The solution depends on the image type you're using but this demo should get you started. See in-line comments for more information.
The demo shows the normalized RGB values but if you want to classify colors by color names you'll need to add an additional step.
Pure color definitions in RGB space are
- White [1 1 1]
- Blue: [0 0 1]
- Yellow [1 1 0]
- Green [0 1 0]
- Red [1 0 0]
But unless you're using exactly those colors you'll need a more rigorous color classification that classifies floating point RGB values into one of the color categories.
% Read and display image
I = imread('peppers.png');
imshow(I)
axis on
% Identify regions of interest
% coordinates should be the [x,y] areas of the image and in this example
% coordinates is also the image indicies. If you image specifies x and
% y values such that the indicies do not equal the x and y data, you'll
% need an additional step to get convert the (x,y) coordinates to indicies.
coordinates = [320 150; 100 230; 250 335];
hold on
plot(coordinates(:,1), coordinates(:,2), 'wo', 'LineWidth', 2)
% Get color values at specified indicies
cval = I(coordinates(:,2), coordinates(:,1), :);
RGB = [diag(cval(:,:,1)), diag(cval(:,:,2)), diag(cval(:,:,3))];
% Convert to normalized RGB values - you may not need this step if your
% image already uses normalized RGB values.
rgbNorm = double(RGB)./ double(intmax('like',RGB));
rgbStr = compose('[%.2f, %.2f, %.2f]',rgbNorm);
% Display color values next to selected areas - for demo purposes only.
text(coordinates(:,1)+10, coordinates(:,2)-10, rgbStr, ...
'Color', 'W', ...
'VerticalAlignment','baseline', ...
'BackgroundColor', 'k')
% Show the three colors
figure()
hold on
x = 1:size(rgbNorm,1);
scatter(x, 0.5, 1500, rgbNorm, 'filled')
xlim([-3,5])
axis off
