Crop identification using deep learning on LUCAS crop cover photos

Crop classification via deep learning on imagery is a key method for delivering timely and accurate crop-specific information to various stakeholders. Automatic labelling is essential when collecting large volumes of data. One such data collection is the EU's Land Cover Land Use Survey (LUCAS), and in particular, the recently published LUCAS Cover photos database. We select and publish a subset of LUCAS Cover photos for 12 major crops across the EU, to deploy, benchmark, and identify the best configuration of Mobile-net for the classification task. The work has produced a dataset of 169,460 images of mature crops for the 12 classes, out of which 15,876 were manually selected as representing a clean sample without any foreign objects or unfavorable conditions.

YORDANOV Momchil;  D'ANDRIMONT Raphael;  MARTINEZ SANCHEZ Laura;  LEMOINE Guido;  FASBENDER Dominique;  VAN DER VELDE Marijn; 

2023-08-04

MDPI

JRC132339

1424-8220 (online),   

https://doi.org/10.3390/s23146298,    https://publications.jrc.ec.europa.eu/repository/handle/JRC132339,   

10.3390/s23146298 (online),