Together with the Estonian University of Life Sciences, Minasuudanforest OÜ is developing a method that identifies the main tree species in Estonian forests by remote sensing. The project aims to build and test a machine learning approach that uses Sentinel-2 multispectral time series and forest structure metrics from LiDAR data. The target is an overall accuracy of at least 80% in identifying each stand’s dominant tree species.
The project will deliver a validated tree species classification model, prototype tree species maps for at least three pilot areas, and a technical final report with a roadmap for turning the method into a new service on the MS Forest platform. The result will give forest owners more frequent and cost-effective tree species data. It will also support forest biodiversity assessment under the EU Nature Restoration Regulation.
The total project cost is €28,234, of which the European Union’s contribution is €15,484.
Co-funded by the European Union.