*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

I have had a similar problem on our confocal system.
The UnwarpJ plugin for ImageJ  (http://bigwww.epfl.ch/thevenaz/UnwarpJ/) 
allows you to load two images of your multicolor beads, run the affine 
transform procedure to warp one of the images to get a match.

In the Advanced Options, if you click on "Save transformation" the computed 
deformation is saved at the end of the process. All the B-spline coefficients, 
and relevant information to recover this deformation is saved in a file specified 
by the user through a File Dialog. Use "Load transformation" in the I/O menu of 
the Toolbar to read this transformation and apply it to the image of a real 
sample.

Stan Vitha
Microscopy and Imaging Center
Texas A&M University

On Wed, 27 Mar 2013 06:41:49 -0500, Simon Walker 
<[log in to unmask]> wrote:

>*****
>To join, leave or search the confocal microscopy listserv, go to:
>http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
>*****
>
>Hello,
>Not a confocal question, but I know you are a knowledgeable crowd.  We are 
>in the process of setting up a screen looking at the colocalisation between 
>two subcellular compartments which appear as small red and green spots.  
>Initial tests have revealed an image registration problem where the same 
>fluorescent beads are not correctly overlaid in the red and green channels.  
>Curiously the registration off-set is not consitent across the image, so a 
>straightforward pixel-shift correction will not work.  Can anyone suggest a 
>good (preferarbly ImageJ based) method to create and subsequently apply a 
>shift-correction map so that the green and red channels correctly align?  I've 
>had a quick look at the registration options in FIJI and there seem to be 
>multiple options which might do the trick, but I don't know how any of them 
>work and which is the best solution.  Advice please!
>Thanks,
>Simon
>
>P.S. I realise that an alternative solution would be to identify and correct the 
>source of the registration problem, but I've looked at the most likely causes 
>and not been able to do this.