A Sentinel-2 SAFE archive data reader plugin for Mapchete.
Use the files from the example folder and download a Sentinel-2 product SAFE file. The example process takes the red, green and blue bands, enhances the colors and saves the output as a tile pyramid using PNG files.
# install rio-color before
pip install rio-color
# host an OpenLayers instance at localhost:5000 to view the output (zoom 8 or higher)
mapchete serve rgb.mapchete --memory --input_file S2A_MSIL1C_20170421T100031_N0204_R122_T33TUL_20170421T100541.SAFE.zip
# create a tile pyramid for zooms 8 to 14 in output/ directory:
mapchete execute rgb.mapchete -z 8 14 --input_file S2A_MSIL1C_20170421T100031_N0204_R122_T33TUL_20170421T100541.SAFE.zipTry out editing the process parameters in rgb.mapchete to inspect their behaviour or the python code in example_process.py. You can do this while mapchete serve is running, it will automatically reload.
Simply provide the path to the SAFE directory or SAFE.zip file as input in a mapchete process configuration to read Sentinel-2 data.
The read() function has the following flags:
- indexes : integer or list
- band number or list of band numbers
- resampling : str
- resampling method
- mask_nodata : bool
- mask out nodata (values in all bands equal 0) areas (default: True)
- mask_clouds : bool
- mask out clouds (default: False)
- mask_white_areas : bool
- mask out white (values over 4096) areas; might just work on RGB bands! (default: False)
- return_empty : bool
- returns empty array if True or raise MapcheteEmptyInputTile exception if False (default: False)
Python = 2.7 Mapchete = 0.13
Python >3 is currently not supported.
Make sure GDAL and OpenJPEG are installed. It is advised to use OpenJPEG 2.2.0 or higher to efficiently read JPEG2000 files. Also, rasterio has to be rebuilt to ensure it uses OpenJPEG instead of the standard JasPer driver:
# e.g. install rasterio version 1.0a9:
pip install --no-binary :all: rasterio==1.0a9The driver is available via pip:
pip install mapchete-safeOr clone the repository and run:
pip install -r requirements.txt
python setup.py installYou can also create a Anaconda (conda) environment via environment.yml file:
conda env create -f environment.ymlRun tests from the repository's root directory:
python setup.py testMIT License
Copyright (c) 2015, 2016, 2017 EOX IT Services