Hubble's angular resolution is 0.04 arc seconds. https://en.wikipedia.org/wiki/Angular_resolution

The closest star has an angular diameter of 1.02 milliarc seconds. https://space.fandom.com/wiki/Proxima_Centauri

So, in most images of stars, each star should be approximately 1 pixel wide due to discretization (of course with variable intensity localized to that pixel), assuming we could remove the noise from the imaging system.

Has anyone had success with blind deconvolution (or other techniques) given images of stars with the cross telescope artifacts? I am attaching an image that I was not able to deconvolve.

enter image description here

  • $\begingroup$ Of course, if an object is less than a pixel in angular diameter, it can be divided between adjacent pixels in an image, but given the ratio of imaging resolution to object resolution for stars, this should not be very prevalent. $\endgroup$
    – Alex
    Commented May 23 at 4:18
  • 1
    $\begingroup$ Yes, people have been successful with deconvolution. Likely not blind deconvolution, as you can easily derive the PSF from these systems. One problem with your image is that many pixels are clipped, you don’t know the true intensity of those pixels. $\endgroup$ Commented May 23 at 14:38
  • $\begingroup$ @Cris Luengo Good point about the clipping. $\endgroup$
    – Alex
    Commented May 24 at 23:49


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.