How can I make mean filter for color image?

I am trying to do something like this for a color image:

I was trying to use plockproc function but kept getting an error; it is my first time working with an image in matlab but i feel like it should not be hard to do. Can anyone help please?

Answer 1

[Image Analyst]

Gleb, you just had to make a minor modification to my code to call it 3 times, once for each color channel, and then recombine the results into an RGB image.

meanFilterFunction4 = @(theBlockStructure) mean2(theBlockStructure.data(:)) * ones(1, 1, class(theBlockStructure.data));
% Now,here we actually to the actual filtering.
blockyImageR = blockproc(redChannel, blockSize, meanFilterFunction4);
blockyImageG = blockproc(greenChannel, blockSize, meanFilterFunction4);
blockyImageB = blockproc(blueChannel, blockSize, meanFilterFunction4);
[blockRows, blockColumns] = size(blockyImageR)

% Recombine separate color channels into a single, true color RGB image.
rgbImage2 = cat(3, blockyImageR, blockyImageG, blockyImageB);

Attached is a full demo using the standard peppers demo image and using a variety of input and output block sizes.

Let me know if you still don't know how to adapt it to what you want.

[Gleb Bulatovskiy]

It works, thank you!

Answer 2

[Image Analyst]

See attached blockproc demo, which does this.

[Gleb Bulatovskiy]

This code doesnt work on Mac for some reason, however it works on windows.

It works for a black and white image, and if I use color image, it makes it black and white first and then processes it and outputs black and white image. What I need to do is the same thing but I need the output to be colored image as well.

Answer 3

[Jan]

See https://www.mathworks.com/matlabcentral/fileexchange/24812-blockmean:

X = RGBimage;  % 2D matrix, 3D RGB array or any other array
V = 100;       % Block size for 1st dim
W = 100;       % and for 2nd dim

S  = size(X);
M  = S(1) - mod(S(1), V);
N  = S(2) - mod(S(2), W);
MV = M / V;
NW = N / W;

% Cut and reshape input such that the 1st and 3rd dimension have the lengths V
% and W:
XM = reshape(X(1:M, 1:N, :), V, MV, W, NW, []);
Y  = sum(sum(XM, 1), 3) ./ (V * W);

% Remove singleton dimensions:
S(1) = MV;
S(2) = NW;
Y    = reshape(Y, S);