N-RDD2024:Road damage and defects

Published: 20 November 2024| Version 5 | DOI: 10.17632/27c8pwsd6v.5
Contributors:
,

Description

The RDD2022 dataset contains road images from six countries (India, Japan, Czech Republic, Norway, China, and USA). However, in the presented dataset, four damage types were considered. There are many road defects in road networks. The edited and updated dataset is called N-RDD2024. 10 different types of defects were considered in this dataset. The defect classes labeled are longitudinal cracks (D00), transverse cracks (D10), alligator cracks (D20), repaired cracks (D30), potholes (D40), pedestrian crossing blurs (D50), lane line blurs (D60), manhole covers (D70), patchy road sections (D80) and rutting (D90), respectively. The process of detecting and classifying all defects in road pavement will become more robust for institutions/organizations and researchers.

Files

Institutions

Erzurum Teknik Universitesi

Categories

Image Processing, Object Detection, Machine Learning, Surface Damage, Road Construction, Road Transportation, Roadway Design, Road Safety, Damage Classification

Licence