Comments on: Connected component labeling – Part 5 http://blogs.mathworks.com/steve/2007/05/11/connected-component-labeling-part-5/ 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: max http://blogs.mathworks.com/steve/2007/05/11/connected-component-labeling-part-5/#comment-24822 max Tue, 31 Jan 2012 20:22:06 +0000 http://blogs.mathworks.com/steve/2007/05/11/connected-component-labeling-part-5/#comment-24822 Is there any idea to do labeling the non-binary image. Thank you in advance. :) Is there any idea to do labeling the non-binary image.
Thank you in advance. :)

]]> By: Anisa http://blogs.mathworks.com/steve/2007/05/11/connected-component-labeling-part-5/#comment-24113 Anisa Thu, 21 Apr 2011 12:42:05 +0000 http://blogs.mathworks.com/steve/2007/05/11/connected-component-labeling-part-5/#comment-24113 <pre> Hi Steve, The information u provided is very useful for me.I am finding a connected component in binary image using bwlabel.But I am not getting the proper result.What actually I want to find in face image is nose,eyes,mouth etc.How can I find it? </pre> Hi Steve, The information u provided is very useful for me.I am finding a connected component in binary image using bwlabel.But I am not getting the proper result.What actually I want to find in face image is nose,eyes,mouth etc.How can I find it? ]]> By: Mark http://blogs.mathworks.com/steve/2007/05/11/connected-component-labeling-part-5/#comment-23647 Mark Sat, 23 Oct 2010 03:00:25 +0000 http://blogs.mathworks.com/steve/2007/05/11/connected-component-labeling-part-5/#comment-23647 Thanks Steve :) That clears the point. Thanks Steve :) That clears the point.

]]> By: Steve http://blogs.mathworks.com/steve/2007/05/11/connected-component-labeling-part-5/#comment-23646 Steve Fri, 22 Oct 2010 12:46:14 +0000 http://blogs.mathworks.com/steve/2007/05/11/connected-component-labeling-part-5/#comment-23646 Mark—Unfortunately, intuition doesn't count for much when it comes to algorithm performance on modern memory architectures. The flood-fill approach isn't very cache friendly. Mark—Unfortunately, intuition doesn’t count for much when it comes to algorithm performance on modern memory architectures. The flood-fill approach isn’t very cache friendly.

]]> By: Mark http://blogs.mathworks.com/steve/2007/05/11/connected-component-labeling-part-5/#comment-23645 Mark Fri, 22 Oct 2010 09:14:52 +0000 http://blogs.mathworks.com/steve/2007/05/11/connected-component-labeling-part-5/#comment-23645 Hi Steve, What's the advantage of 2 passes algorithm over flood-fill? I find the latter pretty intuitive. The only thing I can think of is that 2 passes algorithm *might* take less memory, but for common 1000x1000 images it wouldn't be a problem? Hi Steve,

What’s the advantage of 2 passes algorithm over flood-fill? I find the latter pretty intuitive.

The only thing I can think of is that 2 passes algorithm *might* take less memory, but for common 1000×1000 images it wouldn’t be a problem?

]]> By: Steve http://blogs.mathworks.com/steve/2007/05/11/connected-component-labeling-part-5/#comment-23595 Steve Mon, 11 Oct 2010 11:49:12 +0000 http://blogs.mathworks.com/steve/2007/05/11/connected-component-labeling-part-5/#comment-23595 Ariel—You could label the five grayscale classes separately and then use regionprops to determine which have area under 4000. You'll need to write some code yourself to do it. Ariel—You could label the five grayscale classes separately and then use regionprops to determine which have area under 4000. You’ll need to write some code yourself to do it.

]]> By: ariel http://blogs.mathworks.com/steve/2007/05/11/connected-component-labeling-part-5/#comment-23594 ariel Sun, 10 Oct 2010 21:21:25 +0000 http://blogs.mathworks.com/steve/2007/05/11/connected-component-labeling-part-5/#comment-23594 Steve, I want to lable conected components in greyscale image. I used kmeans to the image and it's now in 5 different grayscale classes. I want to lable all components(in diffent grayscale) with an area under 4000 What can I do? Steve, I want to lable conected components in greyscale image.
I used kmeans to the image and it’s now in 5 different grayscale classes. I want to lable all components(in diffent grayscale) with an area under 4000
What can I do?

]]> By: jeevan http://blogs.mathworks.com/steve/2007/05/11/connected-component-labeling-part-5/#comment-23584 jeevan Fri, 08 Oct 2010 09:19:35 +0000 http://blogs.mathworks.com/steve/2007/05/11/connected-component-labeling-part-5/#comment-23584 how to do color image labelling how to do color image labelling

]]> By: Steve http://blogs.mathworks.com/steve/2007/05/11/connected-component-labeling-part-5/#comment-22487 Steve Wed, 16 Dec 2009 11:32:41 +0000 http://blogs.mathworks.com/steve/2007/05/11/connected-component-labeling-part-5/#comment-22487 Fauzi—<tt>regionprops</tt> Fauzi—regionprops

]]> By: Fauzi http://blogs.mathworks.com/steve/2007/05/11/connected-component-labeling-part-5/#comment-22485 Fauzi Wed, 16 Dec 2009 10:42:12 +0000 http://blogs.mathworks.com/steve/2007/05/11/connected-component-labeling-part-5/#comment-22485 Thx Steve. I'm using this CCL method to filter out small blob of noises from silhouette image. What I do is bwlabel the binary image and subtract the labeled component that has area lower than a threshold using a long loop. Is there any shortcut or Matlab code that can compute the area of each of the connected component? It really takes time to finish to filter out the blobs. Thx Steve. I’m using this CCL method to filter out small blob of noises from silhouette image. What I do is bwlabel the binary image and subtract the labeled component that has area lower than a threshold using a long loop.

Is there any shortcut or Matlab code that can compute the area of each of the connected component? It really takes time to finish to filter out the blobs.

]]>