Author Topic: DeNoise or Deconvolution. Which Should Come First?  (Read 2983 times)

Offline dgbarar

  • PixInsight Enthusiast
  • **
  • Posts: 91
DeNoise or Deconvolution. Which Should Come First?
« on: 2014 January 25 15:30:54 »
Hi,

Not sure if this question has been asked before.  But which process should happen first on linear images: deconvolution or de-noise? What are everyone's thoughts and why.

I have tried it both ways.  My impression is that I get a smoother (better?) result if I de-noise the linear image with TGVDN prior to deconvolution.

Don Barar
« Last Edit: 2014 January 25 16:52:53 by dgbarar »
AP 155 EDF
Paramount ME
ST10 XME/CFW 8A

Offline Carlos Milovic

  • PTeam Member
  • PixInsight Jedi Master
  • ******
  • Posts: 2172
  • Join the dark side... we have cookies
    • http://www.astrophoto.cl
Re: DeNoise or Deconvolution. Which Should Come First?
« Reply #1 on: 2014 January 25 16:59:11 »
Hopefully, the new TGVRestoration process that we are writingmright now will address both problems,at the same time.... Anyway, I think that a mild noise reduction should help the deconvolution.
Regards,

Carlos Milovic F.
--------------------------------
PixInsight Project Developer
http://www.pixinsight.com

Offline jkmorse

  • PixInsight Padawan
  • ****
  • Posts: 931
  • Two questions, Mitch . .
    • Jim Morse Astronomy
Re: DeNoise or Deconvolution. Which Should Come First?
« Reply #2 on: 2014 January 29 02:36:30 »
Carlos,

Can you say a little more about the TGVRestoration process you are working on?  Reading between the lines it sounds like it will combine noise reduction AND deconvolution?  If so that would be an amazing tool.  Of course, now that you've got us interested, no one will leave you alone until its published  >:D

Jim 
Really, are clear skies, low wind and no moon that much to ask for? 

New Mexico Skies Observatory
Apogee Aspen 16803
Planewave CDK17 - Paramount MEII
Planewave IFR90 - Astrodon LRGB & NB filters
SkyX - MaximDL - ACP

http://www.jimmorse-astronomy.com
http://www.astrobin.com/users/JimMorse

Offline Carlos Milovic

  • PTeam Member
  • PixInsight Jedi Master
  • ******
  • Posts: 2172
  • Join the dark side... we have cookies
    • http://www.astrophoto.cl
Re: DeNoise or Deconvolution. Which Should Come First?
« Reply #3 on: 2014 January 29 06:06:37 »
Well, not much to say at the moment... we are still working on the C++ code. Hopefully it will be ready in a few days, or a couple of weeks. I've tested the algorithm in matlab, in the context of reconstruction of magnetic resonance images under heavy undersampling, and it works quite well. We still need to test it with propper astronomical images, and see how it deals with noise. My feeling is that it will generate less ringing, and help reduce noise... but we still need to test it with real raw images.
Regards,

Carlos Milovic F.
--------------------------------
PixInsight Project Developer
http://www.pixinsight.com