Go Back   IceInSpace > Equipment > Astrophotography and Imaging Equipment and Discussions

Reply
 
Thread Tools Rate Thread
  #1  
Old 07-04-2015, 11:46 AM
Andy01's Avatar
Andy01 (Andy)
My God it's full of stars

Andy01 is offline
 
Join Date: Mar 2012
Location: Melbourne
Posts: 3,253
Please list your favourite star removal techniques.

Hi everyone -

Can I ask your collective wisdom for some help please?

I'd like to find a sure fire, foolproof and accurate method of perfect Star removal for processing Nebulae images, that doesn't take forever!

My focus is primarily NB imaging but something that works on RGB images as well would be very useful. I usually want to remove the false coloured NB stars and replace them with RGB stars.

There are various methods and actions around using PS, Startools & Pixinsight but has anyone got one that works all the time? (and in a perfect world is cross platform, I use a mac).

I'll list what I have tried so far... with various degrees of success. *Disclaimer - I could be using them incorrectly so my apologies if that's the case.

Straton - http://www.zipproth.com/Straton/ PC only (15 E) I have only tried the demo, so not sure about this one yet.

Troy's Astro actions - http://troypiggo.blogspot.com.au/201...o-actions.html - Cross platform PS Action - (free) sort of works but can leave dark circles and overall loss of detail. A fair bit of cloning/cleanup is required afterwards.

Astro Anarchy (JP Metsavainio) http://astroanarchy.blogspot.com.au/...ps-action.html
Almost identical to Troy's above, with virtually the same results.

Startools Magic Module - http://startools.org/drupal/ Mask and select stars then reduce /shrink them with several methods.
I have found this feature generally useful but tricky to get realistic results with large stars.

From Rob's AP http://www.astro-photography.net/Star-Removal.html - Using PS - Select>Colour range>Highlights>feather/expand>fill(content aware)
A bit hit & miss for me so far - probably depends on the image and scale - can result in loss of detail, and still lots of cloning/cleanup afterwards - needs a lot of trial & error to make it work well.

From Multiweb - http://www.astropic.net/astro/tut_01/
A good step by step tutorial on Star removal and tone mapping from Marc, with links to an old thread on this topic here on IIS - http://www.iceinspace.com.au/forum/s...ad.php?t=62131

http://darkhorizons.emissionline.com/stars.htm I found this when creating this thread, but havn't tried it out for myself yet.

Please add your own preferred, reliable method below ....

Many thanks in advance,

Andy
Reply With Quote
  #2  
Old 07-04-2015, 03:29 PM
Eden's Avatar
Eden (Brett)
Registered Rambler

Eden is offline
 
Join Date: Jan 2014
Location: Melbourne, Australia
Posts: 399
Hi Andy,

I purchased Straton a few months back, having used it on several star-heavy targets such as the Carina complex with very good results. I've attached some example images for you to have a look at. The authors main selling point appears to be the claim that it delivers better results than JP Metsavainio's "Gold Standard" tone-mapping method.

My only issue with it is the lack of multi-processor support, so working on large images or very dense star fields can be slow going. The example image shown here took approximately 18 minutes of processing time on my 4-core Intel laptop.

Although the end results are more or less the same as what you would expect from JP's tone-mapping method, it's certainly a lot quicker than doing it manually. I also found that the options which allow you to fine-tune the amount of star removal work very well.

15 Euro well spent, IMHO.
Attached Thumbnails
Click for full-size image (Eta Carina Starless Reference Color 8bit.jpg)
193.4 KB143 views
Click for full-size image (Eta Carina Starless Reference Color 8bit_withoutStars.jpg)
141.4 KB154 views
Reply With Quote
  #3  
Old 07-04-2015, 03:44 PM
Amaranthus's Avatar
Amaranthus (Barry)
Thylacinus stargazoculus

Amaranthus is offline
 
Join Date: Jan 2014
Location: Judbury, Tasmania
Posts: 1,203
I find in Startools, it is best to create a 'fat star' mask and use "Tighten" in Magic (a few iterations), and then switch to Heal, increase the mask further, and really bump up the sampling. You get quite a nice result with this 2-step process.
Reply With Quote
  #4  
Old 07-04-2015, 05:53 PM
troypiggo's Avatar
troypiggo (Troy)
Bust Duster

troypiggo is offline
 
Join Date: Nov 2008
Location: Brisbane, Australia
Posts: 4,846
Jeez mate. Great idea for a thread, but I'm embarrassed you put my clunky PS action in there. TBH I use PI exclusively these days.
Reply With Quote
  #5  
Old 07-04-2015, 05:55 PM
RickS's Avatar
RickS (Rick)
PI cult recruiter

RickS is offline
 
Join Date: Apr 2010
Location: Brisbane
Posts: 10,584
Andy,

I find that clouds work very well and they're widely available

Bad jokes aside, I think that star removal is an inherently difficult problem and one that doesn't lend itself to simple automation. Stars vary widely in size (and even shape on a bad night), in brightness and they appear in a range of dark and light backgrounds.

What works for me is a carefully tuned star mask in PixInsight, sometimes using PixelMath to combine small and large scale masks. Getting that right is the hard part. Once I have a good mask the rest is easy - just use one of the wavelet tools to remove small scale detail (MMT or ATWT) and maybe a MorphologicalTransformation to clean up afterwards.

Cheers,
Rick.
Reply With Quote
  #6  
Old 07-04-2015, 06:20 PM
clive milne
Registered User

clive milne is offline
 
Join Date: May 2011
Location: Freo WA
Posts: 1,443
Surely continuum subtraction should be on the list as well?
Reply With Quote
  #7  
Old 07-04-2015, 10:51 PM
Amaranthus's Avatar
Amaranthus (Barry)
Thylacinus stargazoculus

Amaranthus is offline
 
Join Date: Jan 2014
Location: Judbury, Tasmania
Posts: 1,203
The biggest problem I typically face is with halos reflected against the nebulae. In a galaxy field with a dark background, it is typically a simple task. With some bright stars set against nebulosity, it can be a real pain!
Reply With Quote
  #8  
Old 08-04-2015, 10:06 AM
Placidus (Mike and Trish)
Narrowing the band

Placidus is offline
 
Join Date: Mar 2011
Location: Euchareena, NSW
Posts: 3,719
G'day colleagues,

Because I write my own image processing software, I should love to understand how to do this not at a "Use this tool in this package" level, but at a mathematical algorithmic level. I have no answer, but let me talk around the problem for a bit.

One reason for removing stars (and then replacing them) is to avoid ringing when deconvolving or sharpening. I tackle that (as does PixInsight) by using an anti-ringing step between each deconvolution step. The other reason is to control purple haloes when using the Hubble palette (R=SII, G=Ha, B=OIII), in order to gain an astrophysical understanding of what we are looking at. This is my main concern today.

The origin of the typical purple halos is that there's usually say 10 times more light in a nebula from H-alpha than there is from OIII or SII. That makes the image overwhelmingly green. We then stretch the red and blue channels, either to try to bring out what OIII and SII there is, for astrophysical reasons, or to make the image more colour balanced, for aesthetic reasons. The result is that the stars get bigger in red and blue than they are in green.

Our goal is to try to work out what the background would look like if the star wasn't there. The human eye is amazingly good at that, but we don't know how we do it.

One way to think about this is it is the same as trying to guess what is the pattern in a tablecloth under a coffee cup. We try to mathematically model how the brightness changes with position in the background well outside the star, and then use that model to guess what is "under" the star. My own star-removing tool does this by fitting a general conic section z = ax + by + cxy + dxx + eyy + f to the brightness well outside the star, and then use that model to interpolate under the star. This is done using multiple linear regression. The input data to the model is weighted to the fuzzy extent that the particular pixel is dark. Bright pixels are either other small stars, or part of the halo or diffraction spike of the main star, and we want to ignore a pixel to the extent that we think it is part of a star.

This approach works astonishingly well on gently changing background, but fails utterly if there is say a shock front crossing through the star. For that, we would need a much more sophisticated model. It also fails if there are two very bright stars quite close to each other. These special cases can be handled manually for example in PhotoShop, by cloning a bit of background that looks like the bit that we think should be under the star. That's not very scientific - it is finger painting - but is perhaps justifiable artistically.

There is another approach, which I've fiddled with but not yet got working well. Instead of thinking of the star as an opaque coffee cup, where all information of what lies below has been lost and has to be guessed, think of the star as more like a spreading tea-stain. If we can mathematically model the stain, we can subtract the stain, to reveal what is underneath. We don't need to know about shock fronts for example. They will be revealed when we subtract the stain. This method will only work well for the outer edges of the stain (the halo). The centre of the stain (the burned out star) is opaque, and all information about what is underneath is already lost.

An approach to this "stain modelling" method is to assume that the star is radially symmetrical, and that it has some sort of expected profile, or example (prior to curves) a Gaussian profile with the brightest central region clipped to 65535.

All possible approaches to removing stars or haloes, regardless of package, must ultimately use one or both of these two approaches:

(a) Look at the surrounding tablecloth and try to guess what is under the coffee cup.
(b) Look at the tea-stained cloth, and from what you know about tea-stains, try to work out what the unblemished cloth would look like.

I hope that when you tackle the problem using existing commercial tools, it helps to think in this way because it will help understand why said tools are not quite working as expected.

Best,
Mike
Reply With Quote
  #9  
Old 08-04-2015, 10:52 AM
RickS's Avatar
RickS (Rick)
PI cult recruiter

RickS is offline
 
Join Date: Apr 2010
Location: Brisbane
Posts: 10,584
Nicely summarised, Mike

It's worth noting that provided you're going to put the stars back in a narrowband image after processing the colour there's no need to do a perfect job of star removal. I've got some excellent results from just doing a big star reduction on the Oiii and Sii masters before combining them with Ha.

If I had the time I'd have a go at writing a PixInsight process to do star removal. Unfortunately, I don't think it is something that would be amenable to scripting and my C++ is a bit rusty.

Cheers,
Rick.
Reply With Quote
  #10  
Old 08-04-2015, 11:09 AM
Andy01's Avatar
Andy01 (Andy)
My God it's full of stars

Andy01 is offline
 
Join Date: Mar 2012
Location: Melbourne
Posts: 3,253
Quote:
Originally Posted by Eden View Post
15 Euro well spent, IMHO.
I'll have to give it a proper try then, cheers

Quote:
Originally Posted by Amaranthus View Post
I find in Startools, it is best to create a 'fat star' mask and use "Tighten" in Magic (a few iterations), and then switch to Heal, increase the mask further, and really bump up the sampling. You get quite a nice result with this 2-step process.
Thanks Barry, I'll try that out

Quote:
Originally Posted by troypiggo View Post
I'm embarrassed you put my clunky PS action in there.
Thanks for posting it anyway Troy, I used it a lot with reasonable results but the cloning afterwards was taking too long, hence the question.

Quote:
Originally Posted by RickS View Post
What works for me is a carefully tuned star mask in PixInsight
Thanks Rick - Guess I'm going to have to bite the bullet and delve into Pixinsight soon

Quote:
Originally Posted by clive milne View Post
Surely continuum subtraction should be on the list as well?
Sorry Clive, ?

Quote:
Originally Posted by Amaranthus View Post
The biggest problem I typically face is with halos reflected against the nebulae.
I feel your pain Barry...

Quote:
Originally Posted by Placidus View Post
... it will help understand why said tools are not quite working as expected.
Yes very true, and interesting - but I'm personally just looking for an aesthetically pleasing result that works

Quote:
Originally Posted by RickS View Post
provided you're going to put the stars back in a narrowband image after processing the colour
Yes exactly, My intent is to replace the NB ones with RGB stars for a more "realistic" looking image. But not at the expense of hours of time and loss of detail data.



Thanks everyone for your input - seems it's a common problem, will be interesting to see what others contribute as well.

Andy
Reply With Quote
  #11  
Old 08-04-2015, 11:33 AM
RickS's Avatar
RickS (Rick)
PI cult recruiter

RickS is offline
 
Join Date: Apr 2010
Location: Brisbane
Posts: 10,584
Wrt Clive's comment... Astrodon makes a red continuum filter. It's basically a 5nm Ha filter with the centre bandwidth shifted slightly (645nm instead of 656.3nm). The idea is that you can subtract a continuum sub from a Ha sub and broadband sources like stars are removed without affecting the nebulosity.

Like all other star removal methods the results are not perfect.

Cheers,
Rick.
Reply With Quote
  #12  
Old 08-04-2015, 12:28 PM
clive milne
Registered User

clive milne is offline
 
Join Date: May 2011
Location: Freo WA
Posts: 1,443
Hi Andy,
Rick's explanation is succinct enough,
Plus, there is a fair amount written in the professional literature (on continuum subtraction) available online.

I was reading an interesting whitepaper on the subject a couple of months ago which detailed the process of using RGB (or UGRIZ, whatever) channels rather than wasting time with dedicated offband filters. I just tried to dig it up again but haven't managed to find it.

As Rick indicated, it is by no means without issue but does have some advantages over software solutions that rely on detecting stars by their guasian profile, the main one being that the end result isn't confused by such things as shock fronts, double stars or other imaging artifacts. It is also effective for pulling out emission line sources from the spiral arms of galaxies, etc)
Reply With Quote
  #13  
Old 08-04-2015, 12:31 PM
RickS's Avatar
RickS (Rick)
PI cult recruiter

RickS is offline
 
Join Date: Apr 2010
Location: Brisbane
Posts: 10,584
Quote:
Originally Posted by clive milne View Post
I was reading an interesting whitepaper on the subject a couple of months ago which detailed the process of using RGB (or UGRIZ, whatever) channels rather than wasting time with dedicated offband filters. I just tried to dig it up again but haven't managed to find it.
That sounds interesting. Please pass it on if you happen to come across it again, Clive.
Reply With Quote
  #14  
Old 08-04-2015, 12:33 PM
clive milne
Registered User

clive milne is offline
 
Join Date: May 2011
Location: Freo WA
Posts: 1,443
One of the issues (with continuum subtraction) is that for a given exposure budget, the time it takes to shoot the continuum frame eats into your signal whilst contributing a noise component of its own.

No free lunch.

Well, not unless you can infer the continuum from one or more of the RGB channels.
Reply With Quote
  #15  
Old 08-04-2015, 12:34 PM
clive milne
Registered User

clive milne is offline
 
Join Date: May 2011
Location: Freo WA
Posts: 1,443
Quote:
Originally Posted by RickS View Post
That sounds interesting. Please pass it on if you happen to come across it again, Clive.

Sure.

I'm scratching my head trying to remember where I came accross it.
From memory, my google search at the time was specific to metalicity of local group dwarf irregular galaxies inferred from Halpha to Oiii ratio.

The (obvious) caution probably needs to be added that broadband channels include the emission lines, so it is not as straightforward as a simple subtraction.

Anyway, I'll keep digging.

regards,
c

Last edited by clive milne; 08-04-2015 at 01:55 PM.
Reply With Quote
  #16  
Old 08-04-2015, 03:06 PM
troypiggo's Avatar
troypiggo (Troy)
Bust Duster

troypiggo is offline
 
Join Date: Nov 2008
Location: Brisbane, Australia
Posts: 4,846
IIRC, Bert (avandonk) had a continuum filter and was playing around and posting some images in the deep space section some time ago.

Edit:
http://www.iceinspace.com.au/forum/s...ight=continuum
http://www.iceinspace.com.au/forum/s...ight=continuum
Reply With Quote
  #17  
Old 09-04-2015, 02:26 PM
Geoff45's Avatar
Geoff45 (Geoff)
PI rules

Geoff45 is offline
 
Join Date: Jul 2006
Location: Sydney
Posts: 2,631
http://trappedphotons.com/blog/?p=731
Star removal in PixInsight
Reply With Quote
  #18  
Old 12-04-2015, 04:15 PM
Alasdair
Registered User

Alasdair is offline
 
Join Date: Apr 2015
Location: Melbourne, Australia
Posts: 61
This is an interesting problem. As far as I can tell, most commercial programs use some version of Photoshop's "dust and scratch removal filter", about which I can find nothing about its implementation. It's probably some despeckling technique, one method of which uses what's called a Lee filter: http://angeljohnsy.blogspot.com/2014/08/lee-filter.html
However, most modern despeckling methods use wavelets, and I don't know if Photoshop includes a wavelet transform.

From a naive perspective, I'd imagine that a "star" is a white blob (possibly with blurred edges) which is significantly brighter than its surroundings. This means that you should be able to remove stars by setting a threshold level, and removing all blobs for which their mean value and the mean value of their surrounds differ by at least that threshold level. (This may have to be done in an adaptive way, changing the threshold across the image, maybe by checking the variance of the surrounding neighbourhood: this is looking like a Wiener filter). Having removed the blobs, the holes can be filled in by using an interpolation technique.

I'll have a a go at this (if time permits) over the next week or so, and see what I get.
Reply With Quote
Reply

Bookmarks

Thread Tools
Rate This Thread
Rate This Thread:

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT +10. The time is now 08:59 AM.

Powered by vBulletin Version 3.8.7 | Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.
Advertisement
Testar
Advertisement
Bintel
Advertisement