https://github.com/chenzhaiyu/footprint-detection
Building Footprint Detection from Satellite Images
https://github.com/chenzhaiyu/footprint-detection
building footprint
Last synced: about 2 months ago
JSON representation
Building Footprint Detection from Satellite Images
- Host: GitHub
- URL: https://github.com/chenzhaiyu/footprint-detection
- Owner: chenzhaiyu
- Created: 2019-03-08T08:27:22.000Z (over 6 years ago)
- Default Branch: master
- Last Pushed: 2021-04-06T15:37:31.000Z (over 4 years ago)
- Last Synced: 2025-08-03T04:23:00.951Z (2 months ago)
- Topics: building, footprint
- Language: Jupyter Notebook
- Homepage:
- Size: 2.94 MB
- Stars: 16
- Watchers: 1
- Forks: 6
- Open Issues: 1
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Building Footprint Detection
Building footprint detection from satellite/aerial images with Mask R-CNN and image matting (KNN matting, closed-form matting and Grabcut).

## Usage
Run `python detect.py` for detection, `python evaluate.py` for evaluation, and `python regularize.py` for regularizing the footprint. You may have to change the hardcoded paths in these scripts.
## Dataset
You can download the [cropped aerial image tiles and raster labels](https://study.rsgis.whu.edu.cn/pages/download/3.%20The%20cropped%20aerial%20image%20tiles%20and%20raster%20labels.zip) from [WHU Building Dataset](https://study.rsgis.whu.edu.cn/pages/download/building_dataset.html).
## File Structure
```
│ detect.py
│ evaluate.py
│ mrcnn_demo.ipynb
│
├─matting
│ closed_form_matting.py
│ grabcut.py
│ knn_matting.py
│ solve_foreground_background.py
│
├─models
│ models_here.txt
│
├─mrcnn
│ config.py
│ model.py
│ parallel_model.py
│ utils.py
│ visualize.py
│ __init__.py
│
└─polygon
regularize.py
```