Fields of The World: A Field Guide for Extracting Agricultural Field Boundaries

TL;DR

This paper introduces the FTW ecosystem, a comprehensive benchmark and tools for extracting and analyzing agricultural field boundaries globally, enabling improved crop monitoring and classification with limited labeled data.

Contribution

It presents a large-scale benchmark dataset, pre-trained segmentation models, and inference tools for extracting agricultural field boundaries across multiple countries.

Findings

Achieved macro F1 scores of 0.65-0.75 for crop type classification with limited labels.

Mapped crop types at the field level across 24 countries using FTW data.

Demonstrated scalable inference over 4.76 million km² of agricultural land.

Abstract

Field boundary maps are a building block for agricultural data products and support crop monitoring, yield estimation, and disease estimation. This tutorial presents the Fields of The World (FTW) ecosystem: a benchmark of 1.6M field polygons across 24 countries, pre-trained segmentation models, and command-line inference tools. We provide two notebooks that cover (1) local-scale field boundary extraction with crop classification and forest loss attribution, and (2) country-scale inference using cloud-optimized data. We use MOSAIKS random convolutional features and FTW derived field boundaries to map crop type at the field level and report macro F1 scores of 0.65--0.75 for crop type classification with limited labels. Finally, we show how to explore pre-computed predictions over five countries (4.76M km\textsuperscript{2}), with median predicted field areas from 0.06 ha (Rwanda) to 0.28 ha (Switzerland).