SatFlow: Generative model based framework for producing High Resolution Gap Free Remote Sensing Imagery
This addresses the need for cloud-free, high-resolution remote sensing for agricultural and environmental monitoring, representing an incremental improvement in data fusion and inpainting techniques.
The authors tackled the problem of frequent, high-resolution remote sensing imagery being limited by clouds and low temporal frequency by developing SatFlow, a generative model-based framework that fuses MODIS and Landsat data to produce gap-free imagery, demonstrating reliable cloud imputation.
Frequent, high-resolution remote sensing imagery is crucial for agricultural and environmental monitoring. Satellites from the Landsat collection offer detailed imagery at 30m resolution but with lower temporal frequency, whereas missions like MODIS and VIIRS provide daily coverage at coarser resolutions. Clouds and cloud shadows contaminate about 55\% of the optical remote sensing observations, posing additional challenges. To address these challenges, we present SatFlow, a generative model-based framework that fuses low-resolution MODIS imagery and Landsat observations to produce frequent, high-resolution, gap-free surface reflectance imagery. Our model, trained via Conditional Flow Matching, demonstrates better performance in generating imagery with preserved structural and spectral integrity. Cloud imputation is treated as an image inpainting task, where the model reconstructs cloud-contaminated pixels and fills gaps caused by scan lines during inference by leveraging the learned generative processes. Experimental results demonstrate the capability of our approach in reliably imputing cloud-covered regions. This capability is crucial for downstream applications such as crop phenology tracking, environmental change detection etc.,