compare-cloud-masks

This repository contais the code used in the analysis section of the paper entitled Comparison of cloud cover detection algorithms for land use classification of the Amazon tropical forest

Directories:

* cloud_notebook         R notebook files.
* data/masks/image_tiles Shapefiles of tiling systems of satellite imagery.
* qgis                   Qgis files including maps, virtual rasters, and layers' styles.
* scripts                Utilitary script file (bash, python, and R)

Algorithms

Fmask 4.0

* Code downloaded from this git [repostory](https://github.com/gersl/fmask)
* Algorithm description in this [paper](doi.org/10.1016/j.rse.2019.05.024)
* Run by creating a container from the docker image container in server e-sensing6:
+ Repository: fmask
+ Tag: 4.0
+ Image Id: 74de3d7c5a73
+ Virtual size 13.3 GB
* While creating a dockerfile, take into account an issue related to MATLAB's runtime and the LD_LIBRARY_PATH. See below.

MATLAB Runtine is unable to find some of the required libraries. A workaround is to replace Ubuntu's LD_LIBRARY_PATH for the one below. However, this mixes up Ubuntu's configuration. MATLAB runtime:

XAPPLRESDIR=/usr/local/MATLAB/MATLAB_Runtime/v95/X11/app-defaults
export LD_LIBRARY_PATH="/lib:/usr/lib:/usr/local/lib"
LD_LIBRARY_PATH="${LD_LIBRARY_PATH}:/usr/local/MATLAB/MATLAB_Runtime/v95/runtime/glnxa64:/usr/local/MATLAB/MATLAB_Runtime/v95/bin/glnxa64:/usr/local/MATLAB/MATLAB_Runtime/v95/sys/os/glnxa64:/usr/local/MATLAB/MATLAB_Runtime/v95/sys/opengl/lib/glnxa64"
export PATH=$PATH:$LD_LIBRARY_PATH

MAJA

* Algorithm description in this [paper](https://www.mdpi.com/2072-4292/7/3/2668)
* MAJA is ran by the Sen2Agri instances on server e-sensing6.

s2cloudless

* Algorithm desciption  in this [paper](https://medium.com/sentinel-hub/improving-cloud-detection-with-machine-learning-c09dc5d7cf13)
* Run it using the script call_s2cloudless.py

Sen2Cor

* Algorithm description in this [paper](https://www.spiedigitallibrary.org/conference-proceedings-of-spie/10427/2278218/Sen2Cor-for-Sentinel-2/10.1117/12.2278218.full)
* Run it using the docker container prepared by the Brazil Data Cube containers.
* Repository: sen2cor_2
* Tag: latest
* Image Id: ba7400c9bba6
* Virtual size 2.034 GB

brazil-data-cube / compare-cloud-masks