Login  Register

Re: selection

Posted by Wayne Rasband on Apr 25, 2006; 9:11pm
URL: http://imagej.273.s1.nabble.com/selection-tp3702986p3702991.html

>   Ooops! ... I was trying to make a point, and others may
>   have missed it as well.  What you are suggesting cannot be
>   automated.  What would be ideal would to create a ROI, or
>   many ROIs based on a binary image generated by a variety
>   of means.  For example, I'd at least like to create ROIs
>   based after thresholding, but there seems to be no way to
>   "select" and work with what is (or isn't) masked(?)

One way to convert a thresholded image to a set of ROIs is to check
"Record Starts" in the particle analyze and then use doWand(x,y) to
generate ROIs from the recorded coordinates. Here are some macros that
demonstrate this technique:

   http://rsb.info.nih.gov/ij/macros/CircularParticles.txt
   http://rsb.info.nih.gov/ij/macros/ConvertMaskToCompositeROI.txt
   http://rsb.info.nih.gov/ij/macros/ConvexHull.txt
   http://rsb.info.nih.gov/ij/macros/FeretsDiameter.txt
   http://rsb.info.nih.gov/ij/macros/LabelParticlesWithArea.txt
   http://rsb.info.nih.gov/ij/macros/RoiManagerAddParticles.txt

-wayne