Re: FWHM on line selection?

Posted by Michael Schmid on
URL: http://imagej.273.s1.nabble.com/FWHM-on-line-selection-tp5004777p5004782.html

Hi Stanislav,

if you have only a single peak in your line scan, the built-in curve
fitter of ImageJ should work.

Either manually:
In the profile plot, type 'copy'
Open Analyze>Tools>Curve Fitting
Paste the data into the data area
Select 'Gaussian' or 'Gaussian (no offset)'
Press 'Fit'

Or with a macro - this one assumes you have a plot already, and the plot
is in the foreground:
  Plot.getValues(xpoints, ypoints);
  Fit.doFit("Gaussian", xpoints, ypoints);
  print("sigma="+Fit.p(3));

Note that the built-in curve fitting assumes a peak, not a depression,
i.e., it will search for a Gaussian fit with the peak somewhere near the
maximum of the curve.

Michael
______________________________________________


> -----Original Message-----
> From: ImageJ Interest Group [mailto:[hidden email]] On Behalf Of
> Vitha, Stanislav


> Sent: Friday, September 13, 2013 10:22 AM
> To: [hidden email]
> Subject: FWHM on line selection?
>
> I would like to  calculate FWHM from a line selection to measure
> microscope resolution.
> My sample contains cytoskeletal filaments (similar to microtubules). I
> draw a line selection across the individual filament  and plot the
> intensity profile. From that profile I can eyeball the full width at half
> maximum (FWHM), but I would prefer an actual calculation.
> I am aware of the FWHM plugin that can calculate FWHM of a spot in X and Y
> direction,  but since my filaments are running in random directions, I do
> not think I can use it.
>
> Suggestions and hints would be appreciated.
> Thanks!
>
> Stan Vitha
> Microscopy and Imaging Center
> Texas A&M University

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