Trustworthy Super-Resolution of Multispectral Sentinel-2 Imagery With Latent Diffusion

Remote sensing super-resolution aims to enhance the spatial details of satellite images by introducing meaningful high-frequency features while avoiding hallucinations and spectral distortions. High-resolution imagery is usually not publicly available, whereas low-resolution imagery is freely availa...

Full description

Saved in:
Bibliographic Details
Main Authors: Simon Donike, Cesar Aybar, Luis Gomez-Chova, Freddie Kalaitzis
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Subjects:
Online Access:https://ieeexplore.ieee.org/document/10887321/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850023215865266176
author Simon Donike
Cesar Aybar
Luis Gomez-Chova
Freddie Kalaitzis
author_facet Simon Donike
Cesar Aybar
Luis Gomez-Chova
Freddie Kalaitzis
author_sort Simon Donike
collection DOAJ
description Remote sensing super-resolution aims to enhance the spatial details of satellite images by introducing meaningful high-frequency features while avoiding hallucinations and spectral distortions. High-resolution imagery is usually not publicly available, whereas low-resolution imagery is freely available with a much higher revisit rate, such as the Sentinel-2 multispectral imaging mission. Cross-sensor super-resolution has the potential to bridge this gap, providing high spatial and temporal resolution imagery which are otherwise unavailable for many remote sensing users and applications. With the recent advancements in diffusion models, many methodologies have emerged which take advantage of their generative power to perform super-resolution. We propose an adapted latent diffusion approach, since image diffusion is computationally prohibitive to be applied to large Earth observation datasets. Contrary to standard latent diffusion, we encode the low-resolution image to condition the diffusion process, forcing better spectral consistency with the input imagery. The model includes visible and near-infrared bands. To ensure trustworthy results, we utilize the probabilistic nature of diffusion models to generate pixel-level uncertainty maps. This confidence metric is crucial for real-world applications, such as environmental monitoring, land cover classification, and change detection, where accurate surface feature reconstruction and spectral consistency are essential. The uncertainty map allows users to evaluate the reliability of the product for these tasks. The proposed model super-resolves Sentinel-2 imagery at 10 to 2.5 m and is the first multispectral remote sensing (RS) super-resolution diffusion model efficient enough to process large-scale RS datasets, as well as the only model providing a pixelwise uncertainty metric.
format Article
id doaj-art-537dbbf0065c4db5a949e77280242360
institution DOAJ
issn 1939-1404
2151-1535
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
spelling doaj-art-537dbbf0065c4db5a949e772802423602025-08-20T03:01:27ZengIEEEIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing1939-14042151-15352025-01-01186940695210.1109/JSTARS.2025.354222010887321Trustworthy Super-Resolution of Multispectral Sentinel-2 Imagery With Latent DiffusionSimon Donike0https://orcid.org/0000-0002-4440-3835Cesar Aybar1https://orcid.org/0000-0003-2745-9535Luis Gomez-Chova2https://orcid.org/0000-0003-3924-1269Freddie Kalaitzis3https://orcid.org/0000-0002-1471-646XImage Processing Laboratory (IPL), University of Valencia, Valencia, SpainImage Processing Laboratory (IPL), University of Valencia, Valencia, SpainImage Processing Laboratory (IPL), University of Valencia, Valencia, SpainUniversity of Oxford, Oxford, U.K.Remote sensing super-resolution aims to enhance the spatial details of satellite images by introducing meaningful high-frequency features while avoiding hallucinations and spectral distortions. High-resolution imagery is usually not publicly available, whereas low-resolution imagery is freely available with a much higher revisit rate, such as the Sentinel-2 multispectral imaging mission. Cross-sensor super-resolution has the potential to bridge this gap, providing high spatial and temporal resolution imagery which are otherwise unavailable for many remote sensing users and applications. With the recent advancements in diffusion models, many methodologies have emerged which take advantage of their generative power to perform super-resolution. We propose an adapted latent diffusion approach, since image diffusion is computationally prohibitive to be applied to large Earth observation datasets. Contrary to standard latent diffusion, we encode the low-resolution image to condition the diffusion process, forcing better spectral consistency with the input imagery. The model includes visible and near-infrared bands. To ensure trustworthy results, we utilize the probabilistic nature of diffusion models to generate pixel-level uncertainty maps. This confidence metric is crucial for real-world applications, such as environmental monitoring, land cover classification, and change detection, where accurate surface feature reconstruction and spectral consistency are essential. The uncertainty map allows users to evaluate the reliability of the product for these tasks. The proposed model super-resolves Sentinel-2 imagery at 10 to 2.5 m and is the first multispectral remote sensing (RS) super-resolution diffusion model efficient enough to process large-scale RS datasets, as well as the only model providing a pixelwise uncertainty metric.https://ieeexplore.ieee.org/document/10887321/Deep learninglatent diffusionmodel uncertaintyremote sensing (RS)Sentinel-2super-resolution (SR)
spellingShingle Simon Donike
Cesar Aybar
Luis Gomez-Chova
Freddie Kalaitzis
Trustworthy Super-Resolution of Multispectral Sentinel-2 Imagery With Latent Diffusion
IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Deep learning
latent diffusion
model uncertainty
remote sensing (RS)
Sentinel-2
super-resolution (SR)
title Trustworthy Super-Resolution of Multispectral Sentinel-2 Imagery With Latent Diffusion
title_full Trustworthy Super-Resolution of Multispectral Sentinel-2 Imagery With Latent Diffusion
title_fullStr Trustworthy Super-Resolution of Multispectral Sentinel-2 Imagery With Latent Diffusion
title_full_unstemmed Trustworthy Super-Resolution of Multispectral Sentinel-2 Imagery With Latent Diffusion
title_short Trustworthy Super-Resolution of Multispectral Sentinel-2 Imagery With Latent Diffusion
title_sort trustworthy super resolution of multispectral sentinel 2 imagery with latent diffusion
topic Deep learning
latent diffusion
model uncertainty
remote sensing (RS)
Sentinel-2
super-resolution (SR)
url https://ieeexplore.ieee.org/document/10887321/
work_keys_str_mv AT simondonike trustworthysuperresolutionofmultispectralsentinel2imagerywithlatentdiffusion
AT cesaraybar trustworthysuperresolutionofmultispectralsentinel2imagerywithlatentdiffusion
AT luisgomezchova trustworthysuperresolutionofmultispectralsentinel2imagerywithlatentdiffusion
AT freddiekalaitzis trustworthysuperresolutionofmultispectralsentinel2imagerywithlatentdiffusion