I have a (continuous) function f and a vector v with values of f over an interval of integers as components, v = [f(0), ..., f(n)].

v is affected from Poisson noise such that the components are not equal to f(i) but to a poisson sample with mean f(i).

I need a way to remove noise from data, I found something about Poisson Noise in images but I don't think it's the same because of the correlation between pixel of the same image.

  • $\begingroup$ If you don't have any prior you don't have much to do. Prior could be about the signal (Bandwidth, Being smooth, etc...) or the noise (Mean value, correlation, etc...). What information do you have? $\endgroup$ – Royi Oct 16 '19 at 11:49
  • $\begingroup$ I can assume a specific form for the function which relate the data that I'm using. For example it could be nice to find something that work with linear functions $\endgroup$ – user12147352 Oct 16 '19 at 20:53
  • $\begingroup$ If your function is indeed linear, share the model with us and we'll try building estimator for the model give Poisson Noise. Probably, if the noise is white, you'll get classic Least Squares solution. $\endgroup$ – Royi Oct 16 '19 at 21:18

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.