FDWD-Net: Feature-Decoupled and Window-Differentiated Network for Remote Sensing Image Super-Resolution
Open Access
- 25 April 2026
- journal article
- research article
- Published by MDPI AG in Remote Sensing
- Vol. 18 (9) , 1316
Abstract
Super-resolution reconstruction of remote sensing images has significant application value in fields such as smart cities, land monitoring, and traffic management. However, current super-resolution methods often overlook the differences between semantic and texture feature representations. This limitation makes it difficult to collaboratively preserve semantic structures and fine details during reconstruction, thereby affecting overall reconstruction quality. To address these challenges, this paper proposes a novel remote sensing image super-resolution network based on feature decoupling and differential window design, termed FDWD-Net. Specifically, we introduce an Adaptive Energy-driven Channel Selection module and a Multi-Directional Gradient-based Semantic–Texture Decoupling module to identify informative channels from the feature maps and decouple them into semantic and texture representations for independent optimization. Furthermore, we design a Differential Window-based Cross-scale Interaction module that dynamically adjusts window sizes based on local texture complexity, enabling adaptive feature modeling and effective multi-scale information interaction. Experimental results confirm that our method surpasses existing mainstream models on several remote sensing datasets. It also performs better in preserving structures and restoring detailed information.Keywords
Funding Information
- National Natural Science Foundation of China (No. 62577044)
- Natural Science Basic Research Program of Shaanxi Province (No. 2025JC-YBQN-906)
- Key Projects for Research and Development of Shaanxi Province (No. 2025CY-YBXM-067)
- Education Department of Shaanxi Province (No. 24JK0651)
- Key Scientific Research Projects of the Shaanxi Provincial Education Department (No. 24JR152)
This publication has 35 references indexed in Scilit:
- High-frequency structure transformer for magnetic resonance image super-resolutionPattern Recognition, 2025
- CMANet: A CNN-Mamba aggregation network for face super-resolutionPattern Recognition, 2025
- Tensor Transformer for hyperspectral image classificationPattern Recognition, 2025
- Ship target detection in SAR images based on SimAM attention YOLOv8IET Communications, 2024
- Fully 1 × 1 Convolutional Network for Lightweight Image Super-resolutionMachine Intelligence Research, 2024
- SRConvNet: A Transformer-Style ConvNet for Lightweight Image Super-ResolutionInternational Journal of Computer Vision, 2024
- AFCN: An attention‐directed feature‐fusion ConvNeXt network for low‐voltage apparatus assembly quality inspectionIET Image Processing, 2024
- Activating More Pixels in Image Super-Resolution Transformer2023
- Pre-Trained Image Processing Transformer2021
- Accurate Image Super-Resolution Using Very Deep Convolutional Networks2016