Recognition of different colors in one picture

Hello,

I need a help with one algorithm, or at least a suggestion how to start (because I'm Matlab beginner). I need to recognize the colors from the picture. I need my algorithm to take both halves of the picture and recognize beige on one half and blue on the other one. That's the basic. All I can do for now is to load a picture... I also need it to recognize if there is something wrong with it (like big black spot on picture). I think It could be doable with some kind of area where I would calculate mean color or something like that, but I don't how to do it (or maybe there is some better way in Matlab to do it).

Thank you very much for all your help and suggestions!

Answer 1

[Vidhi Agarwal]

Hi @John Mulcahey,

I understand you have a query regarding an algorithm to recognizes colours in different halves of an image and identifies any large black spots. You can achieve the same by following the below steps:

• Start by loading the image into MATLAB.

imagePath = 'your_image.jpg'; % Replace with your image file path
img = imread(imagePath);

• Split the image into two halves.

[rows, cols, ~] = size(img);
% Split the image into two halves
leftHalf = img(:, 1:floor(cols/2), :);
rightHalf = img(:, floor(cols/2)+1:end, :);

• Analyse each half to identify the predominant colours.

leftHSV = rgb2hsv(leftHalf);
rightHSV = rgb2hsv(rightHalf);

% Define thresholds for beige and blue in HSV
beigeThreshold = [0.1 0.1 0.5; 0.2 0.3 1]; % Adjust these values based on your image
blueThreshold = [0.55 0.4 0.2; 0.75 1 1];

% Detect beige in the left half
beigeMaskLeft = (leftHSV(:,:,1) >= beigeThreshold(1,1) & leftHSV(:,:,1) <= beigeThreshold(2,1)) & ...
                (leftHSV(:,:,2) >= beigeThreshold(1,2) & leftHSV(:,:,2) <= beigeThreshold(2,2)) & ...
                (leftHSV(:,:,3) >= beigeThreshold(1,3) & leftHSV(:,:,3) <= beigeThreshold(2,3));

% Detect blue in the right half
blueMaskRight = (rightHSV(:,:,1) >= blueThreshold(1,1) & rightHSV(:,:,1) <= blueThreshold(2,1)) & ...
                (rightHSV(:,:,2) >= blueThreshold(1,2) & rightHSV(:,:,2) <= blueThreshold(2,2)) & ...
                (rightHSV(:,:,3) >= blueThreshold(1,3) & rightHSV(:,:,3) <= blueThreshold(2,3));

% Calculate the proportion of each color
beigeProportion = sum(beigeMaskLeft(:)) / numel(beigeMaskLeft);
blueProportion = sum(blueMaskRight(:)) / numel(blueMaskRight);

• Check for any large black spots in the image.

blackThreshold = 0.1;

% Create a mask for black areas
blackMask = (img(:,:,1) < blackThreshold) & (img(:,:,2) < blackThreshold) & (img(:,:,3) < blackThreshold);

% Calculate the proportion of black areas
blackProportion = sum(blackMask(:)) / numel(blackMask);

For given image the output is given below:

For better understanding of color recognition using HSV, refer to the given documentation:

• https://www.mathworks.com/help/matlab/ref/hsv.html

Hope that Helps!