NOA – RECAP crop identification processing chain

The application provides an object-based crop identification processing chain, which exploits time series of Sentinel-2 data and applies Random Forest and Support Vector Machine algorithms.

The application is developed with RECAP H2020 project funding and it is currently being validated with the participation of Paying Agencies and farmers in Spain, Greece and Lithuania.

Potential end users: Primarily Paying Agencies in Europe with a mandate to systematically monitor the implementation of the Common Agricultural Policy (CAP). Other potential customers are agro-consultants and the farmers themselves by reducing the unintentional error rate in their declarations.

ONDA Data used

  • Time-series of Sentinel-2 data, spanning the cultivation periods

Third Party data

  • Farmer declarations, delivered by the Paying Agencies and stored locally on to NOA virtual instances on ONDA