CGAL users discussion list

[Andre Massing <>]

Hi again,


Andre Massing wrote:
> Hi all,
>
> I  discovered a slightly strange behavior when using the AABB tree. It 
> does not discover the right number of intersections between a list of 
> triangles (3) and a given test triangle (3 intersections only at one 
> vertex) *although* the use of the pure CGAL::do_intersect function 
> detects the intersection correctly. It seems to me (only a vague guess) 
> that the foregoing bounding box  intersection for finding good 
> candidates computes some false negatives and throws two actual 
> intersections cases away.
>
> The attached code illustrates the example.

I just wanted to ask, whether the aforementioned described behavior is a 
bug in the AABB tree? Or is there anything else I have to take into 
account or can do about?

Greetings,
Andre