CGAL users discussion list

[Pádraig Ó Conbhuí <>]

@vimarsh: http://doc.cgal.org/latest/Surface_mesher/Surface_mesher_2mesh_a_3d_gray_image_8cpp-example.html#a5 should be a good place to start if you have a gray level image in an inr format, or can generate an inr image from your data.

If you can write an interpolating function for your data, you could use http://doc.cgal.org/latest/Surface_mesher/Surface_mesher_2mesh_an_implicit_function_8cpp-example.html#_a1.

It might be worth looking at the CGAL::Image_3 class implementation to see how it's done there. You might also be able to use the Image_3 class to represent your gridded data (ie construct an Image_3 with the same dimensions/values as your data), then use the value function to get some interpolated values from it, and finally wrap that in a function or class so Implicit_surface_3 can use it.

@gunshi: One approach you could take is to triangulate your data and write some interpolation function over your triangles/tetrahedra. You could use an AABB tree to do efficient lookup of tetrahedra that intersect a query point and use a std::map of vertex handles to values to get the values at each vertex of the found tetrahedron. You can then use barycentric coordinates to determine the interpolated value inside the tetrahedron ad your query point.

I haven't done something like that myself yet, so I can't give you any decent pointers on it, but that's how I'd start it. A more efficient approach might involve using vertices in your triangulation that also include an index, so you could keep a list of vertex values instead of a map for quicker lookup.

Regards,

Paddy

On Fri, Nov 13, 2015 at 5:40 PM, gunshi <> wrote:

Hi, i had the same problem, i have a distribution of points where some points
have negative scalar values, and some have positive scalar values,and i want
the zero level isosurface.
This doesn't exactly fit the 'point set/cloud' description or else i'd have
used the point set processing module (because from what i understand, in
that case we should only have points which we know are definitely inside the
object, and the package functions find the best fitting contour for those
points).
Any chance you've found a way to do this?
Thanks.



--
View this message in context: http://cgal-discuss.949826.n4.nabble.com/isosurface-construction-from-signed-distance-function-point-set-tp4661319p4661337.html

Sent from the cgal-discuss mailing list archive at Nabble.com.

--
You are currently subscribed to cgal-discuss.
To unsubscribe or access the archives, go to
https://sympa.inria.fr/sympa/info/cgal-discuss