anon
anon

Reputation:

Var of added gaussian noise

I have basically this:

I = im2double(imread('lena.png'));

% Add gaussian noise
J = imnoise(I,'gaussian',0,0.05);

Now having both undegraded image I and noised image J how can I estimate that the variance used for gaussian noise was 0.05?

Upvotes: 2

Views: 290

Answers (2)

Oliver Charlesworth
Oliver Charlesworth

Reputation: 272487

You can obtain the noise waveform via subtraction. You can obtain the sample variance with var().

Of course, the sample variance is only an estimate for the true variance.

Upvotes: 2

atw13
atw13

Reputation: 719

Subtract J from I, square each difference, and take the average. Since you know the mean (zero) you can use the number of pixels as the denominator in your average.

Upvotes: 2

Related Questions