0

I want to detect circles of a shooting target. So the target might have circles on there already, but I am only looking for the NEW bullet holes.

What I was thinking is to start with the target, sort of know where the circles are with the target not being shot. Then once you have that, its almost like greenscreening the current circles so they will not be detected again. From there I would then to pickup only the bullet holes.

How would this be done ? And is this the best approach?

Harry
  • 13,091
  • 29
  • 107
  • 167
  • I don't know if it's the best, but one approach to detecting *new* things (e.g. holes) is to subtract two images that were taken in the same conditions at different points in time. – Martin Maillard Nov 10 '12 at 12:31
  • Could you elaborate on this approach? Im not sticking with my approach that was just the way I thought of it – Harry Nov 10 '12 at 12:32
  • This may help: http://stackoverflow.com/questions/189943/how-can-i-quantify-difference-between-two-images. – Martin Maillard Nov 10 '12 at 12:36

0 Answers0