非正规聚居区检测 | Informal Settlement Detection
Detects informal settlements (urban villages / shantytowns) by fusing texture irregularity, building density, and spectral mixing, serving urban renewal surveys and human settlement environment monitoring.
Core algorithm: quantifies texture disorder with the local standard deviation; combines building density with NDVI (low vegetation), weighted on an absolute physical scale to derive an informal score in [0, 1]; thresholding then extracts the informal settlements. High texture + high density + low NDVI → high score.
Dependencies / 依赖
pip install 'numpy' 'rasterio' 'scipy'
Usage / 使用方法
Basic usage
python geoskill-informal-settlement-detection.py --bbox 116.0 39.0 117.0 40.0 [other arguments]
Examples
Example 1 (Synthetic data (offline))
python geoskill-informal-settlement-detection.py --bbox 116.0 39.0 117.0 40.0 --synthetic --output-dir ./out
Example 2 (Usage 2)
python geoskill-informal-settlement-detection.py --input multispectral.tif --footprints fp.tif --output-dir ./out
Example 3 (Usage 3)
python geoskill-informal-settlement-detection.py --bbox 121.0 31.0 122.0 32.0 --threshold 0.6 --output-dir ./out --quiet
Example 4 (Usage 4)
python geoskill-informal-settlement-detection.py --input ms.tif --kernel-size 7 --output-dir ./out
Example 5 (Usage 5)
python geoskill-informal-settlement-detection.py --bbox 116.0 39.0 117.0 40.0 --synthetic --threshold 0.4 --output-dir ./out --quiet
Output / 输出
| File | Format | Description |
|---|---|---|
informal_score.tif |
GeoTIFF | Two bands: band1 = informal score, band2 = classification mask |
informal_stats.json |
JSON | Mean score, informal ratio, threshold |
output-manifest.json |
JSON | Run manifest |
Data Source / 数据源 / Source
Local multispectral GeoTIFF (Red, NIR) + optional building footprints; --synthetic mode simulates a scene with half informal and half formal areas.
Privacy / 隐私声明 / Privacy
- Runs offline by default;
--syntheticmode is fully offline with no network access. - All processing is performed locally; user data is never uploaded.
License / License
MIT
name: geoskill-informal-settlement-detection description: 'Detect informal settlements by fusing texture irregularity, building morphology and spectral mixing into a classification score.'
非正规聚居区检测 | Informal Settlement Detection
融合纹理不规则性、建筑密度与光谱混合检测非正规聚居区（城中村/棚户区），服务于城市更新调查与人居环境监测。
核心算法：用局部标准差量化纹理无序度；结合建筑密度与 NDVI（低植被），以绝对物理标度加权得到非正规评分 [0,1]，阈值分割出非正规聚居区。高纹理 + 高密度 + 低 NDVI → 高评分。
依赖
pip install 'numpy' 'rasterio' 'scipy'
使用方法
基本用法
python geoskill-informal-settlement-detection.py --bbox 116.0 39.0 117.0 40.0 [其他参数]
示例
示例 1（合成数据（离线））
python geoskill-informal-settlement-detection.py --bbox 116.0 39.0 117.0 40.0 --synthetic --output-dir ./out
示例 2（用法 2）
python geoskill-informal-settlement-detection.py --input multispectral.tif --footprints fp.tif --output-dir ./out
示例 3（用法 3）
python geoskill-informal-settlement-detection.py --bbox 121.0 31.0 122.0 32.0 --threshold 0.6 --output-dir ./out --quiet
示例 4（用法 4）
python geoskill-informal-settlement-detection.py --input ms.tif --kernel-size 7 --output-dir ./out
示例 5（用法 5）
python geoskill-informal-settlement-detection.py --bbox 116.0 39.0 117.0 40.0 --synthetic --threshold 0.4 --output-dir ./out --quiet
输出
| 文件 | 格式 | 说明 |
|---|---|---|
informal_score.tif |
GeoTIFF | 双波段：band1=非正规评分，band2=分类掩膜 |
informal_stats.json |
JSON | 评分均值、非正规比例、阈值 |
output-manifest.json |
JSON | 运行清单 |
数据源 / Source
本地多光谱 GeoTIFF（Red, NIR）+ 可选建筑足迹；--synthetic 模式模拟非正规区与正规区各半的场景。
隐私声明 / Privacy
- 默认离线运行，
--synthetic模式完全无网络。 - 所有处理在本地完成，不上传用户数据。
License
MIT