From GIS to Remote SensingI'm glad to announce that the Semi-Automatic Classification Plugin version 9 (codename "Foundation") is under development.
It will be released in early 2026, in time for the release of QGIS 4, which will be based on Qt 6 framework. Therefore, the Semi-Automatic Classification Plugin version 9 will be ported to run on Qt 6 versions of QGIS.
In addition, there will be several new features, in particular:
- new classification algorithms, in particular the introduction of foundation models (hence the plugin codename "Foundation");
- a new simplified interface, in addition to the traditional user interface, designed for new users in order to focus on the classification process;
- new process that should ease the installation of the required dependency "Remotior Sensus".
The development of the Semi-Automatic Classification Plugin version 9 is running in parallel with the development of the Python library "Remotior Sensus", which will be updated to version 0.6, and include several performance enhancements, in addition to the foundation models. Foundation models are a particular type of machine learning models trained on huge amounts of satellite data, which can be adapted to several classification purposes.
This will be the first attempt to integrate deep learning foundation models into the Semi-Automatic Classification Plugin, in particular by using training data on features extracted by the model, thereby easing the processes required for running a pretrained model.
For any comment or question, join the Facebook group or GitHub discussions about the Semi-Automatic Classification Plugin.