Noise removal using filters

I have to remove the noise using Component median filtering and vector median filtering ,can any one tell wehere i can get codes please

Answer 1

[Vidhi Agarwal]

Hi @FIR,

Component Median Filter applies the median filter independently to each color channel (R, G, B), and Vector Median Filter Considers each pixel as a vector (R,G,B) and selects the pixel in the window with the minimum total distance (usually Euclidean distance) to all others.

Sample code for component median filter is given below:

function output = componentMedianFilter(I, windowSize)
    % I: RGB image
    % windowSize: odd integer (e.g., 3)
    output = zeros(size(I));
    for c = 1:3 % R, G, B
        output(:,:,c) = medfilt2(I(:,:,c), [windowSize windowSize]);
    end
    output = uint8(output);
end

following is the sample code for Vector Median filter:

function output = vectorMedianFilter(I, windowSize)
    % I: RGB image
    % windowSize: odd integer (e.g., 3)
    [rows, cols, ~] = size(I);
    pad = floor(windowSize / 2);
    paddedI = padarray(double(I), [pad pad], 'symmetric');
    output = zeros(size(I));

    for i = 1:rows
        for j = 1:cols
            % Extract window
            window = paddedI(i:i+2*pad, j:j+2*pad, :);
            % Reshape to Nx3 matrix (N = windowSize^2)
            vectors = reshape(window, [], 3);
            
            % Compute pairwise L2 distances
            D = pdist2(vectors, vectors);
            totalDist = sum(D, 2);
            
            % Find vector with minimal total distance
            [~, idx] = min(totalDist);
            output(i, j, :) = vectors(idx, :);
        end
    end
    output = uint8(output);
end

To understand to more about "medfilt2" refer to the following documentation: https://www.mathworks.com/help/images/ref/medfilt2.html

Hope this helps!