Skip to content

Calculate Best Fit Circle

Use this dialog to calculate the best fit circle and center point, as well as the associated normal vector if needed, from measured points. This can, for example, be point measurements on piles.

Where to Find the Dialog

Right-click in the horizontal list field for points, and select Tools - Calculate Best Fit Circle...

Using the Dialog

To select the relevant points, the program uses proximity evaluation based on a distance criterion specified by the user.

Procedure

  1. Select the relevant points in the horizontal list window.
  2. Right-click and choose Tools - Calculate Best Fit Circle...
  3. Adjust the distance criterion if it does not fit your dataset.
  4. Check what you want to create: circle and/or normal vector.
  5. Select the application layer where the result should be saved.
  6. Click OK to execute the function.

Example with the options for 3D circle as polygon and normal vector enabled

Result

Within each circle group, the program first calculates the best fit plane for the points, as well as the mean error MeanErr_Circle for the height deviations between the points and this plane (will be zero if there are only 3 points). The points are then transformed to the plane, and the best fit circle is calculated using the least squares method, along with the mean error MeanErr_Heights for the deviations between the points and the circle (will be zero if there are only 3 points).

During the calculations, the lines are enriched with the attributes MeanErr_Circle, MeanErr_Heights, and MeanErr_BottomPoint.

The mean error at the bottom point is calculated as follows: MeanErrorBottomPoint = MeanErrorHeight * NormalvectorLength / Radius