CGAL users discussion list

[lhc_liuyu <>]

The method to compute the normal vector of each vertex in a any size point cloud data can be found in journal of CAGD; I forget the title of the paper;the details of the method is: find a surface(paraboloid,such as) to fit the vertices in neigbor field (a given radius),and the normal of the a surface can be considered  as the normal of vertex.  

在2009-12-25，"Zheng,Guang" <> 写道：

Hi,

Has anyone compute the normal vector for a large size point cloud data (more than 10M), it seems that the algorithm provided by CGAL is not so efficient for this kind of size dataset. Anyone has any suggestions about this? 

Right now, my computer has been running more than 10 hours (8CPU and 4G RAM, windows 64bit). 

Thanks very much!

--
Guang