N Largest Empty Circles

[micycle1]

Could org.locationtech.jts.algorithm.construct.LargestEmptyCircle be straightforwardly adapted to find the N largest (ideally non-overlapping) empty circles?

[dr-jts]

Maybe... That would require keeping a growing list of the centres of constructed largest circles, and including their (virtual) circles in the "distance-to" computation for further candidate centres.

Intriguing idea, for sure!