{"name": "Drone AI Georef", "package_name": "drone_ai_georef", "description": "Automatic georeferencing of drone photos by AI feature matching against a basemap / Georreferenciaci\u00f3n autom\u00e1tica de fotos de dron con IA.", "about": "Georeferences drone photos (JPG/TIF) without manual ground control points. It reads GPS, altitude and heading from EXIF/XMP, renders the basemap around the photo (Esri World Imagery, OpenStreetMap or any raster/WMS/XYZ layer of the project, e.g. your own orthophoto), finds common patterns with SIFT/ORB or deep-learning matchers (DISK+LightGlue, LoFTR) and estimates the transformation with MAGSAC++/RANSAC plus a refinement pass. Outputs a GeoTIFF, a .points file for the QGIS Georeferencer and a footprint layer with quality indicators. Requires OpenCV (the plugin offers to install opencv-python-headless with pip, after asking). The AI matchers additionally need torch and kornia (optional, ~250 MB). User interface in Spanish. | Georreferencia fotos de dron sin puntos de control manuales: lee GPS/altura/rumbo, descarga el mapa base, reconoce patrones con IA y genera GeoTIFF, .points y huellas con indicadores de calidad.", "homepage": "https://github.com/geosebas05/drone_ai_georef", "repository": "https://github.com/geosebas05/drone_ai_georef", "tracker": "https://github.com/geosebas05/drone_ai_georef/issues", "author": "Geosebas", "tags": ["drone", "uav", "dron", "georeference", "georeferencing", "georreferenciacion", "ai", "deep learning", "feature matching", "sift", "loftr", "lightglue", "orthophoto", "raster"], "downloads": 4, "latest_version": "0.2.1", "versions": [{"version": "0.2.1", "experimental": true, "qgis_min": "3.22.0", "qgis_max": "4.99.0", "downloads": 4, "uploaded_by": "geosebas05", "upload_datetime": "2026-10-21T22:35:57.576991"}]}