我一直在看cgal绑定。
我有一个8000 x，y，z坐标的numpy数组
[991.495 , 268.0944, 192.412 ],
[991.496 , 268.0823, 192.4133],
[991.4993, 268.1071, 192.4103],
我想使用CGAL，我正在尝试使用简单的triangulation_3 https://github.com/CGAL/cgal-swig-bindings/wiki/BindingsExamples
最终，我想将点云转换为多面体并将其保存为xyz文件，但我想一次只能执行一次。
我已经安装了所有依赖项，并运行了一些示例。
这是cgal示例：
from CGAL.CGAL_Kernel import Point_3
from CGAL.CGAL_Triangulation_3 import Delaunay_triangulation_3
from CGAL.CGAL_Triangulation_3 import Delaunay_triangulation_3_Cell_handle
from CGAL.CGAL_Triangulation_3 import Delaunay_triangulation_3_Vertex_handle
from CGAL.CGAL_Triangulation_3 import VERTEX
from CGAL.CGAL_Triangulation_3 import Ref_Locate_type_3
from CGAL.CGAL_Kernel import Ref_int
L=[]
L.append( Point_3(0,0,0) )
L.append( Point_3(1,0,0) )
L.append( Point_3(0,1,0) )
T=Delaunay_triangulation_3(L)
n=T.number_of_vertices()
V=[]
V.append( Point_3(0,0,1) )
V.append( Point_3(1,1,1) )
V.append( Point_3(2,2,2) )
n = n + T.insert(V)
assert n==6
assert T.is_valid()
lt=Ref_Locate_type_3()
li=Ref_int()
lj=Ref_int()
p=Point_3(0,0,0)
c = T.locate(p, lt, li, lj)
assert lt.object() == VERTEX
assert c.vertex(li.object()).point() == p
v = c.vertex( (li.object()+1)&3 )
nc = c.neighbor(li.object())
nli=Ref_int()
assert nc.has_vertex( v, nli )
T.write_to_file("output")
T1 = Delaunay_triangulation_3()
T1.read_from_file("output")
assert T1.is_valid()
assert T1.number_of_vertices() == T.number_of_vertices()
assert T1.number_of_cells() == T.number_of_cells()