finding neighbor of a position

6 vues (au cours des 30 derniers jours)
Mohammad Golam Kibria
Mohammad Golam Kibria le 26 Juin 2011
Hi, I have a matrix.
I =
1 0 0
2 5 0
0 0 3
0 0 0
I know the position of 5 in I is 6 linear index.
is there any easy function to have the 8 other neighbors of 5. Thanks

Réponse acceptée

Oleg Komarov
Oleg Komarov le 26 Juin 2011
EDITED: should be fine now
I =[ 1 0 0
2 5 0
0 0 3
0 0 0];
l = 8;
sz = size(I);
% row, col subs of center
[r,c] = ind2sub(sz,l); % c = ceil(l/4); r = mod(l,4)+ c*sz(1);
% Calculate 8 neighbors
neigh(1:8,1:2) = [r+[-1;0;1;-1;1;-1;0;1] c+[-1;-1;-1;0;0;1;1;1] ];
% Only those in the range
neigh = neigh(all(neigh,2) & neigh(:,1) <= sz(1) & neigh(:,2) <= sz(2),:);
% Convert to position
idx = (neigh(:,2)-1)*sz(1) + neigh(:,1);
  3 commentaires
Oleg Komarov
Oleg Komarov le 27 Juin 2011
Hopefuly now is ok. Tested initial and final position.
Mohammad Golam Kibria
Mohammad Golam Kibria le 28 Juin 2011
Thanks

Connectez-vous pour commenter.

Plus de réponses (3)

Sean de Wolski
Sean de Wolski le 27 Juin 2011
idx = find(conv2(double(I==5),ones(3),'same'))
%This includes the 6, but that could easily be taken care of with setdiff.

Wolfgang Schwanghart
Wolfgang Schwanghart le 26 Juin 2011
  3 commentaires
Wolfgang Schwanghart
Wolfgang Schwanghart le 27 Juin 2011
I = [ 1 0 0;
2 5 0;
0 0 3;
0 0 0];
% find the neighbors of the elements where I = 5
I5 = I==5;
[ix,ixn] = ixneighbors(I,I5)
ix =
6
6
6
6
6
6
6
6
ixn =
10
7
2
5
9
1
11
3
% thus, ixn are the linear indices of the neighbors of the indices ix.
% You'll find the values associated with the neighbors by
I(ixn)
ans =
0
0
2
0
0
1
3
0
Mohammad Golam Kibria
Mohammad Golam Kibria le 28 Juin 2011
thanks,this also works fine for me

Connectez-vous pour commenter.


Andrei Bobrov
Andrei Bobrov le 26 Juin 2011
idl = 6;
idxs = ...
nonzeros(bsxfun(@plus,idl - [1 0 -1]',size(I,1)*[-1 0 1]).*[1 1 1;1 0 1;1 1 1])
CORRECTED 06/27/2011 10:05 MSK
idl = 6;
s = size(I);
I0 = zeros(s+2);
I0(2:end-1,2:end-1) = reshape(1:numel(I),s);
idxs = nonzeros(I0(bsxfun(@plus,find(I0==idl) - [1 0 -1]',(s(1)+2)*[-1 0 1])).*[1 1 1;1 0 1;1 1 1])
MORE variant (06/27/2011 11:12 MSK)
s = size(I);
[ii jj] = ind2sub(s,idl);
v = [-1 -1 -1;0 0 0;1 1 1];
R=ii+v;
C=jj+v';
loc = (R<=s(1) & R>=1&C<=s(2) & C>=1&[1 1 1;1 0 1;1 1 1])>0;
idxl = sub2ind(s,R(loc),C(loc));
MORE variant 2 (06/27/2011 11:35 MSK) with idea of Oleg
s = size(I);
[ii jj] = ind2sub(s,idl);
R = ii + [-1 0 1 -1 1 -1 0 1];
C = jj + [-1 -1 -1 0 0 1 1 1];
loc = (R<=s(1) & R>=1&C<=s(2) & C >= 1 )>0;
idxl = sub2ind(size(I),R(loc),C(loc));
LAST variant (06/27/2011 13:43 MSK)
I1 = zeros(size(I));
I1(idl)=1;
idx = find(bwdist(I1,'chessboard')==1)
or
idx = find(bwdist(I==5,'chessboard')==1)
  3 commentaires
Andrei Bobrov
Andrei Bobrov le 27 Juin 2011
Thanks Oleg! Corrected...
Mohammad Golam Kibria
Mohammad Golam Kibria le 28 Juin 2011
Thanks this also works fine for me

Connectez-vous pour commenter.

Catégories

En savoir plus sur Measurements and Feature Extraction dans Help Center et File Exchange

Tags

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by