Color differentation in the image

Hi All,

Is there any code to find differentation in the image?

Answer 1

[Image Analyst]

What do you mean by differentiation? I have several color segmentation demos in my File Exchange if that's what you mean: http://www.mathworks.com/matlabcentral/fileexchange/?term=authorid%3A31862

[hamed abdulaziz]

Thanks for your response,I have many coloured images some of them contains some color (red,blue,brown,and black)what I need to know how can I detect the color variation in the segmented image(ROI),could you help me by demo please?

[Image Analyst]

If you have objects with a variety of vivid colors, then it's probably best to threshold the saturation image. See my Color segmentation using HSV demo in my File Exchange, or see this answer: http://www.mathworks.com/matlabcentral/answers/116126#answer_124379 Or perhaps you'd like this: http://www.mathworks.com/matlabcentral/fileexchange/37197-dem-diffused-expectation-maximisation-for-image-segmentation

[hamed abdulaziz]

Please could you take a look to some my images that attached

[Image Analyst]

OK, once you attach them.

[hamed abdulaziz]

[hamed abdulaziz]

[hamed abdulaziz]

[hamed abdulaziz]

Where these image are melanoma and the color variation is one of the features to detect this disease

[Image Analyst]

Yes, calculating delta E and then getting the histogram, mean, and std dev of that should be good.

[hamed abdulaziz]

Could you explain by demo please,thanks in advance

[Image Analyst]

This demo shows you how to calculate the delta E image and get its histogram. http://www.mathworks.com/matlabcentral/fileexchange/31118-color-segmentation-by-delta-e-color-difference.

[hamed abdulaziz]

OK, I will accept please could adapt your excellent demo for calculating delta E and then getting the histogram, mean, and std dev for use it with my images ,thanks

[hamed abdulaziz]

Image Analyst : please could you help me for calculating delta E and then getting the histogram, mean, and std dev for use it with my images ,thanks

[Image Analyst]

Make sure you take the mask into account.

pixelsInsideMask = deltaE(mask); % 1D vector of pixels only within mask.
counts = hist(pixelsInsideMask, numberOfBins);
meanDeltaE = mean(pixelsInsideMask(:));
SDDeltaE = std(pixelsInsideMask(:));

[hamed abdulaziz]

Ok , I tried your demo but I need to all operation automatically (not clicking by mouse) what I need is to get the color variation as a indication for disease feature

[Image Analyst]

You just change/adapt it so that it takes the mean color from your whole image (or the masked part). Do you know how to do that?

[hamed abdulaziz]

I don't know how to do that?please could you change the demo to work with my images,thanks.

[Image Analyst]

Find the mask something like this:

% Extract the individual red, green, and blue color channels.
redChannel = rgbImage(:, :, 1);
greenChannel = rgbImage(:, :, 2);
blueChannel = rgbImage(:, :, 3);
% Find where all of them are zero.
mask = redChannel == 0 & greenChannel == 0 & blueChannel == 0;

Then convert rgb to lab and get mean lab like this:

meanL = mean(lChannel(mask));
meanA = mean(aChannel(mask));
meanB = mean(bChannel(mask));

[hamed abdulaziz]

Then how can I get the DELTA E after getting the meanL ,meanA,and meanB ?

[Image Analyst]

Like it shows in the demo:

  % Make uniform images of only that one single LAB color.
  LStandard = LMean * ones(rows, columns);
  aStandard = aMean * ones(rows, columns);
  bStandard = bMean * ones(rows, columns);
  
  % Create the delta images: delta L, delta A, and delta B.
  deltaL = LChannel - LStandard;
  deltaa = aChannel - aStandard;
  deltab = bChannel - bStandard;
  
  % Create the Delta E image.
  % This is an image that represents the color difference.
  % Delta E is the square root of the sum of the squares of the delta images.
  deltaE = sqrt(deltaL .^ 2 + deltaa .^ 2 + deltab .^ 2);
  
  % Mask it to get the Delta E in the mask region only.
  maskedDeltaE = deltaE .* mask;
  % Get the mean delta E in the mask region
  % Note: deltaE(mask) is a 1D vector of ONLY the pixel values within the masked area.
   meanMaskedDeltaE = mean(deltaE(mask));

where meanL is just Lmean by a different name.

[hamed abdulaziz]

how can determine the the number of(rows and columns) in the

LStandard = LMean *

ones(rows, columns);

and can I use the value of deltaE in the

deltaE = sqrt(deltaL

.^ 2 + deltaa .^ 2 +

deltab .^ 2);

as a feature for color variation?

I tried this code

clear all;

clc;

close all;

% rgbImage= imread('D:\final_segmentation\segmentation_abnormal\200AC.BMP');

rgbImage= imread('D:\3.BMP');

% Find the mask something like this:

% Extract the individual red, green, and blue color channels.

redChannel = rgbImage(:, :, 1);

greenChannel = rgbImage(:, :, 2);

blueChannel = rgbImage(:, :, 3);

% Find where all of them are zero.

mask = redChannel ==

0 & greenChannel ==

0 & blueChannel ==

0; % Then convert rgb to lab and get mean lab like this:

% Convert image from RGB colorspace to lab color space.

cform =

makecform('srgb2lab'

);

lab_Image =

applycform(im2double

(rgbImage),cform); % Extract out the color bands from the original image % into 3 separate 2D arrays, one for each color component.

LChannel = lab_Image(:, :, 1);

aChannel = lab_Image(:, :, 2);

bChannel = lab_Image(:, :, 3);

LMean =

mean(LChannel(mask));

aMean =

mean(aChannel(mask));

bMean =

mean(bChannel(mask));

[rows,

columns]=size(LChann

el);

% Make uniform images of only that one single LAB color.

LStandard = LMean *

ones(rows, columns);

aStandard = aMean *

ones(rows, columns);

bStandard = bMean *

ones(rows, columns); % Create the delta images: delta L, delta A, and delta B.

deltaL = LChannel - LStandard;

deltaa = aChannel - aStandard;

deltab = bChannel - bStandard; % Create the Delta E image. % This is an image that represents the color difference. % Delta E is the square root of the sum of the squares of the delta images.

deltaE = sqrt(deltaL .^ 2 + deltaa .^ 2 + deltab .^ 2);

but I always get deltaE = zero

thank you

[Image Analyst]

That's from a line earlier in the demo:

[rows, columns, numberOfColorBands] = size(rgbImage);

[hamed abdulaziz]

THANK YOU I ATTACHED WAHT I TRIED (MY_DELTA_E) CODE HOW CAN USE FOR COLOR VARIATION INDICATOR PLEASE COULD CHECK IT FOR CORRECTNESS THANK YOU

[hamed abdulaziz]

I ATTACHED WAHT I TRIED (MY_DELTA_E) CODE HOW CAN USE DELTA_E FOR COLOR VARIATION INDICATOR PLEASE COULD CHECK IT FOR CORRECTNESS

[Image Analyst]

Looks about right for as far as you got. You just need to complete it to find the variation using code I gave you in my answer:

pixelsInsideMask = deltaE(mask); % 1D vector of pixels only within mask.
counts = hist(pixelsInsideMask, numberOfBins);
meanDeltaE = mean(pixelsInsideMask(:));
SDDeltaE = std(pixelsInsideMask(:));

[hamed abdulaziz]

Please did you check (MY_DELTA_E) CODE wher do I add your code?could you complet it and check it on the above images,thanks

pixelsInsideMask = deltaE(mask); % 1D vector of pixels only within mask.

counts = hist(pixelsInsideMask, numberOfBins);

meanDeltaE = mean(pixelsInsideMask(:));

SDDeltaE = std(pixelsInsideMask(:));