Skip to content

Latest commit

 

History

History
68 lines (49 loc) · 3.17 KB

README.md

File metadata and controls

68 lines (49 loc) · 3.17 KB

CeyRo Traffic Sign and Traffic Light Dataset

classes_grid

Overview

CeyRo is a novel benchmark dataset for traffic sign and traffic light detection which covers a wide variety of challenging urban, sub-urban and rural road scenarios in Sri Lanka. The dataset consists of 7984 total images of 1920 × 1080 resolution with 10176 traffic sign and traffic light instances belonging to 70 traffic sign and 5 traffic light classes.

For more details, please refer to our paper Towards Real-time Traffic Sign and Traffic Light Detection on Embedded Systems.

Download

The train set, the test set and a sample of the CeyRo traffic sign and traffic light dataset can be downloaded from the following Google Drive links.

Annotations

The traffic sign and traffic light annotations are provided as bounding boxes in the PASCAL VOC format. LabelImg can be used to visualize the bounding box annotations (Images and XML files should be copied to the same folder).

annotation_format

Statistics

The number of traffic sign and traffic light instances present in each superclass is listed in the below table. For detailed information please refer this sheet.

Superclass Train Test Total
Danger Warning Signs (DWS) 2833 809 3642
Mandatory Signs (MNS) 453 128 581
Prohibitory Signs (PHS) 650 195 845
Priority Signs (PRS) 115 26 141
Speed Limit Signs (SLS) 735 237 972
Other Signs Useful for Drivers (OSD) 1619 498 2117
Additional Regulatory Signs (APR) 377 123 500
Traffic Light Signs (TLS) 1075 303 1378
Total 7857 2319 10176

Evaluation

The evaluation script requires the following dependencies to be installed with Python 3.

pip install argparse shapely tabulate

The class-wise and overall results can be obtained by running the provided python script as follows. <gt_dir> should contain the ground truth xml files and <pred_dir> should contain the prediciton xml files following the same format as per the ground truth.

python eval.py --gt_dir=<gt_dir> --pred_dir=<pred_dir>

Citation

If you use our dataset in your work, please cite the following paper.

@article{jayasinghe2022towards,
  title={Towards Real-time Traffic Sign and Traffic Light Detection on Embedded Systems},
  author={Jayasinghe, Oshada and Hemachandra, Sahan and Anhettigama, Damith and Kariyawasam, Shenali and Wickremasinghe, Tharindu and Ekanayake, Chalani and Rodrigo, Ranga and Jayasekara, Peshala},
  journal={arXiv preprint arXiv:2205.02421},
  year={2022}
}