8
$\begingroup$

I am looking for sets of astronomical images for testing different kinds of stacking algorithms.

The idea is simple: if one has $N$ images of the same object, the signal-to-noise ratio of the averaged image increases with $\sqrt{N}$ under certain conditions.

Where could I obtain data for testing?

PS. I would propose adding the tags "image", "ccd", "research", "signal processing", and "machine learning".

$\endgroup$
3
  • $\begingroup$ I am working on an idea that needs stacking algorithms, I can provide plenty of data, however, not astronomical. Perhaps we can help each other out. alexanderpodgorny@skype $\endgroup$ Mar 21, 2016 at 23:20
  • $\begingroup$ Unfortunately the idea isn't that simple. Given a perfect camera in perfect circumstances stacking would have no effect. So the actual, real world, circumstances and cameras must be taken into account. Different types of noise will affect the image in different ways. I see several people below have suggested data from professional observatories. But, if you're looking to apply this to consumer level equipment, from crappy skies, you'll want data from consumer level equipment. $\endgroup$ Jan 27 at 22:13
  • $\begingroup$ Here's a video from Dr Robin Glover that goes into the cost/benefits of different types of noise and attempts to eliminate them on consumer level equipment. youtube.com/watch?v=3RH93UvP358 $\endgroup$ Jan 27 at 22:17

4 Answers 4

2
$\begingroup$

It's pretty easy to create a dataset for yourself using remotely controlled telescopes. I've used iTelescope in the past and had a good experience. Their "starter" account costs $20 and should be good to get you going.

$\endgroup$
3
  • 1
    $\begingroup$ Why is this being voted down? He's looking for data; generating it himself is a way to do that. It also lets him control what data is collected and how (which is instructive itself and lets this be more than an academic exercise.) $\endgroup$ Apr 28, 2015 at 15:37
  • 1
    $\begingroup$ I assume @loggy's motivation is to contribute something for the astronomer community. Why should he pay to be able to do it? Anyway, I think the answer is still valid, though maybe not best option. (I did not down-vote.) $\endgroup$
    – mmh
    Apr 29, 2015 at 12:46
  • $\begingroup$ @mmh -- thanks for the explanation. I'm not saying he should have to pay, just that there are options that let him gather exactly the data he needs, even if he doesn't have a telescope. $\endgroup$ Apr 29, 2015 at 14:51
1
$\begingroup$

You could use archival data, e.g., from the WISE mission: http://irsa.ipac.caltech.edu/applications/wise/

The data are reduced as well as photometrically and astrometrically calibrated, which should make the image combination a lot easier. Just type in random coordinates and download data for a field with the desired number of frames.

You can also try other observatories, for example, Gemini, Spitzer, ... to obtain well-calibrated test data.

$\endgroup$
1
$\begingroup$

Try the NASA data catalog, https://data.nasa.gov/data. Also Try the Keck Observatory Archive, KOA, http://nexsci.caltech.edu/archives/koa/index.shtml. Look for raw images. of clusters or galaxies.

$\endgroup$
1
$\begingroup$

Data from observatory archives is a good way to go. Here is another one with tons of imaging datasets: http://archive.eso.org

$\endgroup$

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .