inverse filtering image restoration

clc;
clear all;
close all;
f=rgb2gray(im2double(imread('cameraman.tif')));
f=imresize(f,[256 256])
figure,(imshow(f))
[M,N]=size(f);
% k=2.5;
%  for i=1:size(f,1)
%      for j=1:size(f,2)
%          h(i,j)=exp((-k)*((i-M/2)^2+(j-N/2)^2)^(5/6));
%       end
%  end
h=fspecial('gaussian',260,2);
g=(imfilter(f,h,'circular'));
figure,imshow(g,[]);

G = fftshift(fft2(g));
figure,imshow(log(abs(G)),[]);

H = fftshift(fft2(h));
figure,imshow(log(abs(H)),[]);

F = zeros(size(f));
R=70;
for u=1:size(f,2)
    for v=1:size(f,1)
        du = u - size(f,2)/2;
        dv = v - size(f,1)/2;
        if du^2 + dv^2 <= R^2;
        F(v,u) = G(v,u)./H(v,u);
        end
    end
end

figure,imshow(log(abs(F)),[]);

fRestored = abs(ifft2(ifftshift(F)));
figure,imshow(fRestored, []);


this is my code for inverse filtering 
in which restored image is getting fragmented into 4 parts and getting aligned arbitrarily can someone help me fix this

Answer 1

[Robert D Jordan]

It looks like your second last line is out of order.

If the line reads:

fRestored = abs(ifftshift(ifft2(F)));

It works for me.

You were shifting before you did your IFFT.

Rob

Answer 2

[Alex Zhang]

Hi friend, you met this because of the second last code, it should be:

fRestored = abs(ifftshift(ifft2(F)));

I also met similar problems when I do the image restoration. The fftshif() function should always be used after fft2() or ifft2().