Closed rdrighetto closed 1 year ago
Oh, I see. Is your sample_shape
the same in all dimensions?
We have to be careful to only rotate around the tilt-axis.
Just to answer here: yes, in our case sample_shape
is always cubic and tilt axis is always Y
, but I understand the code needs to be robust to deal with cases that differ from that. Thanks for the new fix -- I will report on that soon!
cryoCARE_extract_train_data.py now runs until completion with the here proposed changes (please check if they make sense). I'm now doing a full run to see if the quality issues are fixed.