Is there a straightforward formula for determining the mean and standard deviation of the gaussian distribution followed by noise applied on an image, knowing the signal-to-noise ratio between the original image and the noisy one?

  • $\begingroup$ If you know the signal power, you can know the noise power from SNR, which is the variance of the noise, and the means is usually assumed zero. $\endgroup$ – BlackMath Oct 8 at 2:45
  • $\begingroup$ Does the expression $SNR = \frac{\mu}{\sigma}$ hold in an image processing context (I found it on wikipedia)? This would boil down to calculating $\sigma = \frac{\mu}{100}$ if the SNR = 20 dB $\endgroup$ – Hakim Oct 8 at 8:09
  • $\begingroup$ If you mean by $\mu$ signal power. $\endgroup$ – BlackMath Oct 8 at 15:28
  • $\begingroup$ In my case, it's just the intensities of the pixels. $\endgroup$ – Hakim Oct 8 at 16:10

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.