Real-Time Pothole Detection System Using Phase Congruency Enhanced YOLO Model on Edge Device
IEEE Access · 2026 · 14 · 90354-90364
DOI: 10.1109/access.2026.3702859 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Access, 2026. Verfasst von Hemanth Padala und 4 weiteren Autor:innen. Die Arbeit nennt 28 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Vellore Institute of Technology (VIT), Chennai Campus, Chennai, India.
Abstract (Originalsprache)
Road defects such as potholes create major safety risks, reduce the lifespan of roads, and disrupt transportation. Traditional inspection methods are usually slow, require a significant workforce, and are not suited for large-scale or real-time monitoring. To solve this, the paper proposes a pothole-detection system that uses Phase Congruency (PC) enhanced You Only Look Once-version 8 nano (YOLOv8n) model which is optimized to run efficiently on a Raspberry Pi 5 (8 GB of internal RAM) edge device. At first, PC model is utilized to extract the phase information from the input images that remains stable regardless of lighting conditions. Then, these phase images are fed into selected channel of YOLOv8n to strengthen edge and texture detection. The proposed model achieves better mAP@0.5, precision and F1-score values than the other compared methods indicating more consistent and reliable pothole detection performance. Finally, the proposed model is deployed on a Raspberry Pi 5 along with a Pi Camera v2 allowing real-time pothole detection directly on the device. The system also integrates a Telegram notification bot for instant alerts and provides a web-based dashboard to visualize detection, monitor live footage, and assess system performance.
Förderung
- Vellore Institute of Technology (VIT), Chennai Campus, Chennai, India
Zitieren
APA
Padala, H., Paul, S., Choudhary, P. P., Vats, P. K., & Raj, A. (2026). Real-Time Pothole Detection System Using Phase Congruency Enhanced YOLO Model on Edge Device. IEEE Access, 14, 90354-90364. https://doi.org/10.1109/access.2026.3702859
Vancouver
Padala H, Paul S, Choudhary PP, Vats PK, Raj A. Real-Time Pothole Detection System Using Phase Congruency Enhanced YOLO Model on Edge Device. IEEE Access. 2026;14:90354-90364. doi:10.1109/access.2026.3702859
BibTeX
@article{Padala2026,
author = {Padala, Hemanth and Paul, Sourabh and Choudhary, Pranav Purushottam and Vats, Prasann Kumar and Raj, Ashutosh},
title = {Real-Time Pothole Detection System Using Phase Congruency Enhanced YOLO Model on Edge Device},
journal = {IEEE Access},
year = {2026},
volume = {14},
pages = {90354--90364},
doi = {10.1109/access.2026.3702859}
}
EUVIMED