Very Deep Super-Resolution of Remotely Sensed Images with Mean Square Error and Var-norm Estimators as Loss Functions

TL;DR

This paper introduces a very deep super-resolution method for remotely sensed images, utilizing novel loss functions and retraining on satellite and drone images to significantly enhance spatial resolution.

Contribution

The paper proposes RS-VDSR and Aero-VDSR models with a new Var-norm estimator loss function, improving super-resolution performance on remote sensing images.

Findings

RS-VDSR outperforms VDSR by up to 3.16 dB PSNR on Sentinel-2 images.

The proposed models achieve better super-resolution quality through novel loss functions.

Retraining on specific datasets enhances the models' effectiveness for remote sensing applications.

Abstract

In this work, very deep super-resolution (VDSR) method is presented for improving the spatial resolution of remotely sensed (RS) images for scale factor 4. The VDSR net is re-trained with Sentinel-2 images and with drone aero orthophoto images, thus becomes RS-VDSR and Aero-VDSR, respectively. A novel loss function, the Var-norm estimator, is proposed in the regression layer of the convolutional neural network during re-training and prediction. According to numerical and optical comparisons, the proposed nets RS-VDSR and Aero-VDSR can outperform VDSR during prediction with RS images. RS-VDSR outperforms VDSR up to 3.16 dB in terms of PSNR in Sentinel-2 images.