image orientaion
Added by @Maria Janssen over 9 years ago
Hi Anchi,
Something weird happened in the past month where the reference image does not have the same orientation anymore as the acquired frames (there is a flip over the y-axis), and we are trying to figure out where this came from.
The command in dmsem.py "image = numpy.fliplr(numpy.rot90(image,1))" seems to only affect the orientation of the averaged image? Can I adjust the orientation of just the bright reference image somewhere?
Thanks,
Mandy
Replies (1)
RE: image orientaion - Added by Anchi Cheng over 9 years ago
Hi, Mandy,
Indeed what we set in pyscope/dmsem.py affects only the orientation of the averaged image and the bright reference image. This has been the case for DM 3.1, unfortunately.
Check the camera configuration in Digital Micrography interface for flip and rotation setup. If you have not
upgrade DM nor leginon, the change of orientation would be from someone changed the configuration there.
I put a picture in the wiki page to help you. See Gatan_K2_installation_and_setup
No, the bright reference is in general handled no different from a normal image, and thus always rotate/flip in the same way during data collection. If you want it in the orientation of the frame movies in DM 3.1, I recommend that you do that as a post processing step, and make a copy when you do so. (It will be the norm image that you want to modify since bright is not used once norm is calculated). This way it will not affect other users since Leginon will just read the file when applying the normalization to the average image.