Comments on: Connected component labeling – Part 3 http://blogs.mathworks.com/steve/2007/03/20/connected-component-labeling-part-3/ Steve Eddins manages the Image & Geospatial development team at The MathWorks and coauthored Digital Image Processing Using MATLAB. He writes here about image processing concepts, algorithm implementations, and MATLAB. Fri, 10 Feb 2012 18:55:59 +0000 hourly 1 http://wordpress.org/?v=3.2.1 By: Gulcan http://blogs.mathworks.com/steve/2007/03/20/connected-component-labeling-part-3/#comment-23814 Gulcan Mon, 27 Dec 2010 21:01:38 +0000 http://blogs.mathworks.com/steve/2007/03/20/connected-component-labeling-part-3/#comment-23814 Sorry for the previous comment. I thought I was writing for the post titled "Almost-connected-component labeling". I posted my comment there as well, thanks. Sorry for the previous comment. I thought I was writing for the post titled “Almost-connected-component labeling”. I posted my comment there as well, thanks.

]]> By: Gulcan http://blogs.mathworks.com/steve/2007/03/20/connected-component-labeling-part-3/#comment-23812 Gulcan Mon, 27 Dec 2010 20:56:35 +0000 http://blogs.mathworks.com/steve/2007/03/20/connected-component-labeling-part-3/#comment-23812 Hello Steve, I have a kinda-similar problem. I am trying to detect buildings and obtain a labeled image after segmentation. I have also binary mask of buildings and I want to train my system for each building object. Thus, I have a labeled image and a binary mask for the desired objects. However, there are several labels in the region of corresponding component in binary mask and I want to merge these regions in labeled image and give them a single label for each object. I mean, I want to merge building segment labels which I know they are part of a building object indeed. I guess traversing all the pixels of objects in the binary mask is the most cumbersome way, but I can't think of any other thing. Any suggestions will be appreciated. Thanks a lot. Hello Steve,
I have a kinda-similar problem. I am trying to detect buildings and obtain a labeled image after segmentation. I have also binary mask of buildings and I want to train my system for each building object. Thus, I have a labeled image and a binary mask for the desired objects. However, there are several labels in the region of corresponding component in binary mask and I want to merge these regions in labeled image and give them a single label for each object. I mean, I want to merge building segment labels which I know they are part of a building object indeed. I guess traversing all the pixels of objects in the binary mask is the most cumbersome way, but I can’t think of any other thing. Any suggestions will be appreciated. Thanks a lot.

]]> By: huambrillo http://blogs.mathworks.com/steve/2007/03/20/connected-component-labeling-part-3/#comment-23495 huambrillo Thu, 16 Sep 2010 17:33:52 +0000 http://blogs.mathworks.com/steve/2007/03/20/connected-component-labeling-part-3/#comment-23495 Mr. Steve , why did you have to put 1s on the diagonal of A? Mr. Steve , why did you have to put 1s on the diagonal of A?

]]> By: Steve http://blogs.mathworks.com/steve/2007/03/20/connected-component-labeling-part-3/#comment-23490 Steve Thu, 16 Sep 2010 12:18:02 +0000 http://blogs.mathworks.com/steve/2007/03/20/connected-component-labeling-part-3/#comment-23490 Jorge—What do want to know about it? Jorge—What do want to know about it?

]]> By: jorge http://blogs.mathworks.com/steve/2007/03/20/connected-component-labeling-part-3/#comment-23486 jorge Thu, 16 Sep 2010 05:35:02 +0000 http://blogs.mathworks.com/steve/2007/03/20/connected-component-labeling-part-3/#comment-23486 do you know about K nearest neighbors? can you explain this into a post? http://www.mathworks.com/matlabcentral/fileexchange/15562-k-nearest-neighbors do you know about K nearest neighbors? can you explain this into a post?
http://www.mathworks.com/matlabcentral/fileexchange/15562-k-nearest-neighbors

]]> By: Steve http://blogs.mathworks.com/steve/2007/03/20/connected-component-labeling-part-3/#comment-23463 Steve Fri, 10 Sep 2010 17:21:18 +0000 http://blogs.mathworks.com/steve/2007/03/20/connected-component-labeling-part-3/#comment-23463 Stefano—To compute the connected components of the graph represented by the adjacency matrix. Stefano—To compute the connected components of the graph represented by the adjacency matrix.

]]> By: stefano http://blogs.mathworks.com/steve/2007/03/20/connected-component-labeling-part-3/#comment-23461 stefano Fri, 10 Sep 2010 17:15:44 +0000 http://blogs.mathworks.com/steve/2007/03/20/connected-component-labeling-part-3/#comment-23461 ok Steve thanks anyway. but why do you have to put 1s on the diagonal of A so that you put dmperm? . what is the propose of dmperm? why do you use this command? ok Steve thanks anyway. but why do you have to put 1s on the diagonal of A so that you put dmperm? . what is the propose of dmperm? why do you use this command?

]]> By: Steve http://blogs.mathworks.com/steve/2007/03/20/connected-component-labeling-part-3/#comment-23453 Steve Thu, 09 Sep 2010 16:25:28 +0000 http://blogs.mathworks.com/steve/2007/03/20/connected-component-labeling-part-3/#comment-23453 Stefano—I don't understand your first question. As to the second, I am not personally familiar the details of the dmperm algorithm; you'll need to research that yourself. Stefano—I don’t understand your first question. As to the second, I am not personally familiar the details of the dmperm algorithm; you’ll need to research that yourself.

]]> By: stefano http://blogs.mathworks.com/steve/2007/03/20/connected-component-labeling-part-3/#comment-23452 stefano Thu, 09 Sep 2010 15:25:53 +0000 http://blogs.mathworks.com/steve/2007/03/20/connected-component-labeling-part-3/#comment-23452 hi Steve! how can you “find” the number of nodes with the “find” function? Respect your dmperm ( dulmage-mendelsohn descomposition) I dont understand. Can you explain that algorithm? thank you in advance hi Steve! how can you “find” the number of nodes with the “find” function?

Respect your dmperm ( dulmage-mendelsohn descomposition) I dont understand. Can you explain that algorithm? thank you in advance

]]> By: Steve http://blogs.mathworks.com/steve/2007/03/20/connected-component-labeling-part-3/#comment-23113 Steve Fri, 14 May 2010 11:41:27 +0000 http://blogs.mathworks.com/steve/2007/03/20/connected-component-labeling-part-3/#comment-23113 Amey—The Image Processing Toolbox does not have such a function. Amey—The Image Processing Toolbox does not have such a function.

]]>