Find Circles

Segmentation > Find Circles

Finds circles in the Current Image using an algorithm based on searching within the image’s Hough transform. This function can take some time to complete depending on the size of and number of circles in the image.

1. Method

Sets method for circle finding. Phase Code tends to be faster and more accurate.

• Phase-Code: Uses the phase-code algorithm for circle detection [1]

• Two-Stage: Uses the two-stage algorithm for circle detection [2,3]

2. Polarity

Sets whether circle edges are bright or dark outlines in the image

• Bright: Looks for circles with bright outlines

• Dark: Looks for circles with dark outlines

3. Min. Radius

Minimum radius of circles to be found (Recommended: 10-20)

4. Max. Radius

Maximum radius of circles of the found (Recommended: 20 larger than Min. Radius)

5. Sensitivity

Sets how much contrast needs to be between circles and the background. A higher sensitivity will find more circles, but may increase false positives.

6. Edge Threshold

Another parameter which affects how many circles are found. A lower Edge Threshold finds more circles, but may increase false positives.

Tips

• For best performance we recommend spacing the min and max diameter by 20.
• If a max-min spacing of 20 is too narrow for your application, try running Find Circles over multiple iterations to cover the necessary range, and add the results together using Set Companion Image and Union.

References

[1] T.J Atherton, D.J. Kerbyson. “Size invariant circle detection.” Image and Vision Computing. Volume 17, Number 11, 1999, pp. 795-803.

[2] H.K Yuen, .J. Princen, J. Illingworth, and J. Kittler. “Comparative study of Hough transform methods for circle finding.” Image and Vision Computing. Volume 8, Number 1, 1990, pp. 71–77.

[3] E.R. Davies, Machine Vision: Theory, Algorithms, Practicalities. Chapter 10. 3rd Edition. Morgan Kauffman Publishers, 2005,