Multilayer Degradation Representation-Guided Blind Super-Resolution for Remote Sensing Images

Xudong Kang,Jier Li,Puhong Duan,Fuyan Ma,Shutao Li

DOI: https://doi.org/10.1109/tgrs.2022.3192680

IF: 8.2

2022-01-01

IEEE Transactions on Geoscience and Remote Sensing

Abstract:Remote sensing image super-resolution (SR) aims to boost the image resolution while recovering rich high-frequency details. Currently, most of the SR methods are based on an assumption that the degradation kernel is a specific downsampler. However, the degradation kernel is unknown and sophisticated for real remote sensing scenes, leading to a severe performance drop. To alleviate this problem, we propose a multilayer degradation representation-guided blind SR method for remote sensing images, which mainly consists of three key steps. First, an unsupervised representation learning is exploited to learn the degradation representation from low-resolution images. Then, a degradation-guided deep residual module is designed to model high-order features across different scales from the original images. Finally, a multilayer degradation-aware feature fusion mechanism is proposed to restore the finer details. Experiments on synthetic and real datasets demonstrate that the proposed method can achieve promising performance with respect to other state-of-the-art SR approaches.