A simplified Python implementation of the SAMSTS algorithm for cloud-filling Landsat8 http://www.mdpi.com/2072-4292/10/4/609
Requires Python 2.x
pip install -r requirements.txt
from samsts import run job_name = 'basemap-nj-p14-r32' # which landsat wrs area path = '14' row = '32' # pre-selected acquisition date of images # you want to cloud fill scenes_to_fill = [ "20171107", "20180621" ] run.run( job_name, path, row, scenes_to_fill )
- A "stack" is a full stack of landsat scenes for multiple dates
- A "scene" is all the landsat bands from a particular date. The scene is often referred to by the date string in the format "YYYYMMDD"
- A "cluster" is a point in high-dimensional space that is useful for grouping segments by similarity
- "idxs" often refers to a 2D numpy array in the format [[r,r,r,...], [c,c,c,...]] used to specify the row/col indices of pixels in the stack