I am trying to calculate the rms of some 2D matices, but am unsure if my approach is correct:
Rrms = sqrt( sum(sum((Z1 - mean(mean(Z1))).^2 )) /(wk*wl) )
(where Z1 is a matrix with size wk * wl)
Is this correct, and if not, what should I use?
According to Root Mean Square definition it's just square root of sum of squared values. In terms of Matlab code it will be.
R = sqrt(sum(sum(Z.^2))/prod(size(Z)))
Another way is to use rms function from signal processing toolbox.