How it works?

Hi, can anyone help to explain this code? It's a Oil Painting code taken online...Thanks

A=imread('aish.jpg');

%Define the matrix size of  your convience.

m=5;
n=6;

Image=uint8(zeros([size(A,1)-m,size(A,2)-n,3]));

%Calculate the histogram for each RGB value.

for v=1:3

for i=1:size(A,1)-m
    for j=1:size(A,2)-n
        mymask=A(i:i+m-1,j:j+n-1,v);
        h=zeros(1,256);
        for x=1:(m*n)
            h(mymask(x)+1)=h(mymask(x)+1)+1;
        end

%Maximum occurring value and the position is obtained

[maxvalue,pos]=max(h);

Image(i,j,v)=pos-1;
    end
end
end
figure,imshow(Image);

[Jan]

Please ask a specific question. It is not clear if we should spend the time for explaining what "m=5" does, or how a "for v = 1:3" loop works.

[Superb]

Oh, sorry, since I'm new in Matlab, my question is this

what does this line mean?

mymask=A(i:i+m-1,j:j+n-1,v);

is that it is i:j:k? How to know what is i,j,and k do? I mean is that i is initialize? j is increment value and k is the last value? TQ

[Jan]

You can try this in the command line: 3:4:20

[Image Analyst]

Superb, note the comma so it's a:b, c:d and not a:b:c. See my explanation below.

Answer 1

[Image Analyst]

It looks like they go along the image a pixel at a time, calculating the histogram of a subimage (located with its upper left pixel at the scanning pixel location) separately, then logging the value of the most frequently occurring graylevel for each color channel in a new image at the location of the scaning pixel.

mymask=A(i:i+m-1,j:j+n-1,v);

takes a rectangular block out of the A image going from row1 = i to row2 = (i+m-1). Same for the columns, so it basically gets a little rectangular chunk out of the larger, original image.

The code is not well commented and has poorly chosen variable names. I cleaned it up a bit. Try this version:

% Script to turn an RGB image into a spatially quantized version of it
% that has the appearance of a painting.
clc;    % Clear the command window.
workspace;  % Make sure the workspace panel is showing.
fontSize = 20;

% Read in a standard MATLAB color demo image.
folder = fullfile(matlabroot, '\toolbox\images\imdemos');
baseFileName = 'peppers.png';
fullFileName = fullfile(folder, baseFileName);
% Get the full filename, with path prepended.
fullFileName = fullfile(folder, baseFileName);
if ~exist(fullFileName, 'file')
  % Didn't find it there.  Check the search path for it.
  fullFileName = baseFileName; % No path this time.
  if ~exist(fullFileName, 'file')
    % Still didn't find it.  Alert user.
    errorMessage = sprintf('Error: %s does not exist.', fullFileName);
    uiwait(warndlg(errorMessage));
    return;
  end
end
rgbImage = imread(fullFileName);
% Get the dimensions of the image.  numberOfColorBands should be = 3.
[rows columns numberOfColorBands] = size(rgbImage);
% Display the original color image.
subplot(2, 2, 1);
imshow(rgbImage, []);
axis on;
title('Original Color Image', 'FontSize', fontSize);
drawnow;

%Define the window size.
windowSizeVert = 15;
windowSizeHoriz = 11;
% Initialize an output image.
paintedImage=uint8(zeros([rows-windowSizeVert, columns-windowSizeHoriz, 3]));

%Calculate the histogram for each RGB value.
for colorChannel = 1 : numberOfColorBands
  for row = 1: rows - windowSizeVert
    for column = 1 : columns - windowSizeHoriz
      % Get a little sub-image out.  
      % The sub-image has it's upper left corner at pixel location i,j.
      subImage = rgbImage(row:row+windowSizeVert-1, column:column+windowSizeHoriz-1, colorChannel);
      
      % If you want to see the sub image, uncomment this
      % but it slows it down A LOT!
%       subplot(2, 2, 2);
%       imshow(subImage);
%       drawnow;

      % Compute the histogram by incrementing values.
      histOfWindow = zeros(1,256); % Initialize
      for x = 1:(windowSizeVert * windowSizeHoriz)
        histOfWindow(subImage(x)+1) = histOfWindow(subImage(x)+1) + 1;
      end
      
      % Maximum occurring value and the position is obtained
      % for this color channel.
      [maxvalue, indexAtMax] = max(histOfWindow);
      
      % Assign the intensity value, which is the index at the max - 1
      % since the indexes go from 1-356 but the intensities gto from 0-255.
      paintedImage(row,column,colorChannel) = indexAtMax - 1;
    end
    fprintf('Processed row %d of %d in color channel #%d of %d\n', ...
      row, rows, colorChannel, numberOfColorBands);
  end
end
% Enlarge figure to full screen.
set(gcf, 'units','normalized','outerposition',[0 0 1 1]); % Maximize figure.
% Now we have our output image, so display it.
subplot(2, 2, 3);
imshow(paintedImage);
axis on;
title('"Painted" Image', 'FontSize', fontSize);
msgbox('Done with demo');