Akinlar C., TOPAL C.

IEEE International Conference on Acoustics, Speech and Signal Processing, Kyoto, Japan, 25 - 30 March 2012, pp.1309-1312 identifier identifier

  • Publication Type: Conference Paper / Full Text
  • Doi Number: 10.1109/icassp.2012.6288130
  • City: Kyoto
  • Country: Japan
  • Page Numbers: pp.1309-1312
  • Keywords: Real-Time Circle Detection, Edge Drawing Algorithm, EDLines, Eye Pupil Detection, Circular Traffic Sign Detection
  • Anadolu University Affiliated: Yes


We propose a real-time, parameter-free circle detection algorithm that produces accurate results with very few false positives. The algorithm makes use of the contiguous (connected) set of edge segments produced by our novel edge segment detector, the Edge Drawing (ED) algorithm; hence the name EDCircles. The proposed algorithm first fits line segments to ED's edge segments and then processes these line segments to detect circles in a given image. We show through experimentation that EDCircles works real-time, produces good results, and is very suitable for the next generation real-time automation applications including automatic inspection of manufactured products, human eye iris and pupil detection, circular traffic sign detection, etc.