CGAL users discussion list

[Pol Monsó Purtí Helimap <>]

Thanks Andreas, I'm fiddling with it now.

Considering the tree. What would be the memory impact of having the tree?
Will it provoke a duplication of the point data?

Aside question, is it possible to have a triangle face circulator? As the line face circulator, a method that retrieves all cdt's faces intersected by a triangle.

Cheers,

Pol

On Thu, Mar 24, 2016 at 2:23 PM, Andreas Fabri <> wrote:

On 24/03/2016 13:19, Olivier Devillers wrote:

On 24 Mar 2016, at 12:33, Pol Monsó Purtí Helimap <
<mailto:>> wrote:

Hello all,

How can I intersect a 3d ray with a 2D Constrained Delaunay
Triangulation created from 3d points using the project_xy traits?

I've seen this post:
http://cgal-discuss.949826.n4.nabble.com/Intersection-segment-and-cells-from-3D-Delaunay-triangulation-td4660891.html

But I don't understand andreas for the second approach.

I don't have that many queries, around 200 of them. I might have,
however, lots of points, > 200 milion.

Is it worth it to build a tree? what other approach could I follow?
Iterating over all the faces seems highly inefficient.

for one ray, you could iterate on the faces intersecting the ray in
projection using the line-face-circulator.

Look for
Line Face Circulator
http://doc.cgal.org/4.8/Triangulation_2/classCGAL_1_1Triangulation__2.html#ad720b1a9adc835ed1a27c228eea1e36c

It might still be a rather expensive operation, namely when
you only intersect very few triangles close to mountain peaks,
but traverse all the triangles in the vertical projection
(~sqrt(200 mio)).

Give the AABB tree a try. It will take you an hour to implement.


--
Andreas Fabri, PhD
Chief Officer, GeometryFactory
Editor, The CGAL Project

phone: +33.492.954.912    skype: andreas.fabri

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