A novel generative adversarial network framework for super-resolution reconstruction of remote sensing

IntroductionRemote sensing super-resolution (RS-SR) plays a crucial role in the analysis of remote sensing images, aiming to improve the spatial resolution of images with lower resolutions. Recent advancements in RS-SR research have been largely driven by the integration of deep learning techniques,...

Full description

Saved in:
Bibliographic Details
Main Authors: Ruilin Li, Linzhi Wen, Songtao Shao, Ming Yu, Linda Mohaisen
Format: Article
Language:English
Published: Frontiers Media S.A. 2025-05-01
Series:Frontiers in Earth Science
Subjects:
Online Access:https://www.frontiersin.org/articles/10.3389/feart.2025.1578321/full
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850148567846486016
author Ruilin Li
Linzhi Wen
Songtao Shao
Ming Yu
Linda Mohaisen
author_facet Ruilin Li
Linzhi Wen
Songtao Shao
Ming Yu
Linda Mohaisen
author_sort Ruilin Li
collection DOAJ
description IntroductionRemote sensing super-resolution (RS-SR) plays a crucial role in the analysis of remote sensing images, aiming to improve the spatial resolution of images with lower resolutions. Recent advancements in RS-SR research have been largely driven by the integration of deep learning techniques, especially through the application of Generative Adversarial Networks (GANs), which have shown significant effectiveness in advancing this field. While GAN has achieved notable advancements in this field, its tendency toward pattern collapse often introduces artifacts and distorts textures in the reconstructed images.MethodsThis study introduces a novel RS-SR model, termed the Diffusion Enhanced Generative Adversarial Network (DEGAN), designed to improve the quality of RS-SR images through the incorporation of a diffusion model. At the heart of DEGAN lies an innovative GAN architecture that fuses the adversarial mechanisms of both the generator and discriminator with an integrated diffusion module. This additional component utilizes the noise reduction capabilities of the diffusion process to refine the intermediate stages of image generation, ultimately improving the clarity of the final output and enhancing the performance of remote sensing super-resolution.ResultsIn the test dataset, the peak signal-to-noise ratio (PSNR) increased by 0.345 dB at 2× scaling and 0.671 dB at 4× scaling, while the structural similarity index (SSIM) was improved by 0.0087 and 0.0166, respectively, compared to the current state-of-the-art (SOTA) approach.DiscussionThese results indicate that DEGAN significantly improves the super-resolution reconstruction performance of remote sensing images. The introduction of the diffusion module and attention mechanism effectively reduces noise and enhances image clarity, addressing common issues of texture distortion and artifacts in remote sensing image super-resolution reconstruction.
format Article
id doaj-art-3670d3ab68244b6fb3ea47dfbc71e93d
institution OA Journals
issn 2296-6463
language English
publishDate 2025-05-01
publisher Frontiers Media S.A.
record_format Article
series Frontiers in Earth Science
spelling doaj-art-3670d3ab68244b6fb3ea47dfbc71e93d2025-08-20T02:27:13ZengFrontiers Media S.A.Frontiers in Earth Science2296-64632025-05-011310.3389/feart.2025.15783211578321A novel generative adversarial network framework for super-resolution reconstruction of remote sensingRuilin Li0Linzhi Wen1Songtao Shao2Ming Yu3Linda Mohaisen4College of Computer and Control Engineering, Northeast Forestry University, Harbin, ChinaCollege of Computer and Control Engineering, Northeast Forestry University, Harbin, ChinaCollege of Computer and Control Engineering, Northeast Forestry University, Harbin, ChinaCollege of Computer and Control Engineering, Northeast Forestry University, Harbin, ChinaDepartment of Information Technology, Faculty of Computing and Information Technology, King Abdulaziz University, Jeddah, Saudi ArabiaIntroductionRemote sensing super-resolution (RS-SR) plays a crucial role in the analysis of remote sensing images, aiming to improve the spatial resolution of images with lower resolutions. Recent advancements in RS-SR research have been largely driven by the integration of deep learning techniques, especially through the application of Generative Adversarial Networks (GANs), which have shown significant effectiveness in advancing this field. While GAN has achieved notable advancements in this field, its tendency toward pattern collapse often introduces artifacts and distorts textures in the reconstructed images.MethodsThis study introduces a novel RS-SR model, termed the Diffusion Enhanced Generative Adversarial Network (DEGAN), designed to improve the quality of RS-SR images through the incorporation of a diffusion model. At the heart of DEGAN lies an innovative GAN architecture that fuses the adversarial mechanisms of both the generator and discriminator with an integrated diffusion module. This additional component utilizes the noise reduction capabilities of the diffusion process to refine the intermediate stages of image generation, ultimately improving the clarity of the final output and enhancing the performance of remote sensing super-resolution.ResultsIn the test dataset, the peak signal-to-noise ratio (PSNR) increased by 0.345 dB at 2× scaling and 0.671 dB at 4× scaling, while the structural similarity index (SSIM) was improved by 0.0087 and 0.0166, respectively, compared to the current state-of-the-art (SOTA) approach.DiscussionThese results indicate that DEGAN significantly improves the super-resolution reconstruction performance of remote sensing images. The introduction of the diffusion module and attention mechanism effectively reduces noise and enhances image clarity, addressing common issues of texture distortion and artifacts in remote sensing image super-resolution reconstruction.https://www.frontiersin.org/articles/10.3389/feart.2025.1578321/fullsuper-resolutiondiffusion modelgenerative adversarial networkremote sensing image reconstructionattention mechanism
spellingShingle Ruilin Li
Linzhi Wen
Songtao Shao
Ming Yu
Linda Mohaisen
A novel generative adversarial network framework for super-resolution reconstruction of remote sensing
Frontiers in Earth Science
super-resolution
diffusion model
generative adversarial network
remote sensing image reconstruction
attention mechanism
title A novel generative adversarial network framework for super-resolution reconstruction of remote sensing
title_full A novel generative adversarial network framework for super-resolution reconstruction of remote sensing
title_fullStr A novel generative adversarial network framework for super-resolution reconstruction of remote sensing
title_full_unstemmed A novel generative adversarial network framework for super-resolution reconstruction of remote sensing
title_short A novel generative adversarial network framework for super-resolution reconstruction of remote sensing
title_sort novel generative adversarial network framework for super resolution reconstruction of remote sensing
topic super-resolution
diffusion model
generative adversarial network
remote sensing image reconstruction
attention mechanism
url https://www.frontiersin.org/articles/10.3389/feart.2025.1578321/full
work_keys_str_mv AT ruilinli anovelgenerativeadversarialnetworkframeworkforsuperresolutionreconstructionofremotesensing
AT linzhiwen anovelgenerativeadversarialnetworkframeworkforsuperresolutionreconstructionofremotesensing
AT songtaoshao anovelgenerativeadversarialnetworkframeworkforsuperresolutionreconstructionofremotesensing
AT mingyu anovelgenerativeadversarialnetworkframeworkforsuperresolutionreconstructionofremotesensing
AT lindamohaisen anovelgenerativeadversarialnetworkframeworkforsuperresolutionreconstructionofremotesensing
AT ruilinli novelgenerativeadversarialnetworkframeworkforsuperresolutionreconstructionofremotesensing
AT linzhiwen novelgenerativeadversarialnetworkframeworkforsuperresolutionreconstructionofremotesensing
AT songtaoshao novelgenerativeadversarialnetworkframeworkforsuperresolutionreconstructionofremotesensing
AT mingyu novelgenerativeadversarialnetworkframeworkforsuperresolutionreconstructionofremotesensing
AT lindamohaisen novelgenerativeadversarialnetworkframeworkforsuperresolutionreconstructionofremotesensing