[Paper Review] RDD2022: A multi-national image dataset for automatic Road Damage Detection

Deeksha Arya, Hiroya Maeda+3 more|arXiv (Cornell University)|Sep 18, 2022

Infrastructure Maintenance and Monitoring85 citations

TL;DR

RDD2022 provides a large, annotated multi-national road damage image dataset (47,420 images, 55,000+ damage instances) for training and benchmarking deep learning models for automatic road damage detection and classification across six countries.

ABSTRACT

The data article describes the Road Damage Dataset, RDD2022, which comprises 47,420 road images from six countries, Japan, India, the Czech Republic, Norway, the United States, and China. The images have been annotated with more than 55,000 instances of road damage. Four types of road damage, namely longitudinal cracks, transverse cracks, alligator cracks, and potholes, are captured in the dataset. The annotated dataset is envisioned for developing deep learning-based methods to detect and classify road damage automatically. The dataset has been released as a part of the Crowd sensing-based Road Damage Detection Challenge (CRDDC2022). The challenge CRDDC2022 invites researchers from across the globe to propose solutions for automatic road damage detection in multiple countries. The municipalities and road agencies may utilize the RDD2022 dataset, and the models trained using RDD2022 for low-cost automatic monitoring of road conditions. Further, computer vision and machine learning researchers may use the dataset to benchmark the performance of different algorithms for other image-based applications of the same type (classification, object detection, etc.).

Motivation & Objective

• Provide a large-scale, multi-national image dataset for automatic road damage detection and classification.
• Annotate road damage instances to enable supervised learning of deep learning models.
• Promote cross-country benchmarking through a global challenge (CRDDC2022).
• Support municipalities and road agencies in low-cost, automatic road condition monitoring.

Proposed method

• Assemble and annotate a diverse set of road images from six countries.
• Annotate more than 55,000 road damage instances across four damage types.
• Release the dataset as part of the Crowd sensing-based Road Damage Detection Challenge (CRDDC2022).
• Encourage researchers to develop and benchmark detection/classification methods on this dataset.

Experimental results

Research questions

• RQ1 How well can deep learning models detect and classify road damage across multiple countries using a single dataset?
• RQ2 What is the performance of different algorithms for detecting four damage types: longitudinal cracks, transverse cracks, alligator cracks, and potholes?
• RQ3 Can CRDDC2022 standards promote robust cross-country road damage detection methods?

Key findings

• Dataset comprises 47,420 road images from six countries.
• Over 55,000 instances of road damage are annotated.
• Four damage types are annotated: longitudinal cracks, transverse cracks, alligator cracks, potholes.
• Datasets are intended to benchmark deep learning-based detection and classification methods.
• Platforms like CRDDC2022 encourage global participation and benchmarking.
• Municipalities and road agencies may use models trained on RDD2022 for low-cost monitoring.