A new deep learning method for multispectral image time series completion using hyperspectral data

Cheick Tidiani Cissé (1) , Ahed Alboody (1) , Matthieu Puigt (1) , Gilles Roussel (1) , Vincent Vantrepotte (2) , Cédric Jamet (2, 3) , Trung-Kien Tran (2)

1 LISIC - Laboratoire d'Informatique Signal et Image de la Côte d'Opale
2 LOG - Laboratoire d’Océanologie et de Géosciences (LOG) - UMR 8187
3 ULCO - Université du Littoral Côte d'Opale

The massive development of remote sensing allowed many novel applications which bring new challenges. In particular, some applications such as marine observation require a good spatial, spectral, and temporal resolution. In order to tackle the last issue, spatio-temporal fusion of remote sensing data allows to complete a time series of multispectral images from, e.g., hyperspectral images. In this paper, we propose a new deep learning approach to that end. Our main contribution lies in the error completion task which allows to improve the completion performance. We show that our proposed method is able to produce high fidelity predictions with better quality indices than state-of-the-art methods on true images taken from the CIA / LGC database and Sentinel-2 / Sentinel-3 data.

Mots clés: Remote Sensing; Time-Series Completion; Spatio-Temporal Fusion; Deep Learning

Domaines: Traitement du signal et de l'image [eess.SP]; Apprentissage [cs.LG]

Cheick Tidiani Cissé, Ahed Alboody, Matthieu Puigt, Gilles Roussel, Vincent Vantrepotte, et al.. A new deep learning method for multispectral image time series completion using hyperspectral data. 47th IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP 2022), May 2022, Singapour, Singapore. pp.1546-1550, ⟨10.1109/ICASSP43922.2022.9747895⟩. ⟨hal-03540379⟩