The first step processes both data sources for facilitating their integration, while the novelty lies in the second step, where additional pseudo-labels are generated by classifying crop parcels with ...