Multi-Class Pavement Disease Recognition Using Object Detection and Segmentation

Pavement disease is an important factor threatening road safety. Most traditional disease recognition methods often rely on manual detection, which is time-consuming and inefficient. In this work, by introducing the object detection and segmentation into the detection of pavement diseases, a multi-c...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:International Conference on Information Science and Technology S. 211 - 216
Hauptverfasser: Zhang, Kuan, Zheng, Mingkai, Yu, Qing, Liu, Yi
Format: Tagungsbericht
Sprache:Englisch
Veröffentlicht: IEEE 14.10.2022
Schlagworte:
Deep learning
Feature extraction
Image segmentation
Information science
Location awareness
Manuals
Object detection
pavement disease detection
semantic segmentation
target localization
ISSN:2573-3311
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Pavement disease is an important factor threatening road safety. Most traditional disease recognition methods often rely on manual detection, which is time-consuming and inefficient. In this work, by introducing the object detection and segmentation into the detection of pavement diseases, a multi-class pavement disease detection method is proposed. First, diseases are located based on YOLOv4. CSPDarknet53 is used as the backbone network. The feature extraction performance is further improved by spatial pyramid pooling. Then, on the basis of pavement disease location, the pyramid scene parsing network (PSPNet) is employed to extract the pixel of the disease area to realize the accurate analysis of the anomaly. The feasibility of the proposed method is verified by a pavement disease detection experiment using the actual road dataset collected from a province in eastern China, including seven common diseases.
ISSN:2573-3311
DOI:10.1109/ICIST55546.2022.9926783