# how to calculate power spectrum of Gaussian noise from the known variance of it?

can someone explain me the following paragraph, i tried to understand it but i am not getting it exactly, Thank you !

Since you know the variance, you can predict the energy of the noise image, ∑xy|η(x,y)|2 ≈ ∑xyσ2η. Then use the facts that (i) ∑xy|η(x,y)|2∝ ∑uv|N(u,v)|2, and that (ii) for white noise, |N(u,v)| is approximately independent of u,v

• Is this from a particular source? Also, could you be more specific in the notations? Like what is $\eta$? The pixel row and column are $u$ and $v$ and then what is $N$? – Engineer Oct 17 '19 at 20:01