I'm looking for the way to detect the tree rings automatically in the wood dist image. Using the matlab, I have removed the background, and binarized the image.

How can I detect the white and black boundaries? I want to fill that white area.

  1. Is there a way to detect the white and black boundaries?
  2. I want to get rid of the black noise that exists in the white area. is there any other way to make clearer binary image?
  3. It has scratches which has caused by sandpaper machine, so morphological operation doesn't work well.. any other way..?

I would appreciate your help.

enter image description here

  • 1
    $\begingroup$ Phew, tough one. Some kind of "fuzzy" version of the Hough transform to detect the circles maybe (fuzzy Hough transform dies yield a few hits on google)? Or look at radial cuts (once you found the center), try to find the troughs and track them along different cut angles? $\endgroup$
    – Florian
    Jul 16 '19 at 15:07
  • $\begingroup$ Are you just trying to "whiten" the white rings or is the ultimate goal to enumerate the rings? For the former, I am surprised nobody has said "2D Median filter" en.wikipedia.org/wiki/Median_filter For the latter, I have some ideas, but no point testing them if that is not the intent. $\endgroup$ Jul 18 '19 at 20:18
  • 1
    $\begingroup$ You started by binarizing, then try to improve your image. This is always the wrong approach. Instead, first improve your image, and don't binarize until the very end, when the result of the threshold is exactly what you need. It is a lot easier to get useful data by filtering a gray-scale or color image than a binary image. Because binarizing throws away a lot of useful information. $\endgroup$ Aug 15 '19 at 21:31
  • $\begingroup$ To the point made by @CrisLuengo: please post one or more examples of your starting image. $\endgroup$
    – TimWescott
    Apr 12 '20 at 2:01
  • 1
    $\begingroup$ This is a relevant PhD thesis: pub.epsilon.slu.se/2274 $\endgroup$ 15 hours ago

It looks like you could use a low pass filter or another noise reduction filter. Then maybe an edge detection filter.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.