dronnix-io/drone-building-extraction
收藏Hugging Face2026-04-07 更新2026-04-12 收录
下载链接:
https://hf-mirror.com/datasets/dronnix-io/drone-building-extraction
下载链接
链接失效反馈官方服务:
资源简介:
---
license: cc-by-4.0
task_categories:
- image-segmentation
language:
- en
tags:
- geospatial
- remote-sensing
- drone
- building-extraction
- semantic-segmentation
- qgis
- gis
pretty_name: Drone Building Extraction Dataset — Calgary
size_categories:
- n<1K
---
# 🏘️ Drone Building Extraction Dataset — Calgary
A high-resolution drone imagery dataset for binary building segmentation,
captured over a residential area in **Calgary, Alberta, Canada**.
Designed to work directly with [GeoSeg Studio](https://github.com/dronnix-io/GeoSegStudio)
— an open-source QGIS plugin for deep learning semantic segmentation.
---
## 🖼️ Preview
<table>
<tr>
<td align="center"><b>Train Area</b> (red = building polygons)</td>
<td align="center"><b>Test Area</b> (blue = building polygons)</td>
</tr>
<tr>
<td><img src="preview_train.png" width="100%"/></td>
<td><img src="preview_test.png" width="100%"/></td>
</tr>
</table>
---
## 📋 Dataset Summary
| Property | Value |
|----------|-------|
| **Location** | Calgary, Alberta, Canada |
| **Drone** | DJI Matrice 4E (M4E) |
| **Sensor** | RGB + Alpha (4 bands) |
| **CRS** | EPSG:32611 (UTM Zone 11N) |
| **Annotation** | Manual — building polygons (QGIS) |
| **Annotator** | Dronnix team |
| **Capture date** | July 11, 2025 |
| **License** | CC BY 4.0 |
---
## 📁 Dataset Structure
drone-building-extraction/
├── train_raster.tif ← 4-band GeoTIFF orthomosaic (larger area)
├── train_vector.geojson ← Building polygons for training area
├── test_raster.tif ← 4-band GeoTIFF orthomosaic (smaller area)
└── test_vector.geojson ← Building polygons for test area
---
## 🔍 About the Data
- **Rasters:** High-resolution 4-band (RGB + Alpha) orthomosaics produced
from drone imagery using DJI Terra software. The alpha channel marks
areas with no image coverage (no-data mask). Provided as full-area
GeoTIFFs — tiling is handled by GeoSeg Studio's Prepare tab.
- **Vectors:** Building footprint polygons manually digitized in QGIS by
the Dronnix team. Stored as GeoJSON in EPSG:32611.
- **Labels:** Binary — building (foreground) vs. background.
- **Train area:** Larger residential neighborhood in Calgary
- **Test area:** Smaller adjacent region for independent evaluation
---
## 🚀 How to Use
This dataset is designed for use with **GeoSeg Studio**:
1. Install [GeoSeg Studio](https://github.com/dronnix-io/GeoSegStudio) QGIS plugin
2. In the **Prepare** tab, load `train_raster.tif` as the raster layer and `train_vector.geojson` as the vector layer
3. Set tile size, stride, and output directory
4. Run **Apply Clipping** → **Apply Splitting** → **Apply Augmentation**
5. Train your model in the **Train** tab
6. Evaluate on the test set using `test_raster.tif` and `test_vector.geojson`
---
## 📜 License
This dataset is licensed under
[Creative Commons Attribution 4.0 International (CC BY 4.0)](https://creativecommons.org/licenses/by/4.0/).
You are free to use, share, and adapt this dataset for any purpose,
including commercial use, as long as you give appropriate credit to
**Dronnix** ([dronnix.com](https://www.dronnix.com)).
---
## 📬 Citation
If you use this dataset in your research or project, please cite:
Dronnix (2025). Drone Building Extraction Dataset — Calgary.
https://huggingface.co/datasets/dronnix-io/drone-building-extraction
---
<div align="center">
<sub>Created by <a href="https://www.dronnix.com">Dronnix</a></sub>
</div>
提供机构:
dronnix-io



