VMTK:question about the input of vtkvmtkPolyDataCenterlines

I need to extract center lines from airway mask, and VMTK is a very good tool.

VMTK use vtkvmtkPolyDataCenterlines to extract the center lines, and the usage of vtkvmtkPolyDataCenterlines is:

    centerlineFilter = vtkvmtk.vtkvmtkPolyDataCenterlines()
    centerlineFilter.SetInputData(centerlineInputSurface)
    centerlineFilter.SetSourceSeedIds(inletSeedIds)
    centerlineFilter.SetTargetSeedIds(outletSeedIds)
    centerlineFilter.SetRadiusArrayName(self.RadiusArrayName)
    centerlineFilter.SetCostFunction(self.CostFunction)
    centerlineFilter.SetFlipNormals(self.FlipNormals)
    centerlineFilter.SetAppendEndPointsToCenterlines(self.AppendEndPoints)
    centerlineFilter.SetSimplifyVoronoi(self.SimplifyVoronoi)
    centerlineFilter.SetCenterlineResampling(self.Resampling)
    centerlineFilter.SetResamplingStepLength(self.ResamplingStepLength)
    centerlineFilter.Update()

The inletSeedIds/outletSeedIds indicate the ids of source and target points. My question is: how can centerlineFilter obtain the real coordinate (x, y, z)? How can we input the real coordinate to centerlineFilter? For example, if I have the coordinate of two bifurcation points (calculated by other method), how can I input the bifurcation points to centerlineFilter and extract the center line?

Actually, we can use pointlist to input the real coordinate. More details can be found in vmtkPointListSeedSelector .