Abstract:Super-resolution (SR) methods have significantly contributed to the improvement of the spatial resolution of remote sensing (RS) images. The development of deep learning empowers novel methods to learn informative feature representation from massive low-resolution (LR) and high-resolution (HR) image pairs. Conventional RS image SR methods, however, may fail in large-scale (x8 and x9) SR tasks. Specifically, a larger scale factor corresponds to less information in LR images, which is a considerable challenge to SR. To address the issue, we propose a novel method for single RS image SR (SRSISR) based on stratified dense sampling to effectively extract image features. Specifically, the proposed SR dense-sampling residual attention network (SRDSRAN) combines dense sampling and residual learning to improve multilevel feature fusion and gradient propagation and employs local and global attentions to learn important features and long-range interdependence in the channel and spatial dimensions. Meanwhile, we also devise a discriminator model using local and global attentions and with the loss function integrating L(1 )pixel loss, L-1 perceptual loss, and relativistic adversarial loss to obtain the perceptually realistic images. Besides, we introduce a chain training to promote performance and expedite the training process for large-scale SR. Experimental results on UC Merced image and other multispectral data demonstrated that our SRDSRAN outperformed the current state-of-the-art methods quantitatively and in visual quality and obtained a higher classification accuracy in scene classification, proving its potential for applications with other downstream tasks. The code of SRADSGAN will be available at https://github.com/Meng-333/SRADSGAN.

Remote Sensing Image Super-Resolution Using Sparse Representation and Coupled Sparse Autoencoder

A Sparse Representation Based Super-Resolution Image Reconstruction Scheme Utilizing Dual Dictionaries

Single Remote Sensing Image Super-Resolution Via a Generative Adversarial Network with Stratified Dense Sampling and Chain Training

Remote Sensing Image Super-Resolution Based on Improved Sparse Representation

Coupled Deep Autoencoder for Single Image Super-Resolution.

Adaptive Super-Resolution for Remote Sensing Images Based on Sparse Representation with Global Joint Dictionary Model

A Super-resolution Model and Algorithm of Remote Sensing Image Based on Sparse Representation

Hyperspectral Image Super-Resolution Via Convolutional Neural Network

Remote Sensing Images Super-Resolution with Deep Convolution Networks

Super-Resolution Network for Remote Sensing Images Via Preclassification and Deep–Shallow Features Fusion

Image Super-Resolution Algorithm Based on an Improved Sparse Autoencoder

Single Image Super Resolution Based on Sparse Domain Selection

Single Remote Sensing Image Super-Resolution with an Adaptive Joint Constraint Model

Super-Resolution Based on Compressive Sensing and Structural Self-Similarity for Remote Sensing Images

A Super-Resolution Algorithm Based on Hybrid Network for Multi-Channel Remote Sensing Images

Image Super-resolution Based on Sparse Representation With Joint Constraints

Single-frame super-resolution for high resolution optical remote-sensing data products

Single Image Super-Resolution Based on Nonlocal Similarity and Sparse Representation

Single Image Super-Resolution Via Adaptive Sparse Representation and Low-Rank Constraint.

An image super-resolution scheme based on compressive sensing with PCA sparse representation

Super-resolution Method based on CS and Structural Self-similarity for Remote Sensing Images