Label connected components in binary image using non-standard pixel connectivity
1 Ansicht (letzte 30 Tage)
Ältere Kommentare anzeigen
Jori Selen
am 23 Mär. 2020
Kommentiert: Sean de Wolski
am 23 Mär. 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 Kommentare
Akzeptierte Antwort
Sean de Wolski
am 23 Mär. 2020
Bearbeitet: Sean de Wolski
am 23 Mär. 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 Kommentare
Sean de Wolski
am 23 Mär. 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
Weitere Antworten (0)
Siehe auch
Kategorien
Mehr zu Read, Write, and Modify Image finden Sie in Help Center und File Exchange
Produkte
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!