General scope of activities

Presentation

      Since the beginning of SAR imagery, the image processing and understanding team of Telecom Paris is working on the developpement of image processing tools dedicated to these particular data. The following research axes are developed :
      •  Speckle reduction and enhancement of SAR data (deep learning methods, non local approaches, dictionnary based approaches, TV regularization, markovian modeling and adapted optimization tools)
      •  Multi-temporal analysis (deep learning methods, non local approaches, statistical tests, local descriptors)
      • Understanding and information extraction (statistical modeling, phase vortex, graphs of primitives, stable points and corner reflectors, deep learning methods)
      • 3D reconstruction (interferometric and tomographic approaches with regularization priors)
      • Fusion of SAR and optical data (SVM, statistical methods)
      More recently the group is evolving towards multi-spectral and hyperspectral imaging.



      Scientific projects

      Ongoing / previous projects :
      • ASTRAL ANR ASTRID DGA project for the development of deep learning methods dedicated to SAR imaging and applications (2022-2026)
      • Futur & Rupture funding on deep learning methods for forest monitoring using tomographic data (2021-2024)
      • Futur & Rupture funding on deep learning methods for remote sensing image understanding (2020-2023)
      • Exploiting spectral information or SAR time series processing (CNES, 2020-2022) -with MAP5-
      • Multi-temp-BIOMASS project for multitemporal speckle reduction for coming BIOMASS mission (CNES, 2017-2019)
      • SWOT-ADT for water surface detection for coming SWOT mission (CNES, 2018-2022)
      • ALYS ANR ASTRID DGA project for SAR tomography in urban areas (2016-2020) ALYS
      • MIRIAM ANR project for multi-images exploitation (2014-2018)
      • CDS funding (Center for Data Science), funded by the IDEX Paris-Saclay, ANR-11-IDEX-0003-02 (2015)
      • G4M project for underground network detection using GPR (Ground Penetrating Radar (2013-2017)
      • Futur & Rupture and CNES project on surface water and hydrological network extraction with KaRIn SWOT data (2014-2017)
      • CEA/DGA project on multi-sensor SAR registration and analysis (2011-2015)
      • PEPS project BAYARAS on asymetric Bayesian estimators for SAR image analysis (PI M. Sigelle, with J. Darbon) (2013-2014)
      • SWOT on line detection for hydrological network extraction (2010-2013)
      • DGA-REI project on SAR imagery regularization (2009-2011)
      • ANR EFIDIR on ground movement monitoring (2007-2011)
      • CMCU project on multi-temporal SAR image analysis (2008-2011)



      Collaborations


      Academic collaborations:
      • Laboratoire Hubert Curien, Telecom Saint-Etienne (Loic denis)
      • IMD, Bordeaux (Charles Deledalle)
      • MAP5, Paris 5 (Remy Abergel, Andres Almansa)
      • Cedric, CNAM (Clément Rambour)
      • CEA (Beatrice Puyssegur)
      • PolyTech Savoie LISTIC (Emmanuel Trouve)
      • Wuhan University, China (Hong Sun)
      • SupCom Tunis (Ferdaous Chaabane)
      • University of Naples (Parthenope) (Giampaolo Ferraioli, Vito Pascazio)
      • University of Sao Paulo, Brazil, (Roberto Cesar, Roberto Hirata, Talita Perciano)
      • University of Sao Carlos, Brazil, (Nelson Mascarenhas, Michelle Horta)

      Industrial or state partners:
      • CNES
      • ONERA
      • Magellium
      • Thales
      • C-S