Road Damages Detection Object Detection Dataset and Pre-Trained Model by Eric Tam

Images

1385 images

Explore Dataset

The datasets of this project is sourced from Crowdsensing-based Road Damage Detection Challenge (CRDDC2022) (https://crddc2022.sekilab.global). The original challenge only considers four catergoires (classes) due to consistency of datasets collected. However, this project extends the utlitity of the datasets by confining dataset to the collection from Japan (containing classes of 7).

Here are a few use cases for this project:

Municipal Infrastructure Maintenance: City governments can use the "Road Damages Detection" model to efficiently identify and prioritize damaged roads for repair, leading to improved road safety and smoother traffic flow.

Pothole and Road Damage Reporting: Mobile app developers can integrate the model into their apps, allowing users to automatically detect and report road damage, such as potholes, directly to local authorities for faster repairs.

Insurance Claim Assessments: Insurance companies can use the model to analyze road conditions at the time of accidents, helping to determine liability and assess claim validity based on road damage influence on the incident.

Autonomous Vehicle Navigation: Autonomous vehicle systems can use the "Road Damages Detection" model to identify damaged roads ahead and adjust their routing or speed to optimize safety, comfort, and fuel efficiency.

Road Infrastructure Investment Planning: Transportation agencies can use the model to analyze current road conditions, enabling them to allocate funds and resources more effectively based on data-driven infrastructure improvement strategies.

Class of the Defects:

Class Code	Damage Type
0. D00	Longitudinal Crack
1. D10	Transverse Crack
2. D20	Aligator Crack
3. D40	Pothole
4. D43	White Line Blur
5. D44	Cross Walk Blur
6. D50	Manhole Cover (TBC)

Trained Model API

This project has a trained model available that you can try in your browser and use to get predictions via our Hosted Inference API and other deployment methods.

YOLOv8

This project has a YOLOv8 model checkpoint available for inference with Roboflow Deploy. YOLOv8 is a new state-of-the-art real-time object detection model.

Cite This Project

If you use this dataset in a research paper, please cite it using the following BibTeX:

@misc{
                            road-damages-detection_dataset,
                            title = { Road Damages Detection Dataset },
                            type = { Open Source Dataset },
                            author = { Eric Tam },
                            howpublished = { \url{ https://universe.roboflow.com/eric-tam-oz6si/road-damages-detection } },
                            url = { https://universe.roboflow.com/eric-tam-oz6si/road-damages-detection },
                            journal = { Roboflow Universe },
                            publisher = { Roboflow },
                            year = { 2023 },
                            month = { may },
                            note = { visited on 2024-06-12 },
                            }