Number of black pixels with no black pixel as right neighbour

Question

I'm trying to find the number of black pixels with NO black pixel neighbours at any of right, upper right, lower right .

mask1 = [0|1 0|1 0; 0|1 0|1 0; 0|1 0|1 0];
black = double (x28462847_1_1 < 128);
result = conv2 (black, mask1, 'same');
count = sum (result(:) == 1)

I tried using this mask. Does anyone know where I'm going wrong

Answer 1

I think the best way of doing this is:

notblack=black>0; %if its not zero, then its not black

mask=[0 1 0; 1 0 1; 0 1 0]% or mask=[0 0 1; 0 0 1; 0 0 1]; in your case.

result = conv2(double(notblack),mask, 'same'); % sum all values


count=sum(result(:)==sum(mask(:)));

Answer 2

To detect all black pixels (ones) which has no black pixels at any of its 3 (upper/lower) right pixels, you can calculate the cross correlation with the following mask:

mask = [0  0  -1; 
        0  1  -1; 
        0  0  -1];

The result will be 1 for the desired positions and smaller for the others.

result = conv2 (double(black), rot90(mask,2), 'same') == 1; % Note that I used `conv2` with `rot90` to calculate the cross correlation. 
count = sum (result(:)); % count the number of desired elements

Example

For the following image

rng('default')
black = rand(10) > 0.8;

I obtain the following result:

figure
imshow(~black, 'InitialMagnification','fit')
hold on
[x, y] = find(result);
plot(y, x, '*') % highlight the found positions with an asterisk