SatDiffMoE: A Mixture of Estimation Method for Satellite Image Super-resolution with Latent Diffusion Models

TL;DR

SatDiffMoE is a novel diffusion-based fusion algorithm that combines multiple low-resolution satellite images over time to produce high-resolution images with finer details, improving super-resolution performance and efficiency.

Contribution

It introduces a flexible, diffusion-based method capable of fusing an arbitrary number of satellite images for enhanced super-resolution, with improved efficiency over previous approaches.

Findings

Achieves superior super-resolution performance on various datasets.

Reduces model parameters and computational costs.

Handles arbitrary numbers of input images effectively.

Abstract

During the acquisition of satellite images, there is generally a trade-off between spatial resolution and temporal resolution (acquisition frequency) due to the onboard sensors of satellite imaging systems. High-resolution satellite images are very important for land crop monitoring, urban planning, wildfire management and a variety of applications. It is a significant yet challenging task to achieve high spatial-temporal resolution in satellite imaging. With the advent of diffusion models, we can now learn strong generative priors to generate realistic satellite images with high resolution, which can be utilized to promote the super-resolution task as well. In this work, we propose a novel diffusion-based fusion algorithm called \textbf{SatDiffMoE} that can take an arbitrary number of sequential low-resolution satellite images at the same location as inputs, and fuse them into one high-resolution reconstructed image with more fine details, by leveraging and fusing the complementary information from different time points. Our algorithm is highly flexible and allows training and inference on arbitrary number of low-resolution images. Experimental results show that our proposed SatDiffMoE method not only achieves superior performance for the satellite image super-resolution tasks on a variety of datasets, but also gets an improved computational efficiency with reduced model parameters, compared with previous methods.