I have a problem with defect detection at the edge of a circular object. It's hard to describe, so I have a photo that might help a little. I am trying to find red marked areas, for example, what is shown below:

I have already tried matching patterns vision.TemplateMatcher(), but this only works for the image I made in the pattern.
I tried to match it with vision.CascadeObjectDetector(), and I trained it with 150 images. I found only <5% of the correct results with this.
I also tried matching with detectSURFFeatures()and then matchFeatures(), but this only works for very similar defects (when the ribs are not closed, it fails).
Since the defects are close to half the circle, I tried to find it using imfindcircles(), but there I find so many possible results. When I take the one with the highest metric, sometimes I get the right one, but I donβt get close to 30%.
Do any of you have an idea that I can try to find at least more than 50%?
If someone has an idea and you want to try something, I added another picture. Since I'm new, I can only add two photos, but if you need more, I can provide more photos.


source
share