{"name": "BatiScan IA", "package_name": "batiscan", "description": "AI building mapping for QGIS: detect building footprints on satellite imagery (no points needed), or turn field-collected building points into footprints with Segment Anything.", "about": "BatiScan maps buildings from satellite or aerial imagery with AI. The automatic mode uses the Sahel 1 model (BatiScan v6), included in the plugin, a U-Net trained on villages of Mali and on open building footprints from West Africa (30 cm/px), from round huts to rectangular metal and earth roofs. It separates adjacent buildings, draws straight footprints that never overlap, and flags low-score or very small objects (under 20 m\u00b2) for review, with the reason in a control field. Each footprint gets its area, perimeter, centre and orientation, and a geometry check reports invalid, overlapping and duplicate footprints. On 20 hand-checked test zones in two villages kept out of training, it found about 89 percent of the buildings, and about 94 percent of its objects were correct. The model is shipped with the plugin in a compressed form (int8, 25 MB), nothing to download, and is provided for research and testing, not for commercial use. BatiScan also turns building points collected in the field into footprints with Segment Anything (MobileSAM in ONNX, no PyTorch needed), offers a click-to-map tool, a batch tab that counts buildings for hundreds of villages in one run (village boundaries, or points re-centred on the nearest settlement, with a per-village table and a total), a points-versus-footprints quality report, and export of image/mask tiles to train a model for your own area. Any Deepness-compatible ONNX segmentation model can be added. Works with GeoTIFF orthophotos and XYZ/WMS basemaps (Google, Bing, Esri...). Requires the Python packages onnxruntime and opencv, which the plugin offers to install on first use. Less accurate on dense urban areas and very large buildings. Walled yards and flat earth roofs can be confused at this resolution. Training labels include Google Open Buildings (CC BY 4.0), Microsoft Global ML Building Footprints (ODbL) and OpenStreetMap (\u00a9 OpenStreetMap contributors, ODbL).", "homepage": "https://github.com/guindoallayecontact7/batiscan-qgis", "repository": "https://github.com/guindoallayecontact7/batiscan-qgis", "tracker": "https://github.com/guindoallayecontact7/batiscan-qgis/issues", "author": "Allaye GUINDO", "tags": ["building", "buildings", "footprint", "ai", "deep learning", "segment anything", "sam", "onnx", "segmentation", "remote sensing", "africa", "sahel", "mali", "b\u00e2timent"], "downloads": 249, "latest_version": "0.7.0", "versions": [{"version": "0.7.0", "experimental": false, "qgis_min": "3.38.0", "qgis_max": "3.99.0", "downloads": 46, "uploaded_by": "allaye", "upload_datetime": "2026-11-13T22:04:19.299262"}, {"version": "0.5.1", "experimental": false, "qgis_min": "3.38.0", "qgis_max": "3.99.0", "downloads": 101, "uploaded_by": "allaye", "upload_datetime": "2026-11-05T11:01:25.383214"}, {"version": "0.3.0", "experimental": true, "qgis_min": "3.38.0", "qgis_max": "3.99.0", "downloads": 100, "uploaded_by": "allaye", "upload_datetime": "2026-10-26T11:09:18.617527"}]}