Linear Stack Alignment with SIFT
Opened this issue · 2 comments
I am in a lab that has been using the “Linear Stack Alignment with SIFT” macro in Fiji to perform image alignment of time-lapse image stacks. The alignment works well using the default parameters of the macro, using “Translation” as the expected transformation.
The scale of the image stacks reverts to pixels after the alignment process. From looking at the Log window, there is an affine transform that appears to be run on the image. I am assuming that this transform does not preserve euclidean distances.
Is this why the pixel dimensions are lost?
I was told in the image.sc forum that the lack of preserve pixel dimensions was a bug and not a consequence of the algorithm. I'm not so sure.
Is this the place to ask this question?
The px dimensions remain the same. Particularly for translation models, there is no doubt. This is just an oversight by me when I wrote this plugin, I was young and stupid. You can add copying the existing parameters into the result in a recorded macro.