Microwave Radiometer Data Superresolution Using Image Degradation and Residual Network

Microwave radiometers are the key sensors to globally monitor environmental parameters; however, it suffers from its low and nonuniform spatial resolution. In this paper, a superresolution (SR) technique based on image degradation and residual network is proposed to enhance the spatial resolution of microwave radiometer data. Specifically, an improved degradation model is proposed to construct pairs of high-resolution (HR) and low-resolution (LR) data for training and testing. In addition, a new residual network connected by the SR main and gradient auxiliary branches in parallel is designed to achieve SR reconstructions, where eight-channel gradient maps extracted from LR data are input into the auxiliary branch to help to reconstruct. SR results are eventually generated by the trained SR network. Experiments executed on both simulated and actual data demonstrate the soundness and the superiority of the proposed SR technique.

I. Introduction

Satellite-borne microwave radiometer occupies a vital position in earth observation applications because of the ability to capture the information unavailable by optical and infrared sensors in a large field of view. Unfortunately, the poor spatial resolution of radiometer data becomes an obstacle to its application in regional-scale studies [1]. Microwave radiometer-based multi-frequency measurements are puzzled by nonuniform spatial resolutions, but data fusion among multi-frequency channels used for geophysical parameters inversion requires with it the coregistration of their spatial resolutions [1]–[3]. One obvious way to collocate different resolutions consists of averaging the higher resolution to the lower ones. This is not desirable to invert the geophysical parameters that are highly nonlinear with the radiance [4]. In addition, loss of a little information would be occurred [5], [6]. Hence, it is an essential task to reconstruct the radiometer data from a coarser resolution to a finer one.