I am experimenting with image down-sampling and would like to calculate a numerical magnitude for alias-induced moiré patterns. Is it possible to do this by analyzing the frequency-domain (FFT) of the input image and knowing the relative sampling rate?

Assume that I have already applied any pre-sample filtering prior to taking the FFT, so the sampling will be point-based.

  • $\begingroup$ Do you have any sample images handy? $\endgroup$ – Phonon Apr 27 '14 at 4:35
  • $\begingroup$ @Phonon: Something like the "rings image" (large) in the "Resampling Filters" article at imagemagick.org would be a good start, but I would like this to apply to any image. $\endgroup$ – nobar Apr 27 '14 at 15:11
  • $\begingroup$ I wonder if the reason I have no answers is because this question is too easy, ... or too hard. :) $\endgroup$ – nobar Jun 7 '14 at 3:06
  • $\begingroup$ Idea: Compare FFT of an aliased image to that of a reference image. The relative magnitude of frequencies in the aliased image compared to the same frequencies in the reference image represent the magnitude of moire at the corresponding frequencies. $\endgroup$ – nobar Sep 30 '14 at 2:50

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.