Accéder directement au contenu Accéder directement à la navigation
Communication dans un congrès

Cloud-GAN: Cloud Removal for Sentinel-2 Imagery Using a Cyclic Consistent Generative Adversarial Network

Abstract : Cloud cover is a serious impediment in land surface analysis from Remote Sensing images either causing complete obstruction (thick clouds) with loss of information or blurry effects when being semi-transparent (thin clouds). While thick clouds require complete pixel replacement, thin cloud removal is fairly challenging as the atmospheric and land-cover information is intertwined. In this paper, we address this problem and propose a Cloud-GAN to learn the mapping between cloudy images and cloud-free images. The adver-sarial loss in the proposed method constrains the distribution of generated images to be close enough to the underlying distribution of the non-cloudy images. An additional cycle consistency loss is used to further restrain the generator to predict cloud-free images only of the same scene as reflected in the cloudy images. Our method not only rejects the necessity of any paired (cloud/cloud-free) training dataset but also avoids the need of any additional (expensive) spectral source of information such as Synthetic Aperture Radar imagery which is cloud penetrable. Lastly, we demonstrate the efficacy of our technique by training on an openly available and fairly new Sentinel-2 Imagery dataset consisting of real clouds. We also show significant improvement in PSNR values after removing clouds on synthetic images thus validating the competency of our methodology.
Type de document :
Communication dans un congrès
Liste complète des métadonnées

Littérature citée [12 références]  Voir  Masquer  Télécharger

https://hal-enpc.archives-ouvertes.fr/hal-01832797
Contributeur : Nikos Komodakis <>
Soumis le : samedi 11 août 2018 - 16:12:48
Dernière modification le : mercredi 26 février 2020 - 19:06:07
Archivage à long terme le : : lundi 12 novembre 2018 - 12:22:12

Fichier

Template.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

  • HAL Id : hal-01832797, version 1

Collections

Citation

Praveer Singh, Nikos Komodakis. Cloud-GAN: Cloud Removal for Sentinel-2 Imagery Using a Cyclic Consistent Generative Adversarial Network. IGARSS, 2018, Valencia, Spain. ⟨hal-01832797⟩

Partager

Métriques

Consultations de la notice

216

Téléchargements de fichiers

2547