Only analyze what is outlined in black in an image

Hi, I have to write a code that only considers what is within a certain pattern of circles in a black and white image. To do this, I overlayed an image of a bracket of circles outlined in black over the picture that I want to analyze, and now I only want to get data about the parts of the picture that are within the circles. (Specifically, I want to count the number of black pixels vs white pixels within all the circles). Please let me know if there is a way to disregard the parts of the picture that are not outlined by the circles.

Thank you!

Answer 1

[Matt J]

If you have a binary mask of the circles, you can just do

numBlack = nnz( circles & ~BW );
numWhite = nnz( circles & BW);

[Dinuri Rupasinghe]

Thank you for your response! So, I tried to make a binary mask of the circles, and this is what I have:

circles = imfindcircles(pic,[10 50])
numBlack = nnz( circles & ~BW );
numWhite = nnz( circles & BW);

But I get this error:

[Matt J]

imfindcircles does not generate a binary mask.

[Dinuri Rupasinghe]

How can I generate a binary mask?

[Matt J]

If you have the radius R and the center coordinates xc,yc, then the mask for one circle can be created as

nx=100; ny=170; %grid dimensions
xc=40; yc=75; R=20;

circleMask=((1:nx).'-xc).^2 + ((1:ny)-yc).^2<=R^2;

imshow(circleMask)

[Dinuri Rupasinghe]

Thank you for your help. I am concerned that I am doing something wrong with the binary mask because the percentage of black pixels I am getting is over 100. I only want to count the number of black pixels within each circle.

numcirc = size(centers)
dim = size(I)
no_of_pix = dim(1)*dim(2)
numWhite = 0
numBlack = 0

% Calculate the number of black pixels in each circle
for circles = 1:numcirc(1)
    nx=dim(1); ny=dim(2); %dimensions of entire image
    xc=centers(circles,1); yc=centers(circles,2); R=10;
    circleMask=((1:nx).'-xc).^2 + ((1:ny)-yc).^2<=R^2;
    imshow(circleMask)
    numBlack = numBlack + nnz(circles & ~BW);
    numWhite = numWhite + nnz(circles & BW);
end

% Calculate total percentage of regolith collected
percent_black = (numBlack/no_of_pix)

disp('percent black pixels:')
disp(percent_black)

[Matt J]

Perhaps as follows:

numcirc = size(centers)
dim = size(I)
no_of_pix = dim(1)*dim(2)


allCircles=false(dim);

% Calculate the number of black pixels in each circle
for i = 1:numcirc(1)
    nx=dim(1); ny=dim(2); %dimensions of entire image
    xc=centers(i,1); yc=centers(i,2); R=10;
    allCircles=allCircles | ((1:nx).'-xc).^2 + ((1:ny)-yc).^2<=R^2;
end

    numBlack = numBlack + nnz(allCircles & ~BW);
    numWhite = numWhite + nnz(allCircles & BW);
    
percent_black = (numBlack/no_of_pix)

disp('percent black pixels:')
disp(percent_black)

[Image Analyst]

@Dinuri Rupasinghe, please attach the original image you're dealing with.

[Dinuri Rupasinghe]

@Matt J it says "arrays have incompatible sizes for this operation" on the line

 numBlack = numBlack + nnz(allCircles & ~BW);
 numWhite = numWhite + nnz(allCircles & BW);

@Image Analyst This is the original image

[Matt J]

I and BW need to be the same size.

[Dinuri Rupasinghe]

This worked. Thank you!