Abstract:Single-image super-resolution (SISR) for high-resolution (HR) remote sensing image (RSI) acquisition is becoming increasingly valuable and important, and convolutional neural networks (CNNs) have produced considerable progress in this field. In RSIs, many similar geo-objects recur within the same scene, maintaining the same positions in both low resolution (LR) and HR. Based on this observation, we found that these similar geo-objects could be utilized to reconstruct texture details in LR by exploiting the consistent non-local relationships between these geo-objects in LR and HR, thereby improving the quality of SISR. Therefore, we propose a novel graph convolutional network (GCN) for SISR including a dynamic graph attention mechanism to learn the in-scale and cross-scale non-local features of RSIs. In scale, we propose a dynamic graph attention block (DGAB) that adaptively determines non-local patches upon the scene correlation derived from RSIs and further fuses patch-wise non-local information weighed by the attention scores of topological relationships and radiation characteristics in RSIs. Across different scales, we also introduce a dynamic graph attention mixing block (DGAMB) to upsample LR non-local information to HR non-local information. Most SISR methods have the upsampling blocks at the end of the network, ignoring feature extraction in high-dimensional space. To address this problem, DGAMB was designed as an upsampler in the middle of the model, enhancing the level of high-dimensional information extraction from the model. The experiments based on the WHU Building and UC Merced datasets show that our proposed method outperforms state-of-the-art methods. Our code is available at https://github.com/WenjuanZhang-aircas/NSGCN.

Exploiting local detail in single image super-resolution via hypergraph convolution

Lightweight Image Super-Resolution Network Using 3D Convolutional Neural Networks

Single-image Super-Resolution Via Selective Multi-Scale Network

Fast and Accurate Single Image Super-Resolution Via an Energy-Aware Improved Deep Residual Network.

ESKN: Enhanced Selective Kernel Network for Single Image Super-Resolution

Multi-level feature interaction image super-resolution network based on convolutional nonlinear spiking neural model

Image Super-Resolution With Cross-Scale Non-Local Attention and Exhaustive Self-Exemplars Mining

Non-Local Similarity-Based Attentive Graph Convolution Network for Remote Sensing Image Super-Resolution

Cross-scale collaborative network for single image super resolution

Diffused Convolutional Neural Network for Hyperspectral Image Super-Resolution

Multilevel Feature Exploration Network for Image Superresolution

Remote Sensing Hyperspectral Image Super-Resolution via Multidomain Spatial Information and Multiscale Spectral Information Fusion

Hyperspectral Image Super-Resolution via Dual-domain Network Based on Hybrid Convolution

Local–Global Feature Fusion Network for Efficient Hyperspectral Image Super-Resolution

Remote Sensing Image Super-Resolution Using Enriched Spatial-Channel Feature Aggregation Networks

Single-Image Super-Resolution Reconstruction based on the Differences of Neighboring Pixels

Non-Locally up-Down Convolutional Attention Network for Remote Sensing Image Super-Resolution

A New Convolutional Neural Network for Super-Resolution by Global and Local Residual

An Attention-Based Approach for Single Image Super Resolution.

A High-Frequency Focused Network for Lightweight Single Image Super-Resolution

HNSR: Highway Networks Based Deep Convolutional Neural Networks Model for Single Image Super-Resolution