Pavement Distress Detection in High-Resolution UAV Imagery

Journal of Rehabilitation in Civil Engineering
دوره 14، شماره 3 - شماره پیاپی 43، پاییز 2026    اصل مقاله (1.91 M)
نوع مقاله: Regular Paper
شناسه دیجیتال (DOI): 10.22075/jrce.2025.2415
نویسندگان
Mohamed Eid abdelshakour Mohamed* 1؛ Adel A. Esmat1؛ Ahmed M. Hamdy1؛ Amr M. El Sheshtawy2
1Civil Engineering Department, Faculty of Engineering, Al-Azhar University, Cairo, Egypt.
2Civil Engineering Department, Faculty of Engineering, Al-Azhar University, Cairo, Egypt
چکیده
Pavement distress detection is a critical task for ensuring road safety and maintaining transportation infrastructure, particularly in environments with limited resources. This study conducts a comprehensive evaluation of ten YOLO object detection modelscomprising five YOLOv5 and five YOLOv8 variants (n, s, m, l, x) for identifying seven pavement defect classes using 500 high-resolution UAV images. The dataset, manually annotated and split into 70% training and 30% validation sets, was used to train all models under uniform hyperparameter settings. The performance was assessed using standard metrics: mAP@0.5:0.95, mAP@0.5, Recall, Precision and F1-score. Results showed that YOLOv8 consistently outperformed YOLOv5 across all dataset sizes, with YOLOv8l reaching the highest mAP@0.5:0.95 of 0.381, and YOLOv8m providing the best balance between accuracy (0.344), training time (67 minutes), and robustness. Statistical validation using Two-Way ANOVA confirmed significant performance differences (F = 13.81, p = 0.0006, Cohen’s d = 0.73). Further analysis using Repeated Measures ANOVA and Bonferroni-corrected t-tests reinforced YOLOv8l's superiority over other variants. Despite the limited dataset size the findings demonstrate YOLOv8’s effectiveness and reliability in low-resource conditions without the need for image augmentation or preprocessing. The main innovation lies in providing a statistically validated benchmark for UAV-based pavement monitoring under small dataset conditions, highlighting YOLOv8’s superior efficiency and generalization without augmentation or preprocessing. This study provides a reproducible benchmark for real-time UAV-based pavement distress detection, offering insights for deploying lightweight deep learning systems in resource-constrained settings.

تازه های تحقیق
  • UAV-based YOLOv8 outperforms YOLOv5 in pavement distress detection accuracy
  • First statistical validation of YOLO models under limited dataset conditions
  • ANOVA tests confirm YOLOv8 robustness without data augmentation or preprocessing
  • UAV + deep learning provides cost-efficient large-scale pavement monitoring
کلیدواژه‌ها
YOLOv8؛ YOLOv5؛ Pavement distress detection؛ Low-resource environments
مراجع
آمار
تعداد مشاهده مقاله: 1,545
تعداد دریافت فایل اصل مقاله: 1,083
دانشگاه سمنان
پیوندهای مفید
آمار
تعداد نشریات22
تعداد شماره‌ها722
تعداد مقالات10,383
تعداد مشاهده مقاله72,838,630
تعداد دریافت فایل اصل مقاله64,535,197