Author Topic: Making meteor shower composite  (Read 474 times)

Offline alex.vasenin

  • Newcomer
  • Posts: 12
    • View Profile
Making meteor shower composite
« on: 2018 August 13 06:10:56 »
Hello,

I want to make a composite image from several Perseids I shot yesterday. I calibrated and registered all the images, but stuck on assembling. If I integrate images with average function the stars looks nice, but it obviously rejects all the meteors. If I use maximum integration the meteors somewhat visible, but background looks horrible (very grainy). What approach should I use for this task?

Thanks,
Alex

Offline ngc1535

  • PixInsight Old Hand
  • ****
  • Posts: 311
    • View Profile
Re: Making meteor shower composite
« Reply #1 on: 2018 August 19 22:48:49 »
Hi Alex,

I suspect no one answered this question- because it is *not* a simple thing to do. You will end up, very likely, making a composite image in some form.
I had a similar task when I attempted to produce an image of satellite trails that were running through exposures of the Orion Nebula (most of them are Geostationary and at my latitude, they run right through the nebula!).

If you have the patience- you can try the technique I employed. In high level language, this is what you would do in PixInsight:

1. Create a combined star-registered image. This image should have a good background (not noisy) if you combine enough frames.
2. Make certain you are rejecting all meteors and other issues.
3. Subtract this mean combined image from your individual (registered and calibrated) frames.
4. The difference images will contain your meteors. It is likely you will need to noise reduce and black clip these images to blend with your star image.
5. Blend some of the difference images with your mean combined image. There are a number of ways to do this. Max value (lighten) is one...but likely not the best. There are a number of pixel math expressions I can think of that will do a bit better.  At the time I made the image I mention above (around 8 years ago), I was not using PI- but it is no problem to do this in PI.

By the way, the technique I mention above- is known, but oft forgotten. Anyone reading... you really should take note of this little golden nugget. :)

-adam block

 To read more about the satellites, see this article:
https://www.syfy.com/syfywire/earths-equatorial-belt-of-satellites

Offline alex.vasenin

  • Newcomer
  • Posts: 12
    • View Profile
Re: Making meteor shower composite
« Reply #2 on: 2018 August 20 15:18:23 »
Hi Adam,

What an amazing image of satellite belt! WOW!  :surprised:

Thanks for the explanation, I'll definitely try it. Also I'm thinking about using LinearFit to handle background differences.

Alex