CHN6-CUG-Roads-Dataset

Version 1.0

1.Overview

To advance the state of the arts in road extraction, CHN6-CUG Roads Dataset is constructed, which is a manually labeled pixel-level VHR satellite imagery. This data set is a new largescale satellite image data set of representative cities in China. The images are selected from six cities with different levels of urbanization, i.e., the Chaoyang area of Beijing, the Yangpu District of Shanghai, Wuhan city center, the Nanshan area of Shenzhen, the Shatin area of Hong Kong, and Macao.

2.Dataset

dataset tree:

CHN6-CUG
└───train
│   └───gt
│   └───images
└───val
│   └───gt
│   └───images

We collected typical areas of six cities in China from Google Earth. The images are with 50 cm/pixel spatial resolution.

Please note that we do not own the copyrights to these original satellite images. Their use is RESTRICTED to non-commercial research and educational purposes.

3.Download

Download link:

BaiduYun（Password: qjpd）

4.Reference

Please cite this paper if you use this dataset:

@article{zhu2021global,
  title={A Global Context-aware and Batch-independent Network for road extraction from VHR satellite imagery},
  author={Zhu, Qiqi and Zhang, Yanan and Wang, Lizeng and Zhong, Yanfei and Guan, Qingfeng and Lu, Xiaoyan and Zhang, Liangpei and Li, Deren},
  journal={ISPRS Journal of Photogrammetry and Remote Sensing},
  volume={175},
  pages={353--365},
  year={2021},
  publisher={Elsevier}
}

5.Contact

If you have any problem in using the CHN6-CUG Roads Dataset, please contact: yanan.zhang@cug.edu.cn