Speckle reduction in SAR time series using an updating strategy, final project of the course "Remote sensing data: from sensor to large-scale geospatial data exploitation" by F.Tupin, G.Facciolo, E.Dalsasso, C.De Franchis, E.Meinhardt of the Master MVA (Mathematiques, Vision, Apprentissage)
Files :
- Test_models.ipynb contain all tests of our work and plots.
- Data_base_ratio_creation.ipynb is a notebook to create the Database of ratio images
- Dataset.py, GenerateDataset.py, model.py are files useful to create neurone network models
- models contain the pre-trained weights of models useful for our work. model.pth = SARtoSAR / model_ratio.pth = SARratiotoSARratio
- data contain the SAR images useful for our work