Satellite Image Semantic Segmentation - Archive ouverte HAL
Pré-Publication, Document De Travail Année : 2021

Satellite Image Semantic Segmentation

Eric Guérin
Killian Oechslin
  • Fonction : Auteur
Benoît Martinez
  • Fonction : Auteur

Résumé

In this paper, we propose a method for the automatic semantic segmentation of satellite images into six classes (sparse forest, dense forest, moor, herbaceous formation, building, and road). We rely on Swin Transformer architecture and build the dataset from IGN open data. We report quantitative and qualitative segmentation results on this dataset and discuss strengths and limitations. The dataset and the trained model are made publicly available.

Dates et versions

hal-03375737 , version 1 (13-10-2021)

Identifiants

Citer

Eric Guérin, Killian Oechslin, Christian Wolf, Benoît Martinez. Satellite Image Semantic Segmentation. 2021. ⟨hal-03375737⟩
79 Consultations
0 Téléchargements

Altmetric

Partager

More