WUSU is a semantic understanding dataset focusing on urban structure and the urbanization process in Wuhan, the main city of the Yangtze River Economic Belt. The dataset covers key development areas including Jiang'an District and Hongshan District, spanning a total geographic area of nearly
The dataset includes high resolution, tri-temporal, and multi-spectral satellite images of these districts, offering unprecedented detail and continuity in the representation of urban changes.
The images in this dataset were obtained from GF-2 satellite imagery and were selected with a maximum cloud cover of 5%. These images then underwent several pre-processing steps, including PAN and MSS image orthographic correction and image pan sharpening.
Multi-temporal image registration was also performed to ensure suitability for change detection tasks, with a requirement that the root-mean-square error (RMSE) was less than 1 pixel.
Finally, the images were cropped to the following dimensions:
- Images for Hongshan District are of size 6358×6382 pixels, and were obtained for the dates April 22, 2015, September 1, 2016, and November 23, 2018.
- Images for Jiang'an District are of size 7025×5500 pixels, and were obtained for the dates April 22, 2015, December 8, 2016, and November 23, 2018.
The result of this process was a set of 4-band fusion satellite images with a spatial resolution of 1 meter, providing a high-resolution view of the study areas.
The WUSU dataset adopts the Chinese Land Use Classification Criteria (GB/T21010-2017) and the Contents and Indicators of the Chinese Geographical Conditions Census (GDPJ 01-2013).
Class Legend: This system includes the following classes:
| Class | Color |
|---|---|
| 1-Road |  |
| 2-Low building |  |
| 3-High building |  |
| 4-Arable land |  |
| 6-Woodland |  |
| 7-Grassland |  |
| 8-River |  |
| 9-Lake |  |
| 10-Structure |  |
| 11-Excavation |  |
| 12-Bare surface |  |
Several public datasets with spatial resolution better than 10 m for land use/land cover classification, binary change detection and semantic change detection (excluding competition datasets) mentioned in our paper are as follows:
-
ISPRS 2D Semantic Labeling Contest dataset. [Download] [CrossRef]
-
DeepGlobe Land Cover Classification Challenge dataset. [Download] [CrossRef]
-
Dense Labelling Remote Sensing Dataset (DLRSD). [Download] [CrossRef]
-
Land-cOVEr Domain Adaptive semantic segmentation (LoveDA) dataset. [Download] [CrossRef]
-
Satellite-UAV heterogeneous image change detection (HTCD). [Download] [CrossRef]
-
Sun Yat-Sen University (SYSU)-CD dataset. [Download] [CrossRef]
-
Cropland change detection (CLCD) dataset. [Download] [CrossRef]
-
Multisource built-up change (MSBC) datasets. [Download] [CrossRef]
The WUSU dataset is made available under the Open Database License. Any rights in individual contents of the database are licensed under the Database Contents License.The copyright of the WUSU dataset belongs to Intelligent Data Extraction, Analysis and Applications of Remote Sensing(RSIDEA) academic research group, State Key Laboratory of Information Engineering in Surveying, Mapping, and Remote Sensing (LIESMARS), Wuhan University. The WUSU dataset can be used for academic purposes only and need to cite the following paper, but any commercial use is prohibited. Otherwise, RSIDEA of Wuhan University reserves the right to pursue legal responsibility.
If you use WUSU dataset, codebase or models in your research, please consider citing our paper. [Open Access]
@Article{shi2023openWUSU,
AUTHOR = {Shi, Sunan and Zhong, Yanfei and Liu, Yinhe and Wang, Jue and Wan, Yuting and Zhao, Ji and Lv, Pengyuan and Zhang, Liangpei and Li, Deren},
TITLE = {Multi-temporal urban semantic understanding based on GF-2 remote sensing imagery: from tri-temporal datasets to multi-task mapping},
JOURNAL = {International Journal of Digital Earth},
VOLUME = {16},
YEAR = {2023},
ISSUE = {1},
PAGES={3321-3347},
DOI = {10.1080/17538947.2023.2246445}
}
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent