search

LBEM-CH / focus

High Throughput Electron Microscopy Image Processing Software
http://www.focus-em.org
GNU General Public License v2.0
35 stars 14 forks source link

Strage drift-corrected CC-Profile #104

Closed scherers closed 9 years ago

scherers commented 9 years ago

In some areas my drift-corrected CC-profile looks strange as shown in the attachment. This only happens in areas where the peaks are low anyway. Should we mask the CC-profile before running the new drift-correction step?

screenshot from 2014-12-02 13 20 32

sthennin commented 9 years ago

I think the concept of CCunbending is simply not ideal. A single particle approach is the much clearer way to go. If unbending, then it only works where the crystal distortions are weak. We could filter the QUADSERCH PROFDATA file by rejecting all peaks, where the drift of the central peak is much different than the drift of the neighbouring 4 peaks, or where the drift is too strong, or where the CCPEAK height is too low. In all those cases, we should 1) remove that spot from the PROFDATA so that that peak isn't influencing the interpolated 25x25pixel block unbending behaviour of CCUNBEND, and 2) mask the image after unbending so that that image area isn't contributing to the final FFT.