Use FIR Wiener Filter to remove the noise??

Hello

I have an audio file that added white guass noise to it, I need to determine the FIR filter coefficients using the Wiener-Hopf equation

so I started like this and stopped

[X, Fs] = audioread('test.wav'); % read audio file

gnoise = awgn(X, 15); % Add white guass noise

Noise = gnoise - X; %extract the noise

as a result X and Fs would be Fs = 48000; X = 961452x2 double

how can we design a wiener filter to clear the noise? I wrote the code below for that but it didn't cleared

N = 100;

rxx = xcorr(gnoise, N);

rxy = xcorr(gnoise, X, N);

R = toeplitz(rxx);

W = inv(R).*rxy; %%% my W value is not one dimension, I could't get it as one dimension matrix.

Y = filter(W(:), gnoise);

thankks.

Answer 1

[Akash]

Hi Lana,

To design a Wiener filter that can effectively reduce the noise, you might find the MATLAB File Exchange link provided below useful. It contains a design for a Wiener filter where a signal is taken, noise is added, and then the Wiener filter is applied to mitigate the noise's effect.

https://www.mathworks.com/matlabcentral/fileexchange/70123-wiener-filter?s_tid=FX_rc2_behav