Scribd
Upload
10K views

Gaussian Noise &amp All Filters (Matlab Code)

The document applies various filtering techniques to an original image with added gaussian noise. It first applies arithmetic, geometric, and harmonic mean filters. It then applies a contraharmonic mean filter with both positive and negative exponents. Additional filters applied include a median filter, max filter, min filter, midpoint filter, and alpha-trimmed filter. For each filter, the resulting filtered image is plotted alongside the original and noise-added image for comparison.

Uploaded by

Sharath Kumar N
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as TXT, PDF, TXT or read online on Scribd
10K views

Gaussian Noise &amp All Filters (Matlab Code)

The document applies various filtering techniques to an original image with added gaussian noise. It first applies arithmetic, geometric, and harmonic mean filters. It then applies a contraharmonic mean filter with both positive and negative exponents. Additional filters applied include a median filter, max filter, min filter, midpoint filter, and alpha-trimmed filter. For each filter, the resulting filtered image is plotted alongside the original and noise-added image for comparison.

Uploaded by

Sharath Kumar N
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as TXT, PDF, TXT or read online on Scribd
You are on page 1/ 4

K = imread('C:\lena.

bmp');
subplot(2,5,1),imshow(K),title('original image');
I=imnoise(K,'gaussian');
subplot(2,5,6),imshow(I),title('Image with gaussian noise');
%arithmetic mean filter
h = ones(5,5) / 25;
I2 = imfilter(I,h);
subplot(3,5,2),imshow(I2),title('arithmetic mean filtered image');
%geometric mean filter
x=double(I);
y(1,: )=x(1,: );
y(1:512,1)=x(1:512,1);
y(512,: )=x(512,: );
y(1:512,512)=x(1:512,512);
for k=2:511
for j=2:511
y(k,j)=(x(k-1,j-1)*x(k,j-1)*....
x(k-1,j)*x(k+1,j)*x(k,j)*....
x(k,j-1)*x(k-1,j)*x(k+1,j+1)*....
x(k,j+1))^(1/9);
end
end
W=uint8(y);
subplot(3,5,3),imshow(W),title('geometric mean filtered image');
%harmonic mean filter
x=double(I);
y(1,: )=x(1,: );
y(1:512,1)=x(1:512,1);
y(512,: )=x(512,: );
y(1:512,512)=x(1:512,512);
for k=2:511
for j=2:511
y(k,j)=9/(1/(x(k-1,j-1))+1/(x(k,j-1))+....
1/(x(k-1,j))+1/(x(k+1,j))+1/(x(k,j))+....
1/(x(k,j-1))+1/(x(k-1,j))+1/(x(k+1,j+1))+....
1/(x(k,j+1)));
end
end
Z=uint8(y);
subplot(3,5,4),imshow(Z),title('harmonic mean filtered image');
%contrahormonic filter
data=im2double(I);
Q=-2;
sumn=[];
sumd=[];
pixln=0;
pixld=0;
ro=512;
col=512;
for i=1:ro;
for j=1:col;
for m=-1:1;
for n=-1:1;
if (i+m>0 && i+m<ro && j+n>0 && j+n<col && ... % To keep in
dices in limit
1+m>0 && 1+m<ro && ...
1+n>0 && 1+n<col)
pixl1=(data(i+m,j+n)).^(Q+1); % Applicatio
n of Formula
pixl2= (data(i+m,j+n)).^Q;
pixln=pixln+pixl1; % Applicatio
n of Summation
pixld=pixld+pixl2;
end
end
end
reformedimage(i,j)=(pixln/pixld);
pixln=0;
pixld=0;
end
end
subplot(3,5,5),imshow(reformedimage),title('contra-harmonic mean (Q=-ve)');

data=im2double(I);
Q=2;
sumn=[];
sumd=[];
pixln=0;
pixld=0;
ro=512;
col=512;
for i=1:ro;
for j=1:col;
for m=-1:1;
for n=-1:1;
if (i+m>0 && i+m<ro && j+n>0 && j+n<col && ... % To keep in
dices in limit
1+m>0 && 1+m<ro && ...
1+n>0 && 1+n<col)
pixl1=(data(i+m,j+n)).^(Q+1); % Applicatio
n of Formula
pixl2= (data(i+m,j+n)).^Q;
pixln=pixln+pixl1; % Applicatio
n of Summation
pixld=pixld+pixl2;
end
end
end
reformedimage(i,j)=(pixln/pixld);
pixln=0;
pixld=0;
end
end
subplot(3,5,7),imshow(reformedimage),title('contra-harmonic mean (Q=+ve)');
%median filter
A = medfilt2(I,[3 3]);
subplot(3,5,8),imshow(A),title('median filtered image');
%max filter
[m n]=size(I);
c=ones(3,3)/9;
[p q]=size(c);
p=(p-1)/2;
q=(q-1)/2;
z=I;
for i=p+1:m-p
for j=q+1:n-q
mask1=zeros(3,3);
for k=-p: p
for l=-q:q
mask1(p+1-k,q+1-l)=I(i-k,j-l);
end
end
max1=mask1( : );
max1=max(max1);
z(i,j)=max1;
end
end
subplot(3,5,9),imshow(uint8(z)),title('max filtered image');
%min filter
[m n]=size(I);
c=ones(3,3)/9;
[p q]=size(c);
p=(p-1)/2;
q=(q-1)/2;
v=I;
for i=p+1:m-p
for j=q+1:n-q
mask1=zeros(3,3);
for k=-p: p
for l=-q:q
mask1(p+1-k,q+1-l)=I(i-k,j-l);
end
end
min1=mask1( : );
min1=min(min1);
v(i,j)=min1;
end
end
subplot(3,5,10),imshow(uint8(v)),title('min filtered image');

[m n]=size(I);
c=ones(3,3)/9;
[p q]=size(c);
p=(p-1)/2;
q=(q-1)/2;
z=I;
for i=p+1:m-p
for j=q+1:n-q
mask1=zeros(3,3);
for k=-p: p
for l=-q: q
mask1(p+1-k,q+1-l)=I(i-k,j-l);
end
end
mid1=mask1(: );
mid1=(min(mid1)+max(mid1))/2;
z(i,j)=mid1;
end
end
subplot(3,5,12),imshow(uint8(z)),title('mid point filtered image');

%alpha-trimmed filter
m = 3; n = 3; d = 2;
if (d <= 0) | (d/2 ~= round(d/2))
error('d must be a positive, even integer.')
end
inclass = class(I);
g = im2double(I);
f = imfilter(I, ones(m, n), 'symmetric');
for k = 1:d/2
f = imsubtract(f, ordfilt2(I, k, ones(m, n), 'symmetric'));
end
for k = (m*n - (d/2) + 1):m*n
f = imsubtract(f, ordfilt2(I, k, ones(m, n), 'symmetric'));
end
f = f / (m*n - d);

subplot(3,5,13),imshow(f),title('alpha-trimmed filtered image');

You might also like