-2

I'm new to object detection and computer vision, but I'm working on a project where I'm taking pictures of disks and I'm hoping to receive a confidence level. For example, if the disk is kind of round but slightly jagged on the edges it can return "80% circle". Is this possible?

Isabel Zhong
  • 1
  • If you provide it satisfactory criteria for identifying whether something is an "80% circle," then of course it's possible. – Robert Harvey Jan 20 '20 at 23:25

1 Answers1

1

I would check out Hough Circles if I were you. I used this technique for several projects during my MS degree. It works really well and you can set the parameters to give you different margins about what does and doesnt count as a circle. It wont give you a specific confidence level, but there are ways for doing that if thats what youre trying to accomplish. That would be more of a classification problem and you could approach it different ways. Anyway, heres the resource on the Hough Circles...

https://www.pyimagesearch.com/2014/07/21/detecting-circles-images-using-opencv-hough-circles/

Ishiffman120
  • 313
  • 1
  • 14