> 01_points
Data collected in the field by the enumerators containing information regarding the field characteristics and the crops characteristics.
Data collected using mobile devices and dedicated applications (see the Field guide and annexes).
Within each segment each parcel was digitalized (see 02_polygons_prior_field_campaign). 
Each field where crops were identified based on VHR/HR data was visited in the field resulting in a point recorded with all the information mentioned above.

> 02_polygons_prior_field_campaign
Prior to the field campaign, each segment from the sample design has been visually interpreted by CLS using a combination of the most recent available Very High Resolution (VHR) imagery from Google Earth/Bing Maps, Yandex, Planet and Sentinel-2 imagery. 
All field boundaries (including cropland parcels) were digitalized, resulting in polygons that constitute the sample. 
After digitalizing the fields, the main land cover of each field is determined. Based on the VHR and Sentinel-2 imagery a land cover is assigned following the “Main Land Cover” nomenclature:

1. Forest
2. Grassland
3. Cropland
4. Bare soil
5. Urban
6. Shrubland
7. Water
8. Wetland
These are the polygons given to the enumerators and used as a basis for the field campaign.

> 03_polygons_postprocessed_with_field_info
After the field campaign, the data collected in the field are cleaned and post-processed in GIS.
Information recorded at point-based level are associated to the corresponding polygons (spatial and attributes joins).
This dataset is then used as basis for the training/validation samples selection.