[vtkusers] image in spherical coordinates

[Mark Gooding]

Hi all,

I came across the post below in the archives. I have the same problem,
image data which is in spherical coordinates. I wondered if anyone had
made any progress in this area using vtk in the last couple of years?

It would be nice if someone clever out there would extend vtk to
include a sphericalCoordinatesImage class. Sadly, I'm not that good a
programmer.

I am about to start on a "wasteful" solution to enable me to display
the data correctly, but would appreciate hearing from anyone who has
already developed a solution (however in-elegant).

Cheers,

Mark

----->

[vtkusers] class for image data in polar/spherical coordinate system?
Wu Ruoyun MRYWu at ntu.edu.sg
Tue Mar 30 02:49:25 EST 2004

Dear All,

As you know that most of, if not all, ultrasound machines produce data
in polar (2D scanner) or spherical (3D scanner) coordinate system. VTK
could represent these data using vtkStructuredGrid, as long as each
point has its x,y,z coordinates computed and stored. For a set of 3D
image data, however, there could be over 8 million points, each means
8M x 3 x 2 bytes memory requirement. It seems that it is quite
wasteful to do so.

I am wondering if anyone has a class similar to vtkRectilinearGrid,
but representing data in polar or spherical coordinate system. The
topology and geometry are all regular, although the geometry may not
be rectangular in some or all axis. With that class, we only need to
specify the dimension (i,j,k), plus the three angular/linear
intervals, in order to setup the data structure.

Advice on whether it is easy to make one quickly by following the
codes of vtkRectilinearGrid would also be appreciated.

Best Wishes!
Ruoyun