RoyiLau/HR-VSGeo
π This dataset is associated with our TGRS submission: "HR-VSGeo: HR-VSGeo: A High-Resolution Visible-SAR Geo-Referenced Dataset and Evaluation Benchmark for Complex Geological Scenes" HR-VSGeo Series Dataset Dataset Summary The HR-VSGeo Series dataset contains high-resolution Synthetic Aperture Radar (SAR) imagery, suitable for tasks such as object detection, semantic segmentation, and image classification. The data is collected from Google Earth and theβ¦ See the full description on the dataset page: https://huggingface.co/datasets/RoyiLau/HR-VSGeo.
π This dataset is associated with our TGRS submission: **"HR-VSGeo: HR-VSGeo: A High-Resolution Visible-SAR Geo-Referenced Dataset
and Evaluation Benchmark for Complex Geological Scenes"**
HR-VSGeo Series Dataset
Dataset Summary
The HR-VSGeo Series dataset contains high-resolution Synthetic Aperture Radar (SAR) imagery, suitable for tasks such as object detection, semantic segmentation, and image classification. The data is collected from Google Earth and the Gaofen-3 satellite. Annotation work was carried out by the EOC (Elites of CUMT) Lab at China University of Mining and Technology and other collaborators.
Dataset Structure
- Number of images: 5,147 visible light images and 1,319 pairs of visible-SAR data
- Image format: TIFF
- Annotation format: JSON, Shapefile
- Resolution: 2000 Γ 2000 pixels
- Number of classes: 15
Example Images
- Optical Image Example <div style="display: flex; gap: 5px;"> <img src="https://raw.githubusercontent.com/RoyiLau/HR-VSAR/main/Examples/104part1.png?v=1" width="200"/> <img src="https://raw.githubusercontent.com/RoyiLau/HR-VSAR/main/Examples/104part8.png?v=1" width="200"/> <img src="https://raw.githubusercontent.com/RoyiLau/HR-VSAR/main/Examples/90part1.png?v=1" width="200"/> <img src="https://raw.githubusercontent.com/RoyiLau/HR-VSAR/main/Examples/90part15.png?v=1" width="200"/> </div>
- Optical and SAR Image Examples <div style="display: flex; gap: 5px; flex-wrap: wrap;"> <img src="https://raw.githubusercontent.com/RoyiLau/HR-VSAR/main/Examples/108part4.png?v=1" width="200"/> <img src="https://raw.githubusercontent.com/RoyiLau/HR-VSAR/main/Examples/108part4sar.png?v=1" width="200"/> <img src="https://raw.githubusercontent.com/RoyiLau/HR-VSAR/main/Examples/99part15.png?v=1" width="200"/> <img src="https://raw.githubusercontent.com/RoyiLau/HR-VSAR/main/Examples/99part15sar.png?v=1" width="200"/> </div>
- Optical Image with Segmentation Masks <table> <tr> <th>Original</th> <th>Mask</th> <th>Original</th> <th>Mask</th> </tr> <tr> <td><img src="https://raw.githubusercontent.com/RoyiLau/HR-VSAR/main/Examples1/299.png?v=1" width="190"/></td> <td><img src="https://raw.githubusercontent.com/RoyiLau/HR-VSAR/main/Examples1/299mask.png?v=1" width="190"/></td> <td><img src="https://raw.githubusercontent.com/RoyiLau/HR-VSAR/main/Examples1/316.png?v=1" width="190"/></td> <td><img src="https://raw.githubusercontent.com/RoyiLau/HR-VSAR/main/Examples1/316mask.png?v=1" width="190"/></td> </tr> <tr> <td><img src="https://raw.githubusercontent.com/RoyiLau/HR-VSAR/main/Examples1/334.png?v=1" width="190"/></td> <td><img src="https://raw.githubusercontent.com/RoyiLau/HR-VSAR/main/Examples1/334mask.png?v=1" width="190"/></td> <td><img src="https://raw.githubusercontent.com/RoyiLau/HR-VSAR/main/Examples1/516.png?v=1" width="190"/></td> <td><img src="https://raw.githubusercontent.com/RoyiLau/HR-VSAR/main/Examples1/516mask.png?v=1" width="190"/></td> </tr> </table>
Usage
To use the dataset with Hugging Face datasets:
from datasets import load_dataset
dataset = load_dataset("your-username/HR-VSGeo")
