Source Detecting

I came back to my source detection code this morning hoping to improve its detection of the obvious sources in my image and reduce the detection of random things that don’t look like sources at all. I quickly realized the issue was with the threshold being only a couple times the standard deviation from background subtraction. Since this was a Hubble image, the std for the background was a few thousandths of a count, so I needed a threshold at least a hundred times that to eliminate many of the minuscule sources. I also needed to lower the FWHM as the one big source in my image didn’t take up a ton of pixel space and so my FWHM of 10 disqualified it as a source. I played around with these parameters and found some winning combinations. I then took a bigger chunk of the original image so I would have more bright sources to try to identify. I tried with a threshold of 300 times the std of the background and a fwhm of 4. Small section of Hubble image with sources identified with blue circles. FWHM of 4

This source detection does a good job identifying most of the sources in the image, however, it identifies some non-sources around the diffraction spikes of the brightest source. in the bottom.

I tried again raising the fwhm to 8 Small section of Hubble image with sources identified with blue circles. FWHM of 8

This cuts out those extraneous sources, but misses some important smaller sources elsewhere in the image. This difficulty underscores the need for PSF photometry to get accurate source detection and photometric results.

I briefly was looking into a way to try to number these circles so that in the next step it would be easy to tell which apertures in the table came from which sources on the figure. I couldn’t figure it out today, but perhaps it will come in the future. For now, this code works well to produce decent photometry results(this one is for fwhm of 4)

These image links didn’t work right so i just turned them into normal links.