Label connected components in binary image using non-standard pixel connectivity
1 visualización (últimos 30 días)
Mostrar comentarios más antiguos
Jori Selen
el 23 de Mzo. de 2020
Comentada: Sean de Wolski
el 23 de Mzo. de 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 comentarios
Respuesta aceptada
Sean de Wolski
el 23 de Mzo. de 2020
Editada: Sean de Wolski
el 23 de Mzo. de 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 comentarios
Sean de Wolski
el 23 de Mzo. de 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
Más respuestas (0)
Ver también
Categorías
Más información sobre Read, Write, and Modify Image en Help Center y File Exchange.
Productos
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!