Subject: CGAL users discussion list
List archive
- From: Simon Giraudot <>
- To:
- Subject: Re: [cgal-discuss] Reconstruction from colored point cloud
- Date: Tue, 2 Jul 2019 09:30:10 +0200
- Authentication-results: mail2-smtp-roc.national.inria.fr; spf=None ; spf=None ; spf=None
- Ironport-phdr: 9a23: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
Hello,
What do you mean by filters? If you're talking about functions of CGAL Point Set Processing (for example CGAL::grid_simplify_point_set()), some of them reorder the input range, which means you loose the original indices. That could explain why you're experiencing problems if you rely on the original order.
There's a way to avoid altering the point set order using dedicated property maps: the principle is that you create a range of indices (trivial, just the ordered indices from 0 to N-1, N being your number of points), and use a property map referencing the original range to recover the right point/normal. That way, the functions that reorder the input range will just reorder the range of indices and leave your original point range unaltered (then you can just use the range of indices to access the elements you want). You can see an example that does that in Spatial Searching: https://doc.cgal.org/latest/Spatial_searching/index.html#title12
It may seem a bit tricky but in practice it's very handy.
I hope this helps :)
--
Simon Giraudot, PhD
R&D Engineer
GeometryFactory - http://geometryfactory.com/
Le 01/07/2019 à 15:21, cisant a écrit :
Oh God!! Thanks for the answer!! It works very well if I don't use any
filter. But, if I execute some filter on my point set after read the file
with the XYZ coordinates, I can't write the color correctly.
How could I know the index of a point in order to compare with the archive
that my program already read? Maybe searching by the index, I'll be able to
write the right color for the point.
--
Sent from: http://cgal-discuss.949826.n4.nabble.com/
- Re: [cgal-discuss] Reconstruction from colored point cloud, Simon Giraudot, 07/01/2019
- Re: [cgal-discuss] Reconstruction from colored point cloud, cisant, 07/01/2019
- Re: [cgal-discuss] Reconstruction from colored point cloud, Simon Giraudot, 07/02/2019
- Re: [cgal-discuss] Reconstruction from colored point cloud, cisant, 07/01/2019
Archive powered by MHonArc 2.6.18.