GeoAI Suite for QGIS
The GeoAI suite is a set of integrated QGIS extensions for annotation, training, and inference of deep learning models on geospatial data, along with advanced geometric tools for data preparation and post-processing.
Plugins
QAnnotate - Raster Annotation
Allows the creation of annotated datasets for training GeoAI models. Includes raster preparation, manual or assisted annotation, and export of training-ready datasets.
QModel Trainer - Model Training
Allows training deep learning models directly within QGIS from datasets created with QAnnotate, with full configuration of training parameters, data augmentation, device selection, and advanced options.
QPredict - Model Inference
Applies trained models from QModel Trainer on georeferenced raster data to produce raster or vector layers that can be directly used in QGIS.
QToolbox - Raster and Vector Geometric Tools
Includes advanced geospatial tools to prepare, regularize, and improve raster and vector data, particularly for automated production and vectorization workflows.
Suite Philosophy
- Consistent and integrated interface across all plugins
- Guided workflow from annotation to training and inference
- Detailed logging for monitoring and diagnostics
- Flexible and extensible tools for GeoAI needs