This project is a part of Encarna Medina-Lopez's work at the University of Edinburgh. It is designed to interpret GeoTIFF files from the Sentinel-2 MultiSpectral Instrument which gathers satellite data from across the globe. More information about this data can be found here: https://developers.google.com/earth-engine/datasets/catalog/COPERNICUS_S2.
The scripts runProcessing.py and gdalClasses.py work to load and read the data in from a Sentinel-2 GeoTIFF file, identify the presence of clouds, and use this information to modify another GeoTIFF file.
Follow instructions at: https://gdal.org/download.html
Follow instruction at: https://matplotlib.org/users/installing.html
Follow instructions at: https://scipy.org/install.html
Follow instructions at: https://pandas.pydata.org/getting_started.html
Follow instructions at: https://pypi.org/project/mplcursors/
Download the appropriate libraries and scripts from this repository.
Open runProcessing.py in a text editor. In the function chooseFile(), change the file names to match their location on your machine.
You need two files to run these scripts. The raw file from Sentinel-2 should be called original_file, and the processed file from Medina-Lopez's neural net should be called processed_file.
To change specifics about the Matplotlib visualization, changes should be made in the geoDisplay() function in runProcessing.py.