CGAL users discussion list

[Markus Eich <>]

Thank you Andreas,

So lets say I create my alpha shapes
================
    // compute alpha shape
    Alpha_shape_2 as(lp.begin(),lp.end());
================
than I get my first edge from the alpha_shape:

=======================
    Dt::Edge edge,next_edge;
    Alpha_shape_edges_iterator iter =  as.alpha_shape_edges_begin();
    edge = *iter;

==========================
And now I want to get the adjacent edge (say next_edge in clockwise 
order). How can I get it? I might look stupid but I don't get it how to 
circulate the edges of a alpha shape in a sorted order.

Mayby you can point me to an example?

Thank you

Markus







On 06.06.2011 14:07, Andreas Fabri wrote:
> Edge is a std::pair<Face_handle,int>
>
> From the face you can get the two vertices
> with index different from the index in the pair.
>
> http://www.cgal.org/Manual/latest/doc_html/cgal_manual/TDS_2_ref/Concept_TriangulationDSFaceBase_2.html#Cross_link_anchor_1316 
>
>
> andreas
>
>
> On 06/06/2011 13:27, Markus Eich wrote:
> > Thank you for the input. My question is now how can I access the
> > neighbors of an edge using CGAL?
> >
> > Is there an overridden function like as.segment(*it).get_neighbor() ?
> > The link
> >
> > http://www.cgal.org/Manual/latest/doc_html/cgal_manual/TDS_2_ref/Concept_TriangulationDSFaceBase_2.html#Cross_link_anchor_1316 
> >
> >
> >
> > doesn't tell me anything (sorry i am a CGA newbie). If I have the
> > segment A from the Alpha_shape_edges_iterator, how can I get the
> > adjactend Vertices. How can I extend my little example to get the two
> > adjacend edges?
> >
> > ////////////////////
> > for(Alpha_shape_edges_iterator it = as.alpha_shape_edges_begin();it !=
> > as.alpha_shape_edges_end(); ++it) {
> > segments.push_back(as.segment(*it));
> >
> >
> >
> > }
> >
> > ////////////////////
> >
> > Cheers,
> >
> > Markus
> >
> >
> >
> > On 06.06.2011 08:55, Sebastien Loriot (GeometryFactory) wrote:
> > > Markus Eich wrote:
> > > > Dear all,
> > > >
> > > > I am using CGAL in order the get one or more alpha polygons from a
> > > > shape. IN a first approach I assume there are no holes. I generate
> > > > the segements as follows:
> > > >
> > > >
> > > > Alpha_shape_2 as(lp.begin(),lp.end());
> > > >
> > > > // find optimal alpha value
> > > >
> > > > Alpha_iterator opt = as.find_optimal_alpha(1);
> > > > as.set_alpha((*opt)*2.0f);
> > > > as.set_mode(Alpha_shape_2::REGULARIZED);
> > > >
> > > > std::vector<K::Segment_2> segments;
> > > > std::vector<K::Segment_2> sorted_segments;
> > > >
> > > > for(Alpha_shape_edges_iterator it = as.alpha_shape_edges_begin();it
> > > > != as.alpha_shape_edges_end(); ++it)
> > > > segments.push_back(as.segment(*it));
> > > >
> > > > std::cout << segments.size() << " alpha shape edges" << std::endl;
> > > >
> > > >
> > > > This works fine an I get all the segments. My problem is that I need
> > > > a single polygon (in my example I only have one concave shape). If I
> > > > iterate over the std::vector<K::Segment_2> segments vector, the
> > > > segments are like E,A,V,D,C,F , instead of A,B,C,D,E,F. In need the
> > > > segments in an ordered way i.e. that the .target() of A is the
> > > > .source() of B, the .target() of B the source of C, and so on.
> > > >
> > > > How can I achieve this using CGAL? Is there a way to sort the
> > > > segments directly?
> > > There is nothing to do it directly in CGAL.
> > >
> > > This is a simple graph problem: each edge has 2 vertices and two
> > > adjacent edges share one identical vertex. In particular each such
> > > vertex is adjacent to two other vertices. Starting from one vertex
> > > and going from adjacent vertex to adjacent vertex you can reconstruct
> > > the polygon.
> > >
> > > You can use for example a
> > > std::map<Vertex_handle,std::vector<Vertex_handle> >
> > > to get the adjacency between vertices.
> > >
> > > For the definition of an edge as std::pair<Face_handle,int>, see:
> > > http://www.cgal.org/Manual/latest/doc_html/cgal_manual/TDS_2_ref/Concept_TriangulationDSFaceBase_2.html#Cross_link_anchor_1316 
> > >
> > >
> > >
> > >
> > >
> > >
> > > S.
> > >
> > >
> > >
> > >
> > > > Thank you and cheers,
> > > >
> > > > Markus


--
Dipl. Inf. Markus Eich
Researcher

DFKI Bremen
Robotics Innovation Center
Mary-Somerville-Str. 9
28359 Bremen, Germany

Phone: +49 (0)421 17845-4105
Fax:   +49 (0)421 17845-6613
E-Mail: 


Weitere Informationen: http://www.dfki.de/robotik
-----------------------------------------------------------------------
Deutsches Forschungszentrum fuer Kuenstliche Intelligenz GmbH
Firmensitz: Trippstadter Straße 122, D-67663 Kaiserslautern
Geschaeftsfuehrung: Prof. Dr. Dr. h.c. mult. Wolfgang Wahlster
(Vorsitzender) Dr. Walter Olthoff
Vorsitzender des Aufsichtsrats: Prof. Dr. h.c. Hans A. Aukes
Amtsgericht Kaiserslautern, HRB 2313
Sitz der Gesellschaft: Kaiserslautern (HRB 2313)
USt-Id.Nr.:    DE 148646973
Steuernummer:  19/673/0060/3
-----------------------------------------------------------------------