Predicting Landsat Reflectance with Deep Generative Fusion

TL;DR

This paper explores the use of deep generative models to fuse satellite data with different resolutions, enabling the creation of high-resolution imagery from coarser, more frequent data to improve environmental monitoring.

Contribution

It introduces a novel dataset and demonstrates a deep generative fusion model that outperforms existing methods in producing high-resolution satellite reflectance images.

Findings

The model effectively blends MODIS and Landsat data to generate detailed reflectance images.

Deep generative fusion surpasses state-of-the-art algorithms in accuracy.

The approach enhances the temporal and spatial resolution of satellite imagery.

Abstract

Public satellite missions are commonly bound to a trade-off between spatial and temporal resolution as no single sensor provides fine-grained acquisitions with frequent coverage. This hinders their potential to assist vegetation monitoring or humanitarian actions, which require detecting rapid and detailed terrestrial surface changes. In this work, we probe the potential of deep generative models to produce high-resolution optical imagery by fusing products with different spatial and temporal characteristics. We introduce a dataset of co-registered Moderate Resolution Imaging Spectroradiometer (MODIS) and Landsat surface reflectance time series and demonstrate the ability of our generative model to blend coarse daily reflectance information into low-paced finer acquisitions. We benchmark our proposed model against state-of-the-art reflectance fusion algorithms.