Re: Regarding astsegment clump detection

[Mohammad Akhlaghi]

Hi Chandan,

The Segment program doesn't provide semi-major and semi-minor axis 
measurements, do you mean MakeCatalog's output catalog of Segement's 
clumps?

The semi-major and semi-minor axises are calculated differently from the 
FWHM; you can see how they are calculated here:

https://www.gnu.org/software/gnuastro/manual/html_node/Measuring-elliptical-parameters.html

But with MakeCatalog, you can also measure the FWHM of clumps using the 
'--fwhm' option. But be careful with very faint clumps: because even a 
star that is too close to the noise its "Maximum" (for the half-max of 
FWHM) can be very noisy; and thus your FWHM measurements will be very 
noisy in the faint-end.

Regarding your question on the "realistic"-ness of the clumps, a 
signal-to-noise of 3 sounds pretty non-noisy. But the best way is to use 
the '--upperlimit-sigma' column/option of MakeCatalog, which will take 
the clump/object footprint and place it randomly over un-detected parts 
of the image for a large number of times (you can set the number of 
random positions to test with the '--upnum' option); it is described 
more fully in these sections of the manual:

https://www.gnu.org/software/gnuastro/manual/html_node/Upper-limit-magnitude-of-each-detection.html

https://www.gnu.org/software/gnuastro/manual/html_node/Brightness-measurements.html

Cheers,
Mohammad



On 10/19/23 14:44, chandan watts wrote:
> Hi,
>
> Thanks for your reply.
>
> Regarding PSF size, this refers to the FWHM of the PSF image of the target.
> Clump size is determined by the semimajor and semiminor axes provided by
> the astsegment output. To compare the sizes, I check the diameter of clumps
> by measuring twice the semiminor axis or semimajor axis and comparing it
> with the FWHM of the PSF image. The objective is to ensure that the clumps
> detected should have a diameter greater than the FWHM of the PSF image for
> the photometric calculations. However, astsegment is detecting most of the
> clumps with diameters less than FWHM.
> Also, I want to mention that I detected these clumps for very faint
> signals, and astsegment signal-to-noise output varies for these from ~3 -
> 8. Can these clumps be considered as realistic or not?
>
> Regards,
> Chandan