Mark previously helped me out with an approach to unify Golgi spots by Cell, using ReassignObjectNumbers with Unify using a binary image of the slightly shrunken cell outlines and the Centroids option.
This generally works, but fails often enough to cause problems.
It would be great if there were an option to Unify-per-enclosing-object, (e.g., Unify per Cell), without needing to do this more complicated and less robust trick (which becomes a problem when cells are not round). There was a thread in 2008 suggesting that this was on the To-Do list, but hasn’t made it yet.
Any hope for this or a suggestion on how to make it work better now?
Attached are two examples of failures, with two image sets and a CP pipeline. In particular, in the first image set Cells #2 and #14 each are found to contain 2 Golgi’s instead of one. Likewise, in the second image set Cell #2 contains 2 Golgi’s instead of one.
I’m using CP 2.0 r11710 on MacOS X 10.6.8.