Abstract
The majority of optical observations acquired via spaceborne Earth imagery are affected by clouds. While there is numerous prior work on reconstructing cloud-covered information, previous studies are, oftentimes, confined to narrowly defined regions of interest, raising the question of whether an approach can generalize to a diverse set of observations acquired at variable cloud coverage or in different regions and seasons. We target the challenge of generalization by curating a large novel data set for training new cloud removal approaches and evaluate two recently proposed performance metrics of image quality and diversity. Our data set is the first publically available to contain a global sample of coregistered radar and optical observations, cloudy and cloud-free. Based on the observation that cloud coverage varies widely between clear skies and absolute coverage, we propose a novel model that can deal with either extreme and evaluate its performance on our proposed data set. Finally, we demonstrate the superiority of training models on real over synthetic data, underlining the need for a carefully curated data set of real observations. To facilitate future research, our data set is made available online.
Highlights
O N AVERAGE about 55% of the Earth’s land surface is covered by clouds [1], impacting the aim of missions, such as Copernicus, to reliably provide noise-free observations at a high frequency, a prerequisite for applications relying on temporally seamless monitoring of our environment, such as change detection or monitoring [2]–[5]
A major reason for these issues, which is still remaining open nowadays, is the current lack of available large-scale data sets for both training and testing of modern cloud removal approaches. We address this issue by curating and releasing a novel large-scale data set for cloud removal containing over 100 000 samples from over 100 regions of interest (ROIs) distributed over all continents and meteorological seasons of the globe
We demonstrated the declouding of optical imagery by fusing multisensory data, proposed a novel model, and released the, to the best of our knowledge, first global data set combining over a 100 000 paired cloudy, cloud-free, and coregistered synthetic aperture radar (SAR) sample triplets
Summary
O N AVERAGE about 55% of the Earth’s land surface is covered by clouds [1], impacting the aim of missions, such as Copernicus, to reliably provide noise-free observations at a high frequency, a prerequisite for applications relying on temporally seamless monitoring of our environment, such as change detection or monitoring [2]–[5]. While the aforementioned contributions share the common aim of dehazing and declouding optical imagery, the majority of methods are evaluated on narrowly defined and geospatially distinct regions of interest (ROIs) Is this specificity posing challenges for a conclusive comparison of methodology and, may cloud-removal performance on a particular ROI poorly indicate performances on other parts of the globe or at different seasons. It would be desirable for a cloud removal method to be applicable to all regions on Earth, at any season. This generalizability would allow for large-scale Earth observation without the need for costly redesigning or retraining for each individual scene that a cloud removal method is meant to be applied to
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
