YMCA (Y-Model for Classification : ANAGEO) Jupyter Notebook

This Git contains :

Jupyter notebooks

• Preprocessing_Ymodel.ipynb : Notebook use to merge all the Sentinel-1/2 images and to clip it according to the grid and the validation/training patches.
• YMCA-kfold.ipynb : Model based on the given model of Rowel Atienza. It has been modified to reduce the overfitting and to work with the data. The pre-processing code has been modified to work with a kfolding.
• k-fold_processing.ipynb : A Jupyter notebook uses to create the kfold list.

Other folders :

• SRC : Python modules imported from https://github.com/tgrippa/Partimap.
• Modules : Python modules

Python modules in Modules folder

• CircleMaker.py : Module uses to plot a pie plot to inspect the partition of test and validation patches.
• MERGE.py : Used to merge the Sentinel bands in a multibands raster
• PlotHistories.py : Used to plot the Loss and accuracy plot of the models
• assimilation.py : Split the 4D cube created with stackage.py according to the id for training and validation.
• chunkage.py : Python homemade function that warp an image from a vectorized grid.
• stackage.py : A python homemade function that is used to create an array of raster data based on a list of path.

Background of the repository :

This repository has been created in the frame of my intership at ANAGEO. Progressively, I will add more Python projects linked to the goal of my intership.

Contact :

email : julien.govoorts@ulb.be

Name : Julien Govoorts