Releases: wri/sentinel-tree-cover
Releases · wri/sentinel-tree-cover
v1.1.0
This release includes some major updates, including:
- Reduced data throughput requirements by using preprocessed cloud probabilities
- Better removal of clouds and cloud shadows
- Improved mosaicing and tiling of prediction windows
- Improved unit test coverage
- Type hints and docstrings for the majority of code
- Swap float32 file structure to uint16 for reduced storage demand
- Swap numpy array saving to hickle to reduce storage demand
- Improve multitemporal alignment by doing a median correction between interpolated and non-interpolated regions
- Adds example data to the large-area notebooks
- Includes beta integration capabilities with the SEPAL platform
v1.0.0
Source code for "A global method to identify trees inside and outside of forests with medium-resolution satellite imagery"