Label connected components in binary image using non-standard pixel connectivity
1 vue (au cours des 30 derniers jours)
Afficher commentaires plus anciens
Jori Selen
le 23 Mar 2020
Commenté : Sean de Wolski
le 23 Mar 2020
Assume I have the following binary image
BW = [ 1 0 0 0;
0 0 0 0;
0 1 1 0;
0 0 0 0;
0 0 0 0
0 1 1 0 ];
I would like to find all connected components in this binary image, where 2 points are connected if they are at most 2 rows and at most 1 column away from each other and both have the value 1. The function bwlabeln seems to be useful for this task, but it does not provide me the flexibility of specifying the desired pixel connectivity. My problem statement seems to translate to a 5-by-3 pixel connectivity of all ones. Is there any function that can help me here?
The desired labeling of connected components is
L = [ 1 0 0 0;
0 0 0 0;
0 1 1 0;
0 0 0 0;
0 0 0 0
0 2 2 0 ];
and the number of connected components is 2.
0 commentaires
Réponse acceptée
Sean de Wolski
le 23 Mar 2020
Modifié(e) : Sean de Wolski
le 23 Mar 2020
Call imdilate with a custom structuring element of your size. This will then expand the regions so they overlap and bwlabel can then do its thing.
After, do a logical index to remove the added values from imdilate.
Id = imdilate(I, custom_element)
L = bwlabel(Id);
L(~I) = 0;
2 commentaires
Sean de Wolski
le 23 Mar 2020
Hi,
This should work but I may be off by 1 (i.e. test it under other conditions!). More on structuring elements here https://www.mathworks.com/help/images/structuring-elements.html:
BW = [ 1 0 0 0;
0 0 0 0;
0 1 1 0;
0 0 0 0;
0 0 0 0
0 1 1 0 ];
Id = imdilate(BW, ones(2,1));
L = bwlabel(Id);
L(~BW) = 0
Plus de réponses (0)
Voir également
Catégories
En savoir plus sur Read, Write, and Modify Image dans Help Center et File Exchange
Produits
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!