Braaten-Cohen-Yang cloud detection

[firzaariany]

Summary

This PR adds a new cloud detection notebook based on Braaten-Cohen-Yang decision tree and contributes to isseus #9

What's changed

• cloud_and_atmospheric_processing/cohen_braaten_yang_cloud_detection.ipynb: Notebook for detecting cloud using openEO processes. This is based on Braaten-Cohen-Yang.
• The script uses simple threshold tests on the green band and NDGR. To reduce snow misclassification, an additional SWIR constraint is applied.
• Thick clouds are rendered in blue
• Thin clouds are rendered in red

How to run cohen_braaten_yang_cloud_detection.ipynb

• Execute cells from top to bottom
• The notebook will connect to CDSE openEO backend, process Sentinel-2 data, and generate a cloud detection tif image, saved as cby_cloud_detection_image.tif

Validation

• Evalscript result on Sentinel-2 image in Copernicus browser

• Result from openEO processes

Note: Some data appear to be unavailable after running the openEO processes on the CDSE backend. A number of pixels are rendered white, which likely indicates NoData values, even though the Copernicus Browser shows valid data at those locations.