Scribd
Upload
38 views2 pages

Automatic Aperture Selection For Astronomical Image Photometry

The document discusses the author's research project on automatic aperture selection for astronomical image photometry. The goal is to develop a method for computers to identify objects in telescope images to help astronomers with their analysis. Currently, programs used to detect objects and select boundaries (apertures) around them are not entirely adequate and require tweaking. The author aims to first review current aperture selection methods and then identify promising techniques that could be combined to improve automatic detection performance.

Uploaded by

Ruby Wright
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
38 views2 pages

Automatic Aperture Selection For Astronomical Image Photometry

The document discusses the author's research project on automatic aperture selection for astronomical image photometry. The goal is to develop a method for computers to identify objects in telescope images to help astronomers with their analysis. Currently, programs used to detect objects and select boundaries (apertures) around them are not entirely adequate and require tweaking. The author aims to first review current aperture selection methods and then identify promising techniques that could be combined to improve automatic detection performance.

Uploaded by

Ruby Wright
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 2

AUTOMATIC APERTURE SELECTION FOR

ASTRONOMICAL IMAGE PHOTOMETRY

The topic of my research project is the automated selection of apertures for


astronomical image photometry
Or in less fancy words getting a computer to circle things in telescope images so
people can do science on them
Ill start out by showing the context of where this problem lies, and why it is important
that it be addressed
Ive been assigned to the GAMA (Galaxy and Mass Assembly) team: data is taken
from sky surveys (some public surveys e.g. SDSS, and some GAMA-run campaigns)
And from this data, the science objective is to study the formation and assembly of
galaxies and their structure
But due to the shear volume of data coming down from these telescopes and surveys,
it is not feasible for astronomers to manually analyse each and every image
As such, the goal is to at least partially automate the process of analysing these
images: going from raw data to a catalogue of objects with known position and
characteristics
Ive outlined this process here:
o So this first step, source extraction, is where we get the image and try to

identify all the objects that are present (the objects being represented by an
array of pixels called an aperture)
Its important to note that the following steps rely heavily on collecting

accurate apertures initially a bad aperture makes for bogus analysis


Photometry is where the science happens: the apertures identified in source
detection are analysed for their spectra and flux to extract some of their
properties
Astrometry is simply locating the object in the sky

o
o The objects location and classification form the object catalogue from the raw
data

The step Im working on improving is the first one: source detection

The problem faced is that programs used to detect and select apertures for sources
are not entirely adequate, and require a lot of tweaking by astronomers before they
can perform the photometry
So, for an example, these are a couple of images that show the apertures generated
by a program called SourceExtractor
o In this first one we can see that due to the ambient light from other

sources, this central objects aperture has been extended far too large

o And the light from this very bright source in the bottom right has caused
the program to identify several objects when there is only one present

o And in this image on the right we are presented with many false positives
(apertures containing no objects) due to the ambient noise fluctuation in
the background of the image

My objectives are to first review current methods of automated aperture selection


And to subsequently identify some techniques of automated aperture selection that
are promising, and could be utilised or combined to improve performance

You might also like