CGAL users discussion list

["Sebastien Loriot (GeometryFactory)" <>]

Benoît Presles wrote:
> > > Hello Everybody,
> > >
> > > I think a new bug could be present in the "Alpha_shape_2" class (the
> > > problem can be reproduced by using the list of points "fin" attached
> > > in this email).
> > >
> > > First, I compute the alpha shape (GENERAL MODE).
> > > Second, I print the segments which belongs to the alpha shape.
> > > Third, I print the points which belongs to the alpha shape
> > > (Alpha_shape_vertices_iterator).
> > >
> > > The problem is that I get too many points printed, for example the
> > > point (60,50) doesn't belong to the alpha shape but is printed.
> > >
> > > Output of the program:
> > > Reading 11 points from file
> > > Alpha Shape computed
> > > 7 alpha shape edges
> > > 7 0 4 4
> > > 4 4 1 0
> > > 0 0 1 0
> > > 1 0 2.5 0
> > > 2.5 0 4.5 0
> > > 4.5 0 7 0
> > > 16 16 20 20
> > > as_it=4.5 0
> > > as_it=2.5 0
> > > as_it=4 4
> > > as_it=7 0
> > > as_it=1 0
> > > as_it=0 0
> > > as_it=8 12
> > > as_it=16 16
> > > as_it=20 20
> > > as_it=40 40
> > > as_it=60 50
> >
> > It behaves as expected, vertices printed are either REGULAR or SINGULAR
> > in GENERAL mode and REGULAR in REGULARIZED mode.
>
> Maybe I don't understand something but in my point of view, in both 
> cases (GENERAL or REGULARIZED mode), the vertices printed with 
> "Alpha_shape_vertices_iterator" should be the same as the one printed 
> with "Alpha_shape_edges_iterator" and it is not the case here.
A vertex is singular if it's in the alpha complex but
no incident simplex is in the alpha complex.

If you only want to have vertices that are on the boundary but not
singular, you need to use either the REGULARIZED mode (it was designed
for this purpose I guess) or if you need to use the GENERAL mode, just
consider regular vertices.

S.

> Thank you very much for your help,
> Ben_P
>
> > S.