Re: AutoThresholding

Posted by Gabriel Landini on Dec 21, 2009; 5:46pm
URL: http://imagej.273.s1.nabble.com/AutoThresholding-tp3689968p3689969.html

On Monday 21 December 2009, you wrote:
> I have a set of similar images that I am trying to segment using
>  autothresholding so I can calculate a % area. Most of the images come back
>  with % area of 13-22% which compares favorably to my physical
>  measurements. Three of the images come back with a %area of 55%or more.  
>  I am not sure I understand why the results would be so different since the
>  images are quite similar.

Probably there is an active ROI in one of the images and the results
correspond to this ROI.
Or perhaps the images are analysed with a different method which returns a
different area being thresholded.
Post images to some site so others can see what might be wrong.

G