bigearthnet-common 2.3.0

A collection of common tools to interact with the BigEarthNet dataset.

Common BigEarthNet Tools

A personal collection of common tools to interact with the BigEarthNet dataset.

This library provides a collection of high-level tools to better work with the BigEarthNet dataset.

bigearthnet_common tries to:

1. Collect the most relevant constants into a single place to reduce the time spent looking for these, like:
   • The 19 or 43 class nomenclature strings
   • URL
   • Band statistics (mean/variance) as integer and float
   • Channel names
   • etc.
2. Provide common metadata related functions
   • Safe JSON parser for S1/S2
   • Get the original split
   • Get a list of snowy/cloudy patches
   • Convert the old labels to thew new label nomenclature
   • Generate multi-hot encoded label views
3. Easily filter patches and generate subsets as CSV files

Installation

The package is available via PyPI and can be installed with:

• pip install bigearthnet_common

The package has Python-only dependencies and should cause no issues in more complex Conda environments with various binaries.

Review constants

To quickly search for BigEarthNet constants of interest, call:

• ben_constants_prompt or
• python -m bigearthnet_common.constants

Sets generator

To generate sets/subsets from the data and to store them as a CSV file, use:

• ben_build_csv_sets --help

This command-line tool let's the user easily create subsets from common constraints. To generate a CSV file that contains all Sentinel-2 patches from Serbia only durng the Summer and Spring months, call the function as:

• ben_build_csv_sets <FILE_PATH> S2 --seasons Winter --seasons Summer --countries Serbia --remove-unrecommended-dl-patches

Note: By default this tool will always remove the unrecommended patches, i.e. patches that contain seasonal snow, shadows, clouds, or that have no labels in the 19-class nomenclature

Contributing

Contributions are always welcome!

Please look at the corresponding ipynb notebook from the nbs folder to review the source code. These notebooks include extensive documentation, visualizations, and tests. The automatically generated Python files are available in the bigearthnet_common module.

More information is available in the contributing guidelines document.