Generative Adversarial Networks Capabilities for Super-Resolution Reconstruction of Weather Radar Echo Images

Abstract

Improving the resolution of degraded radar echo images of weather radar systems can aid severe weather forecasting and disaster prevention. Previous approaches to this problem include classical super-resolution (SR) algorithms such as iterative back-projection (IBP) and a recent nonlocal self-similarity sparse representation (NSSR) that exploits the data redundancy of radar echo data, etc. However, since radar echoes tend to have rich edge information and contour textures, the textural detail in the reconstructed echoes of traditional approaches is typically absent. Inspired by the recent advances of faster and deeper neural networks, especially the generative adversarial networks (GAN), which are capable of pushing SR solutions to the natural image manifold, we propose using GAN to tackle the problem of weather radar echo super-resolution to achieve better reconstruction performance (measured in peak signal-to-noise ratio (PSNR) and structural similarity index (SSIM)). Using authentic weather radar echo data, we present the experimental results and compare its reconstruction performance with the above-mentioned methods. The experimental results showed that the GAN-based method is capable of generating perceptually superior solutions while achieving higher PSNR/SSIM results.

Super-Resolution of Sentinel-2 Imagery Using Generative Adversarial Networks

“…They believed that the radiometric and geometric differences between both satellites images could be an explanation of the low-value metrics obtained. Some authors have applied ESRGAN architectures as well, to solve different remote sensing SR problem applications [47,48] This work focuses in the analysis of single-image supervised SR techniques using the DL paradigm. Specifically, we present a GAN-based super-resolution model to enhance the 10 m channels of the Sentinel-2 sensor to 2 m with a similar quality as produced by the WorldView satellite.…”

Stochastic Super-Resolution for Downscaling Time-Evolving Atmospheric Fields With a Generative Adversarial Network

“…GANs offer a natural way to model uncertainty using modern machine-learning methods, less dependent on particular statistical assumptions than the traditional methods. Regardless, the uncertainty aspect has been largely ignored in earlier attempts at improving the resolution of climate fields using deep learning even when employing GANs for this problem [16] or for other super-resolution applications related to climate or remote sensing [17]- [19] although a few studies have used GANs to represent uncertainty in other atmospheric data problems [20], [21]. Moreover, while GANs have been recently also used to model the time evolution of atmospheric fields [22], few studies using deep learning have investigated modeling the uncertainty of the generated high-resolution image in a manner consistent with the time evolution of atmospheric fields-a problem analogous to video super-resolution, which has also been studied using GANs [23], [24].…”

A Dual Network for Super-Resolution and Semantic Segmentation of Sentinel-2 Imagery

“…On the other hand, architectures based on Generative Adversarial Networks (GANs) [37], like SRGAN [38] or ESRGAN [39], have been proposed as they produce high resolution images with photo-realistic details. Models based on GANs have also been applied for the super-resolution of remote sensing imagery [10,[40][41][42].…”