 This paper proposes a novel methodology for producing land cover maps at country scale from high-resolution optical images. It utilizes existing databases as reference data for training and validation and employs a supervised classifier to process the entire data set. The resulting maps have a CAPA score of 0.86 and can be delivered quickly after the acquisition of the imagery. Additionally, the confidence map provides information at the pixel level regarding the accuracy of the results. This article was authored by Jordan Glotta, Arthur Vincent, Marcella Arias, and others.