The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences
Download
Publications Copernicus
Download
Citation
Articles | Volume XLIII-B2-2022
Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci., XLIII-B2-2022, 1125–1132, 2022
https://doi.org/10.5194/isprs-archives-XLIII-B2-2022-1125-2022
Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci., XLIII-B2-2022, 1125–1132, 2022
https://doi.org/10.5194/isprs-archives-XLIII-B2-2022-1125-2022
 
30 May 2022
30 May 2022

CLOUDTRAN: CLOUD REMOVAL FROM MULTITEMPORAL SATELLITE IMAGES USING AXIAL TRANSFORMER NETWORKS

D. Christopoulos, V. Ntouskos, and K. Karantzalos D. Christopoulos et al.
  • Remote Sensing Lab, National Technical University of Athens, Greece

Keywords: cloud-free, cloud detection, reconstruction, reflectance, time series, autoregressive models, sentinel-2

Abstract. We present a method for cloud-removal from satellite images using axial transformer networks. The method considers a set of multitemporal images in a given region of interest together with the corresponding cloud masks, and delivers a cloud-free image for a specific day of the year. We propose the combination of an encoder-decoder model employing axial attention layers for the estimation of the low-resolution cloud-free image, together with a fully parallel upsampler that reconstructs the image at full resolution. The method is compared with various baselines and state-of-the-art methods on two Sentinel-2 datasets, showing significant improvements across multiple standard metrics used for image quality assessment.