How do I remove islands in a non-binary greyscale image?

6 views (last 30 days)
I understand that bwareaopen() serves as a similar function however using the command yields a binary image when I need to retain the original intensity range.

Accepted Answer

Cyrus N
Cyrus N on 27 Jun 2017
Edited: Cyrus N on 28 Jun 2017
the trick is to use the bwareaopen() which actually removes the islands then use the binary image to cut out the original islands using OrigionalImageName(~BinaryImageName) = 0;. This uses the island free binary image to cut the islands out of the greyscale original image.
  1 Comment
Image Analyst
Image Analyst on 27 Jun 2017
I've never heard of "bwarefunction()" - it's not built into the Image Processing Toolbox.
To remove "islands" I'd first get the binary image of the islands somehow, for example by thresholding, then I'd call regionfill() to smear the surrounding values in to replace the islands with gray levels.
Also, we don't know if your islands are brighter or darker than the surround because you forgot to attach an image.
I'm not sure of your definition of removing islands, but it looks like it replaces them with zeros. Not sure why that is called a removal.

Sign in to comment.

More Answers (0)

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!