|
|
I want to align a stack of high resolution electron microscope images
that show atoms as black dots and the change of atom positions at an
interface. There is a substantial drift (both a constand drift and
some random one).
If I use Register ROI, Register ROI MI or Register MI Edges, I get an
aligned stack with several wrongly aligned images (jumps).
The main cause of the problem is the periodic lattice of the atoms in
relation to the procedure that all alignments are done to the first
image. (If I skip the first image the jumps occurs at other fromes in
the movie).
Thus it is better to align each image to the previous image. This can
lead to new problems if an image is very poor of if an image
alingment is done by the TEM operator during the recording of the
image stack. The fist problem can be overcome by removal of this
image. The second could be overcome by a procedure (by hand) to
identify these big position changes and add these shifts and image
numbers as input in the alignement procedure.
Has anybody developed tools for such alignments?
henny
--
--------------------------------
Henny Zandbergen, tel 31-15-2782266, fax 31-15-2786730
|
Locked
