How can I count all the black & white pixels from a circular mask passing through a binary image

Hello everybody,

I'm working on a code that uses a circular mask and do a serie of things with it. One of the things I need to do is count all black and white pixels that are inside of this circular mask. I don't know why when I'll count the black pixels inside the mask, I count ALL the black pixels in the image, not only those are inside the mask, but when I do it for the white pixels, it works well.

clc;clear all; close all;

% Carrega imagem
image = imread('img.png');

% Binariza a imagem
level = graythresh(image);
imgBW = im2bw(image,level);

% Identifica as dimensões da imagem
sz = size(imgBW);
[columnsInImage rowsInImage] = meshgrid(1:sz(1), 1:sz(2));

% Cria a máscara circular na imagem

for i=3 % only 3 interations to test
    for j=3
        centerX = i;
        centerY = j;
        radius = 10;

          X = columnsInImage - centerX;
          Y = rowsInImage - centerY;

          circlemask = (X).^2 + (Y).^2 <= (radius).^2;

          circle_image = double(imgBW) .* circlemask;

          angle = atan2(Y, X);
          quadrant1 = circlemask & angle > -pi/2 & angle <= 0;
          quadrant2 = circlemask & angle > -3*pi/2 & angle <= -pi/2;
          quadrant3 = circlemask & angle >  pi/2 & angle <= pi;
          quadrant4 = circlemask & angle > 0 & angle <= pi/2;

          quad1_image = double(imgBW) .* quadrant1;
          quad2_image = double(imgBW) .* quadrant2;
          quad3_image = double(imgBW) .* quadrant3;
          quad4_image = double(imgBW) .* quadrant4;

          ne = length(find(circle_image > 0)) % here I count all white pixels in the whole mask 
          *nw = length(find(circle_image == 0)) % here is the problem...*

          de = ne/nw;

          n(1) = length(find(quad1_image > 0));%here I count all white pixels in each quadrant an it works very well
          n(2) = length(find(quad2_image > 0));
          n(3) = length(find(quad3_image > 0));
          n(4) = length(find(quad4_image > 0));

          ds = min(n)/mean(n);

          matriz_I(i+1,j+1) = de*ds ;
      end
  end

% Plots
subplot(2,4,1);imshow(image);title('Imagem original');
subplot(2,4,2);imshow(imgBW);title('Imagem binarizada');
subplot(2,4,3);imshow(circle_image, []);title('Máscara circular');
subplot(2,4,4);imshow(quad1_image, []);title('Primeiro quadrante');
subplot(2,4,5);imshow(quad2_image, []);title('Segundo quadrante');
subplot(2,4,6);imshow(quad3_image, []);title('Terceiro quadrante');
subplot(2,4,7);imshow(quad4_image, []);title('Quarto quadrante');

[Italo Wieczorek]

PROBLEM SOLVED:

        black_pixels = imgBW; 
        black_pixels(~circlemask) = 1;

        ne = length(find(circle_image > 0))
        nw = length(find(black_pixels == 0))

It turns all the black pixels outside the mask to 1, so you just have to count the black pixels inside the mask. Thanks anyway.