ACCESS-4FI – Automated Crop Classification and yield Estimation online ServiceS for Food Industry.

This project addresses the challenge of quantification and annual monitoring of the agricultural production, as seen from the perspective of different individual actors in the agriculture sector. The project will be executed primarily in Poland, but it aims to expand to a regional scale in the longer-term perspective.

The main objective is to develop a set of user-dedicated EO-based tools:
– An automated crop recognition and classification tool. The tool will concern the crop types typical for intensive agriculture, focusing on wheat, maize, potato and sugar beet – the major crops of human food supply, and additionally, rapeseed – one of the largest source of vegetable oil in the World.
– An automated, per crop field, yield estimation tool. After crop type identification, the second step will be to forecast and also back-cast the yields. For bigger fields (> 2 ha), field zone yield prediction will additionally be implemented, enabling distinguishing different productivity levels.

Together, these tools will allow the considered actors to optimize their operations, reduce risks, and in turn they will have a smoothing effect on the volatile commodity markets. The tools will be based on image time series provided by the Copernicus Programme (Sentinels 1, 2 and 3), other openly available satellite data (Landsat 8 and MODIS), data from private providers, meteorological data (from weather stations and modelled), as well as a large training and validation dataset with per-field information.

The project is being realized by a consortium consisting of SatAgro Sp. z o.o. (project leader) and Department of Earth Observation of Space Research Centre of the Polish Academy of Sciences (CBK PAN) (research partner).

Project duration: January 2019 – December 2020
Funding: European Space Agency
Principal Investigator: SatAgro Sp. z o.o.
Contact in CBK PAN: dr Edyta Woźniak