Joining label

Hi I have the following matrix

I =

     0     1     0     0     1     0
     0     0     0     1     0     0
     0     0     0     0     0     0
     0     0     1     1     0     0
     0     1     0     0     1     0
     1     0     0     0     0     1

and

L=bwlabel(I)

L =

     0     2     0     0     3     0
     0     0     0     3     0     0
     0     0     0     0     0     0
     0     0     1     1     0     0
     0     1     0     0     1     0
     1     0     0     0     0     1

I need to join or merge those label who have minimum distance 2. Is there any one to help?

[Mohammad Golam Kibria]

label 1 and label 3 are in minimum distance 2 but label 2 is not within minimum distance 2 with any of label 1 or 3.

the out put might be like bellow:

0 2 0 0 1 0

0 0 0 1 0 0

0 0 0 1 0 0

0 0 1 1 0 0

0 1 0 0 1 0

1 0 0 0 0 1

Answer 1

[Jan]

To join the regions, exapand them at first. I assume "minimum distance of 2" means a dilation with a 3x3 matrix:

I = [0     1     0     0     1     0; ...
     0     0     0     1     0     0; ...
     0     0     0     0     0     0; ...
     0     0     1     1     0     0; ...
     0     1     0     0     1     0; ...
     1     0     0     0     0     1];
I2 = imdilate(I, ones(3, 3));
L = bwlabel(I2);
% Now remove all points, which are 0 in the original matrix:
L(I == 0) = 0;

Answer 2

[the cyclist]

It is not clear to me what you mean by "join or merge". Perhaps you can tell us what the correct output is for the example you have given?

The find() command might be part of what you need. For example,

 >> [i,j] = find(L>=2);

will give you the (i,j) coordinates of the values of L greater than or equal to 2.

Answer 3

[Andrei Bobrov]

C = nchoosek(1:max(L(:)),2);
[ii jj c] = arrayfun(@(i1)find(bwdist(L==C(i1,1)).*(L==C(i1,2))),1:size(C,1),'un',0);
leq = C(cellfun(@(x)min(x),c)<=2,:);
for jj = 1:size(leq,1),
    leq(ismember(leq(:,1),leq(jj,2)),1) = leq(jj,1);
    L(L==leq(jj,2)) = leq(jj,1);
end

CORRECTED

I1 = I;
L = bwlabel(I1);
for ii = max(L(:)):-1:1
   c{ii} = bwdist(L==ii)==1;
end
I1(sum(cat(3,c{:}),3) > 1)=1
Lnew = bwlabel(I1)