CGAL users discussion list

[Hansjoerg Seybold <>]

Does this also apply to the 2d case or only to 3d meshes
Cheers hj

On May 2, 2011 10:44 AM, "Mariette Yvinec" <> wrote:
> Hi
>
> In fact, cgal 3.8, allows to have a fine control on the sizing field.
> I quote the doc below :
>
> /template<class SizingField>
>
> /
>
>
>
> /Mesh_cell_criteria_3<Tr> mcc ( FT radius_edge_bound, SizingField radius_bound);
> /
>
> Returns an object to serve as default criteria for facets. The type/SizingField/ must
> be a model of the concept/MeshDomainField_3 <http://www.cgal.org/Manual/latest/doc_html/cgal_manual/Mesh_3_ref/FunctionObjectConcept_MeshDomainField_3.html#Cross_link_anchor_1501>/. The behavior and semantic of the arguments are the same
> as above, except that the radius bound parameter is a functional instead of a constant.
>
>
> Thus, the parameter SizingField is just required to be a model
> of MeshDomainField_3, that is some scalar function on the domain.
> In particular it is quite possible to set such a function using a separate
> voxel image as you are planing ....
> and the same is true for the SizingField parameter in Mesh_facet_criteria_3.
>
> Hope this helps,
> Mariette
>
>
>
> Le 02/05/11 01:54, toddoe a écrit :
>> HI All,
>> hope this isn't too silly of a request, but I would really like to control
>> the mesh sizing using a separate voxel image, rather than an analytical
>> function.
>>
>> The reason for this is because i've written a bunch of nice mesh sizing
>> tools in Matlab that use gradient limiting and other mesh or problem
>> specific features that i need. The way I do it now is to generate a
>> supporting 'image' of the voxel object where each sizing voxel contains the
>> local target mesh size (i do this all in Matlab, which i can 'understand'...
>> i've written my own crude tet mesher but want to use CGAL).
>>
>> I have compiled cgal 3.8 and have 3d meshing working quite well (win7,
>> 64bit). Now if I could just locally control the mesh size that would be
>> amazing.
>>
>> I've read the docs and the discussions here and understand that I can set
>> mesh sizes for different domains, but I really want to control mesh size
>> within the domain, and i can't use an analytical function because the mesh
>> size depends upon a non-analytical distribution (viz. the original voxel
>> image intensity, actually bone density, just for one example). Another
>> application is aneurysm, which is a really difficult meshing problem because
>> the vessel wall gets really thin (need local sizing control, some
>> preliminary pics attached below).
>>
>> I hope this explanation makes sense.
>> Basically, I want to control the local mesh size within a domain with a
>> separate voxel image.
>>
>> Any help or ideas are greatly appreciated. I'm only a beginner at C++
>> (really good at matlab and python) but I understand most basic concepts. At
>> least enough that I've been able to modify the examples and add input-output
>> and basic gui functionality.
>>
>> Thanksinadvance,
>> todd
>>
>> p.s. 3-domain mesh of a brain aneurysm... latest fiddling around. the vessel
>> wall is 'transparent' but right now too thick because of lack of local
>> sizing control.
>> about 200k elements.
>> http://cgal-discuss.949826.n4.nabble.com/file/n3489004/brainan1.jpg
>> http://cgal-discuss.949826.n4.nabble.com/file/n3489004/brainan2.jpg
>> http://cgal-discuss.949826.n4.nabble.com/file/n3489004/brainan3.jpg --
>> View this message in context: http://cgal-discuss.949826.n4.nabble.com/Mesh-sizing-using-a-separate-sizing-voxel-image-tp3489004p3489004.html
>> Sent from the cgal-discuss mailing list archive at Nabble.com.
>>
>
> --
> Mariette Yvinec
> Geometrica project team
> INRIA Sophia-Antipolis
>
>
>
>
> --
> You are currently subscribed to cgal-discuss.
> To unsubscribe or access the archives, go to
> https://lists-sop.inria.fr/wws/info/cgal-discuss
>