Segment Anything for Satellite Imagery: A Strong Baseline and a Regional Dataset for Automatic Field Delineation

Carmelo Scribano, Elena Govi, Paolo Bertellini, Simone Parisi, Giorgia Franchini +1 more
6/19/2025
cs.CVcs.AI

No official code found. Best implementation: cscribano/ERAS-dataset (4 stars, last commit 2026-01-13).

Abstract

Accurate mapping of agricultural field boundaries is essential for the efficient operation of agriculture. Automatic extraction from high-resolution satellite imagery, supported by computer vision techniques, can avoid costly ground surveys. In this paper, we present a pipeline for field delineation based on the Segment Anything Model (SAM), introducing a fine-tuning strategy to adapt SAM to this task. In addition to using published datasets, we describe a method for acquiring a complementary regional dataset that covers areas beyond current sources. Extensive experiments assess segmentation accuracy and evaluate the generalization capabilities. Our approach provides a robust baseline for automated field delineation. The new regional dataset, known as ERAS, is now publicly available.

View on arXivView PDF

Code Implementations(1)

cscribano/ERAS-datasetCommunity100%

Emilia-Romagna Agri Seg (ERAS) field segmentation dataset

40PythonJun 19, 20259 months ago

Similar Papers

Investigating the Segment Anything Foundation Model for Mapping Smallholder Agriculture Field Boundaries Without Training Labels

92% match2024

Delineate Anything Flow: Fast, Country-Level Field Boundary Detection from Any Source

92% match2025

Delineate Anything: Resolution-Agnostic Field Boundary Delineation on Satellite Imagery

91% match1 repo1382025

fabSAM: A Farmland Boundary Delineation Method Based on the Segment Anything Model

88% match2025

Multi-Region Transfer Learning for Segmentation of Crop Field Boundaries in Satellite Images with Limited Labels

87% match2024

A region-wide, multi-year set of crop field boundary labels for Africa

86% match2024

Cite this paper

@article{scribano2025segment,
  title  = {Segment Anything for Satellite Imagery: A Strong Baseline and a Regional Dataset for Automatic Field Delineation},
  author = {Carmelo Scribano and Elena Govi and Paolo Bertellini and Simone Parisi and Giorgia Franchini and Marko Bertogna},
  year   = {2025},
  eprint = {2506.16318},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2506.16318v2}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2506.16318.svg)](https://codeofpaper.com/paper/2506.16318)

Discussion