Re: Analyze particles bug?

Posted by Rasband, Wayne (NIH/NIMH) [E] on Apr 27, 2015; 7:56pm
URL: http://imagej.273.s1.nabble.com/Analyze-particles-bug-tp5012617p5012629.html

On Apr 27, 2015, at 12:05 AM, Lachlan Whitehead <[hidden email]> wrote:

>
> Hi All,
>
> After running my morning update of FIJI I seem to be experiencing a bug in the Analyze Particles function which seems to provide a list of many more particles than it should.
> What it appears to be doing is detecting objects correctly, but then assigning every pixel within each object as a separate object (with an area of 0 interestingly).
> Needless to say this is causing all sorts of havoc in my analysis routines.
>
> To replicate the bug you can do the following:
> - Set Measurements to include "Limit to Threshold"*
> - Open blobs test image (or any other image)
> - Threshold and analyze particles (display results and/or add to ROI manager to see what I'm talking about).

This ImageJ 1.49r regression is fixed in the latest daily build (1.49s9). You can upgrade by using the Help>Update ImageJ command and selecting “daily build” from the drop down menu.

-wayne


>
> I've tried this on three PCs running win7, and have had a colleague check it out too.
>
> Any insights would be greatly appreciated.
>
> Thanks,
> -Lachie
>
> * Unsetting "limit to threshold" was found to be a workaround, but will be particularly annoying if I have to alter all my previous macros to unset this setting only to re-set it when performing other measurement steps



--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html