On Tuesday 19 May 2009 10:05:12 David Webster wrote:
> Would you check the results from usintg LocalMinMax.txt. When I run it, I
> get a macro error message "No window with the title Eroded found". This
> appears to be caused by
> "run("RGB Merge...", "red='Original (red)' green='Original (green)'
> blue='Original (blue)'");".
Hi David,
Thanks for reporting this. Yes you are right, I changed this to Merge Channels
command, which solves the issue.
> Also, I am not sure the color coding of the minima and maxima is working
> right. If I run on blobs.gif with a radius of 10, the RGB image is almost
> entirely blue, even though the 3 level "Result of Maxima" shows distinct
> minim and mnaxima.
I wonder if that is because the blobs image has an inverted LUT. So what you
think is maxima is actually minima and viceversa.
To get the results one expects :-) one needs to
run("Blobs (25K)");
run("Grays");
run("Invert");
Just in case also see if you are running with IJ with white as foreground:
run("Options...", "iterations=1 black edm=Overwrite count=1");
run("Colors...", "foreground=white background=black selection=yellow");
However I found another possible bug. Flat (i.e. constant grey level) parts
of the image have pixels which are detected as both local minima and maxima
and I think this was not showing correctly.
Now the Local MinMax will return the rgb only as there are otherwise 4 phases:
min, max, both (plateau) and none. Those are shown in the new image are blue,
red, magenta, greyscale respectively.
I also included a new version of the particles8/4 plugins (which solve an
error introduced in the previous version if one was using size filtering of
particles).
The morphology zip file can be download it from here:
http://www.dentistry.bham.ac.uk/landinig/software/software.htmlCheers
Gabriel
Locked
