How can I count pixel of particular color in image?

I want to count the number of pixels with particular color in the image below. What function can I use to count pixels?

The size of this image is 672*708*3.

Answer 1

[Image Analyst]

For example, to get the number of yellow pixels:

yellowPixels = rgbImage(:,:,1) == 255 & rgbImage(:,:,2) == 255 & rgbImage(:,:,3) == 0;
numYellowPixels = sum(yellowPixels(:));

To get the number of blue pixels:

bluePixels = rgbImage(:,:,1) == 0 & rgbImage(:,:,2) == 0 & rgbImage(:,:,3) == 255;
numBluePixels = sum(bluePixels(:));

[muhaimin mahfudz]

How about 2 dimensional image eg 262144x1 double.

[Image Analyst]

That's a 1-D vector. An RGB image can be thought of as a 2-D image, as opposed to a 3-D volumetric image. Or it can be thought of as a 3-D image because it has 3 color planes and 3 indexes.

I'm not sure how your 1-D vector can be considered to have colored pixels. Please explain or post your image.

[muhaimin mahfudz]

Im trying to compute the amount of blue pixels in the image. However the code did not able to compute due to error states " indexx exceeds matrix dimensions". can advise?

[Image Analyst]

It looks like an indexed or classified image. Just determine that gray level the blue class is and use sum, like

numberOfBluePixels = sum(sum(indexedImage == 2));

if blue has a value 2 in the image.

[muhaimin mahfudz]

pardon me, your codes does not show errors however it does not show the number of pixel for each colour.

[riya shahrin]

how can i determine the number of white, light brown, dark brown, red, gray and black colour pixels in the segmented portion of that picture?

[Image Analyst]

riya, that is not a segmented picture - it's a masked picture. You need to do color segmentation on it so somehow you need to tell it what colors classes are those that you're interested in (the 6 you listed). Then you need some way to classify the colors that are not in your training set so that it can classify them to the nearest of the 6 reference colors. You can use Linear Discriminant Analysis, like my attached demo. If you need more help, please start a brand new question and attach your image and code.

[Iszuzuldin amirull bin mohd janudin]

mask= g_channel > 100, b_channel > 800, r_channel < 800, b_channel < 0, r_channel < 0;

numBlackPixels = sum(~mask(:))

how can i change to yellow mask ?

[Image Analyst]

@Iszuzuldin amirull bin mohd janudin, try using the Color Thresholder app on the Apps tab of the tool ribbon. Segment your image, probably using the HSV color space would be good. then tell it to export the code as a function that you can then call from your program.