Scribd Logo
Upload

Welcome to Scribd!

  • 10K views

    Gaussian Noise &amp All Filters (Matlab Code)

    Uploaded by

    Sharath Kumar N
    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.

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Download as TXT, PDF, TXT or read online from Scribd

    Gaussian Noise &amp All Filters (Matlab Code)

    Uploaded by

    Sharath Kumar N
    10K views4 pages
    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.

    Original Description:

    Matlab Code for noise & All Filters

    Original Title

    Gaussian Noise & All Filters(Matlab Code)

    Copyright

    © Attribution Non-Commercial (BY-NC)

    Available Formats

    TXT, PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    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.

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Download as TXT, PDF, TXT or read online from Scribd
    Download as txt, pdf, or txt
    10K views4 pages

    Gaussian Noise &amp All Filters (Matlab Code)

    Uploaded by

    Sharath Kumar N
    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.

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Download as TXT, PDF, TXT or read online from Scribd
    Download as txt, pdf, or txt
    of 4
    At a glance
    Powered by AI
    The document discusses different image filtering techniques including arithmetic mean, geometric mean, harmonic mean, median, max, min, and alpha-trimmed filters.

    The filters used are arithmetic mean, geometric mean, harmonic mean, median, max, min and alpha-trimmed filters.

    Arithmetic mean takes the average, geometric mean takes the nth root of the product, harmonic mean takes the reciprocal average, median selects the middle value, max selects the maximum, min selects the minimum.

    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