implementation for iterative wiener filter

This is my implementation in the iterative wiener filter in this paper : http://www.tsc.uc3m.es/~navia/LATDS07/IterativeWienerFilter.pdf
I wish it will help anyone
function wiener()
clc;
clear;
f=im2double(rgb2gray(imread('lena.jpg')));
imshow(f);
f=imresize(f,[32 32]);
figure,imshow(f)
[r c]=size(f);
h=fspecial('average');
g=imfilter(f,h,'circular');
s_avg = sum(sum(f))/(r*c);
SNR=90;
n_sigma=s_avg/(10^(SNR/20));
n=n_sigma*randn(size(f));
g=g+n;
[Nf,Mf]=size(g);
[Nh,Mh]=size(h);
L1=floor(Nh/2);
L2=floor(Mh/2);
H=zeros(Nf*Mf);
k=1;
for row=1:Mf,
for col=1:Nf,
hh=zeros(Nf,Mf);
hh(1:Nh,1:Mh)=h;
hh=circshift(hh,[col-1-L1,row-1-L2]);
H(k,:)=hh(:)';
k=k+1;
end
end
%%make vector of m^2*1 of the f,n,g
f=reshape(f',size(f,1)*size(f,2),1);
g=reshape(g',size(g,1)*size(g,2),1);
n=reshape(n',size(n,1)*size(n,2),1);
%%%calculate the autocorrelation matrix of f ,g,n
u=mean(g);
g1=autom(g-u);
Rg=toeplitz(g1);
n1=autom(n);
Rn=toeplitz(n1);%%%%%
Rf=Rg;
steps=10;
mse=zeros(1,steps);
for i=1:steps
B=Rf*H'*inv( (H*Rf*H') +Rn);
fHat=B*(g);
Rf=B*Rg*B';
im=reshape(fHat,[32 32]);
g=reshape(g,[32 32]);
%figure,imshow(im',[]);
mse(1,i) = sum(sum((im(:)-g(:))));
g=reshape(g',size(g,1)*size(g,2),1);
end
t=1:steps;
mse
plot(t,mse);
end
function [Rxx]=autom(x)
N=length(x);
Rxx=zeros(1,N);
for m=1: N+1
for n=1: N-m+1
Rxx(m)=(Rxx(m)+x(n)*x(n+m-1))/N-m+1;
end;
end
end

2 Comments

This should probably go into the File Exchange (FEX)

Sign in to comment.

Answers (0)

Categories

Find more on Just for fun in Help Center and File Exchange

Tags

Asked:

on 1 Aug 2012

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!