How to find number of Pixel green colour image by using webcamera

clc
clear
clear all
images= []
webcamlist()
cam= webcam(2)
preview(cam)
pause(2);
images= snapshot(cam)
imshow(images)
r_channel= images(:,:,1);
g_channel= images(:,:,2);
b_channel= images(:,:,3);
images(:,:,1) =0;
images(:,:,3) = 0;
image(images);
mask= g_channel > 100, b_channel > 800, r_channel < 800, b_channel < 0, r_channel < 0;
numBlackPixels = sum(~mask(:))

gr_ratio = double(g_channel)./double(r_channel);
gb_ratio = double(g_channel)./double(b_channel);
rb_ratio = double(r_channel)./double(b_channel);

% gr_ratio(isnan(g_ratio))=0;
% grayImage = rgb2gray(image);
% binaryImage = grayImage > 20;
% binaryImage = bwareaopen(binaryImage, 50);
% binaryImage = imfill(binaryImage, 'holes');
% grayImage(~binaryImage) = 0;
falseMatrix= zeros(7,7);
targetMatrix= ones(7,7);
checkMatrix= ones(7,7);
fprintf('mask has %d black pixels.\n', numBlackPixels);
if targetMatrix == checkMatrix
    %     cam = camStop;
    return
end

[DGM]

What exactly are you trying to do?

This

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

Is the same as this

mask = g_channel > 100, % only this part is assigned to mask
b_channel > 800,  % this does nothing
r_channel < 800,  % this does nothing 
b_channel < 0,    % this does nothing
r_channel < 0;    % this does nothing

which, other than dumping a bunch of unused garbage to console, does the same as this:

mask = g_channel > 100;

It should stand to reason that not all colors described by this mask are green (e.g. [255 255 255]).

One might expect that this is a set of upper and lower thresholds for each color channel being applied, but the values and relations are nonsense for what I'm assuming is a uint8 image.

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

Why test that B and R are above/below 800, when they won't be above 255? How would B or R ever be less than zero?

Since you're asking how to count "green pixels", it's up to you to define the limits of what constitutes "green".

[Iszuzuldin amirull bin mohd janudin]

Thanks for Explaination

Answer 1

[Image Analyst]

Use the Color Threshold app on the Apps tab of the tool ribbon to define "green". Full demo is attached.

[Iszuzuldin amirull bin mohd janudin]

Thank you. Good Sharing

[Iszuzuldin amirull bin mohd janudin]

Hi, may i know how to detect yellow colour for this coparison?.

[Image Analyst]

@Iszuzuldin amirull bin mohd janudin, certainly. To get yellow...

1. Like I said, run the Color Thresholder app on the Apps tab of the tool ribbon. Did you try this?
2. Choose HSV color space.
3. Adjust the hue arrows to select only yellow.
4. Adjust the Saturation thresholds to be about from 0.25 to 1.
5. Click the Export Function button to get the code that you can call from your main code.

[Iszuzuldin amirull bin mohd janudin]

I have tried it, it worked. Thanks for sharing

Answer 2

[yanqi liu]

sir，may be you should upload your image file to develop some function