An advanced features extraction module for remote sensing image super-resolution

Name: Sultan, Naveed

Organization : Chulalongkorn University. Department of Electrical Engineering

Email : 6570132321@student.chula.ac.th

Name: Hajian, Amir

Organization : Chulalongkorn University. Department of Electrical Engineering

Email : amirhajian85@gmail.com

Name: Supavadee Aramvith

Organization : Chulalongkorn University. Department of Electrical Engineering. Multimedia Data Analytics and Processing Research Unit

Email : supavadee.a@chula.ac.th

keyword: Image enhancement

LCSH: Remote sensing

LCSH: Image processing -- Digital techniques

LCSH: High resolution imaging

Abstract: In recent years, convolutional neural networks (CNNs) have achieved remarkable advancement in remote sensing image super-resolution due to the complexity and variability of textures and structures in remote sensing images (RSIs), which often repeat within the same images but differ across others. Current deep learning-based super-resolution models focus less on high-frequency features, leading to suboptimal performance in capturing contours, textures, and spatial information. State-of-the-art CNN-based methods now focus on feature extraction from RSIs using attention mechanisms. However, these methods are still incapable of effectively identifying and utilizing key content attention signals in RSIs. To solve this problem, this paper proposes an advanced feature extraction module called Channel and Spatial Attention Feature Extraction (CSA-FE) for effectively extracting features using channel and spatial attention incorporated with the standard vision transformer (ViT). The proposed method was trained on the UCMerced dataset at scales of 2, 3, and 4. The experimental results show that the proposed method helps the model focus on specific channels and spatial locations containing high-frequency information, allowing the model to focus on relevant features and suppress irrelevant ones, thereby enhancing the quality of super-resolved images. The model achieved superior performance compared with various existing models.

King Mongkut's University of Technology North Bangkok. Central Library

Address: BANGKOK

Email: library@kmutnb.ac.th

Created: 2024

Modified: 2026-06-02

Issued: 2026-06-02

บทความ/Article

application/pdf

BibliograpyCitation : In Electrical Engineering/Electronics, Computer, Telecommunications and Information Technology (ECTI) Association, Thailand. 2024 21th International Conference on Electrical Engineering/Electronics, Computer, Telecommunications and Information Technology (ECTI-CON 2024) (pp.138-143). Piscataway, NJ : IEEE

eng

©copyrights King Mongkut's University of Technology North Bangkok

ใช้เวลา

0.019976 วินาที