Despeckling Sentinel-1 GRD images by deep learning and application to narrow river segmentation - Equipe Image, Modélisation, Analyse, GEométrie, Synthèse Accéder directement au contenu
Communication Dans Un Congrès Année : 2021

Despeckling Sentinel-1 GRD images by deep learning and application to narrow river segmentation

Résumé

This paper presents a despeckling method for Sentinel-1 GRD images based on the recently proposed framework "SAR2SAR": a self-supervised training strategy. Training the deep neural network on collections of Sentinel 1 GRD images leads to a despeckling algorithm that is robust to space-variant spatial correlations of speckle. Despeckled images improve the detection of structures like narrow rivers. We apply a detector based on exogenous information and a linear features detector and show that rivers are better segmented when the processing chain is applied to images pre-processed by our despeckling neural network.

Dates et versions

hal-03129006 , version 1 (02-02-2021)

Identifiants

Citer

Nicolas Gasnier, Emanuele Dalsasso, Loïc Denis, Florence Tupin. Despeckling Sentinel-1 GRD images by deep learning and application to narrow river segmentation. IGARSS 2021, Jul 2021, Bruxelles, Belgium. ⟨hal-03129006⟩
63 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More