Noise reduction in a SAR image using an updating strategy
Encadrants : Florence Tupin, Emanuele Dalsasso
Email : florence.tupin@telecom-paris.fr
Disponible : OUI
Spécialité : IMA
Nombre d'étudiants : 1
Description : SAR (Synthetic Aperture Radar) images have the advantage of being
acquired at any time (day or night and whatever the weather conditions).
However, they are very noisy because of the speckle noise due to
coherent imaging. The acquisition of several images and their
combination allows to reduce very efficiently the noise present on the
data.
Travail demandé : The aim of this project is to use a set of previously acquired images to
help suppressing the noise in a
newly acquired data. A first strategy will be to compute the mean of the
available dates to create a super-image and train a network to reduce
the noise. In a second step, the interest of updating (the "super-image"
corresponding to the previously denoised date) willbe investigated. In
this situation, the level of change between the two dates should be low
and may improve the noise reduction step.
Experiments will first be done in a supervised framework using the
network artchitecture proposed in [1].
Liens complémentaires : References
[1] Dalsasso, E., Yang, X., Denis, L., Tupin, F., & Yang, W. (2020). SAR
image despeckling by deep neural networks: From a pre-trained model to
an end-to-end training strategy. Remote Sensing, 12(16), 2636.
https://www.mdpi.com/2072-4292/12/16/2636/pdf
[2] Multi-temporal speckle reduction with self-supervised deep neural
networks
Inès Meraoumia, Emanuele Dalsasso, Loïc Denis, Rémy Abergel, Florence Tupin
https://arxiv.org/pdf/2207.11095
