@article{fdi:010093787,
  title = {{S}ub-pixel displacement estimation with deep learning : application to optical satellite images containing sharp displacements},
  author = {{M}ontagnon, {T}. and {G}iffard-{R}oisin, {S}ophie and {D}alla {M}ura, {M}. and {M}archandon, {M}. and {P}athier, {E}. and {H}ollingsworth, {J}.},
  editor = {},
  language = {{ENG}},
  abstract = {{O}ptical image correlation is a powerful method for remotely constraining ground movement from optical satellite imagery related to natural disasters (e.g., earthquakes, volcanoes, landslides). {T}his approach enables the characterization, and identification of the causal factors and mechanisms underlying such processes. {B}y employing sub?pixel correlation algorithms, one can obtain highly accurate (m?to?cm level) displacement fields at high spatial resolution (dm?to?cm) by comparing satellite images acquired before and after a period of movement. {H}owever, this method generally assumes a homogeneous translation of all pixels within a given correlation window, which will lead to biased estimates of ground displacement if the real case is not well represented by such a simplification, especially when resolving ground displacements next to sharp gradients in displacement, such as those found in the near?field of earthquake surface ruptures. {I}n this paper, we present an innovative deep learning method estimating sub?pixel displacement maps from optical satellite images for the retrieval of ground displacement. {F}rom the generation of a realistic simulated database, comprising {L}andsat?8 satellite image pairs containing simulated sub?pixel shifts and sharp discontinuities, we develop a {C}onvolutional {N}eural {N}etwork able to retrieve sub?pixel displacements. {T}he comparison to state?of?the?art correlation methods shows that our pipeline significantly reduces by 32% the estimation bias around fault ruptures, leading to more accurate characterization of the near?field strain in surface rupturing earthquakes. {A}pplication of our model to the 2019 {R}idgecrest earthquake demonstrates the ability of our model to accurately and quickly resolve ground displacement using real satellite images. {C}ode is made available at https://gricad-gitlab.univ-grenoblealpes.fr/montagtr/cnn4l-discontinuities},
  keywords = {},
  booktitle = {},
  journal = {{J}ournal of {G}eophysical {R}esearch : {M}achine {L}earning and {C}omputation},
  volume = {1},
  numero = {4},
  pages = {e2024{JH}000174 [25 ]},
  ISSN = {2993-5210},
  year = {2024},
  DOI = {10.1029/2024jh000174},
  URL = {https://www.documentation.ird.fr/hor/fdi:010093787},
}