Converting Point cloud to Mesh

Hello, I’m trying to do the following: I have main mesh (obj, off) trying to make some affine transformation on it, badly I had to use other source code to do this, I wasn’t able to reach exactly how to use transformations using vtkplotter, so the point now, after applying the transformation I have points and I set them in .ply file

from vtkplotter import *
pcl = load("/home/roaa/Desktop/PhD/Models/OwnDatabase/cat_noise_0.0001.ply")
print("Number of vertices:",len(pcl.points()))
mesh = delaunay2D(pcl.points())
mesh.write("/home/roaa/Desktop/PhD/Models/OwnDatabase/cat_noise_0.0002.off")
mesh1 = recoSurface(pcl.points())
mesh1.write("/home/roaa/Desktop/PhD/Models/OwnDatabase/cat_noise_0.0003.obj")
print("Number of vertices: ",len(mesh1.points()))
print("Number of faces:",len(mesh1.faces()))

and this what I got:

Number of vertices: 352
 Unknown format /home/roaa/Desktop/PhD/Models/OwnDatabase/cat_noise_0.0002.off file not saved. 
Recalculating normals for 352 Points, sample size= 7
Calculating mesh from points with R = 5.278604972180709
Number of vertices:  248764
Number of faces: 147165

It’s so weird to have a number of vertices and faces, especially that I don’t want to perform anything other than getting the mesh after applying transformation on it. What I may be doing false.

Thank you for your support.

[marcomusy]

I don’t think that vtk can do that , you may try something similar with glyphs

from vedo import *

msh = Mesh(dataurl+"bunny.obj").computeNormals().wireframe()

cpts = msh.cellCenters()
shape = Circle(r=0.001, res=24).rotateY(90)
gl = Glyph(cpts, shape,  orientationArray=msh.celldata["Normals"])
gl.c("blue3")#.lighting('off')

show(gl, axes=1)

but will probably become slow for large meshes.

[marcomusy]

Hi, sorry I forgot I disabled settings.renderPointsAsSpheres = False. Try instead:

from vedo import *

msh = Mesh(dataurl + "teapot.vtk").pointSize(10).renderPointsAsSpheres(False)

show(msh, axes=1)