Author Topic: Satellite Trails Removal with Large-Scale Pixel Rejection - Processing Challenge  (Read 761 times)

Offline Kerrywnm1

  • Newcomer
  • Posts: 7
    • View Profile
Hi All,

I have been working on a set of images trying to eliminate the satellite trail superhighway at the lower part of the images. I have used various combinations of settings per Vicent's topic (https://pixinsight.com/forum/index.php?topic=11067.msg68651#msg68651) using the ImageIntegration tool and large-scale pixel rejection, but many of the trails stubbornly refuse to be gone and I think I have gone through all the permutations of the Layers and Growth settings.

I would like to see if anyone has better luck with removing all the trails than I.

Here is a Dropbox link to the zipped images (calibrated and registered): https://www.dropbox.com/s/u0j33fyqq96272e/RegisteredImages.zip?dl=0

The images have a significant light pollution gradient in them, so this will need to be removed using the DynamicBackgroundExtraction tool to see all the trails.

Thanks,
Kerry

Online pfile

  • PTeam Member
  • PixInsight Jedi Grand Master
  • ********
  • Posts: 4537
    • View Profile
this is somewhat time consuming but the technique discussed in this thread will take care of it:

https://pixinsight.com/forum/index.php?topic=8342.0

rob

Offline Kerrywnm1

  • Newcomer
  • Posts: 7
    • View Profile
The process discussed in the thread: https://pixinsight.com/forum/index.php?topic=8342.0, requires you to have at least one registered image that does not contain a satellite trail in the region of question. Believe it or not, I have 51 individual luminance frames making up this image, and not a single frame is devoid of at least two satellite trails along the "satellite superhighway".

--- Kerry

Online pfile

  • PTeam Member
  • PixInsight Jedi Grand Master
  • ********
  • Posts: 4537
    • View Profile
not necessarily, oldwexi is trying to show how to repair an image. but if you simply paint the white (or black) lines over the satellite trails in each image, then the pixel rejection in ImageIntegration will be able to reject the trails more easily than the dim trails.

regardless it is time consuming and somewhat tedious.

rob

Offline GregW

  • Newcomer
  • Posts: 39
    • View Profile
somewhat?
Objects in the mirror are farther than they appear

Offline Kerrywnm1

  • Newcomer
  • Posts: 7
    • View Profile
Well I tried the Pixelmath painting lines technique. But after several tries I decided to use DynamicCrop instead.

--- Kerry

Offline GregW

  • Newcomer
  • Posts: 39
    • View Profile
Well how many individual Subs are there? You could just clone out the meteors on the individual Subs before integrating if they are limited too 50 or so how many more than that would be too much work. I guess that's the Occam's razor approach.
Objects in the mirror are farther than they appear