[vtkusers] Intersection of two iso surfaces

[Bill Lorensen]

More like:
contours->GetOutput()->GetPointData()->SetScalars(weights);

On Sat, Dec 3, 2011 at 5:50 PM, Luke <hazelnusse at gmail.com> wrote:
>> 5b) Store the valuve of the second function in the pointdata scalars
>> of the isosurface. Then run vtkContourFilter
>
> Ok, so you mean to do this:
>
> vtkPoints *Points = contours->GetOutput()->GetPoints();
> vtkPolyData* polydata = vtkPolyData::New();
> polydata->SetPoints(Points);
> polydata->GetPointData()->SetScalars(weights);
>
> Where contours is the vtkContourFilter applied to the sampling of the
> first function, and weights would be a vtkDataArray pointer (probably
> vtkFloatArray or vtkDoubleArray) populated with the values of the
> second function at every point on the isosurface of the first
> function?
>
> Also, can you recommend any techniques to help refine the accuracy of
> the points of the first isosurface?  If, possible, I would like to
> ensure that all of those points satisfy f1(x,y,z) = 0 to machine
> precision.  My initial attempt would be to hold two of the three
> constant and use a Newton-Raphson on the function with the third
> coordinate being considered the independent variable used for the
> Newton-Raphson.  It seems like it might work because the initial guess
> provided would be very accurate, but sometimes root finders don't
> converge, so I would need to be careful about protecting against that.
>
> Thanks,
> ~Luke



-- 
Unpaid intern in BillsBasement at noware dot com