search

NCAR / ucomp-pipeline

Data processing pipeline for UCoMP
Other
6 stars 3 forks source link

Align level 1 images to each other using coronal structures #182

Open jburkepile opened 1 year ago

jburkepile commented 1 year ago

The corona moves over the day under the occulter. This is clear when you look at daily movies. Align the coronal structures between the Level 1 images in order to stabilize the images over the day. Can the image alignment code in the waves program be automated to accomplish this? If not, should we align the coronal structures by applying a coordinate transformation to the images to the 'radius-position angle' system in order to apply an Fourier transform? This removes the problem with having the occulter in the images.

[!NOTE] Time estimate: This is a research project without a definitive implementation, so it is difficult to estimate how long it will take. In any likely scenario, 40+ hours.

mgalloy commented 11 months ago

The centering-offsets branch has code to use predetermined offsets instead of finding its own.

mgalloy commented 11 months ago

The 5 largest differences in centering offsets between the pipeline and Vanessa's code:

20220901.211422.ucomp.1074.l1.3.fts -> 19.518152
20220901.211455.ucomp.1074.l1.3.fts -> 17.655525
20220902.024313.ucomp.1074.l1.3.fts -> 6.134364
20220902.032124.ucomp.1074.l1.3.fts -> 6.111406
20220901.204012.ucomp.1074.l1.3.fts -> 5.574274