0

I am looking to be able to plot an stl file in Python. I can do it in Matlab easily enough with this code:

phacon = stlread("Spine_PhaconModel.stl");

hold on

figure = trisurf(phacon,'FaceColor',[0.6 0.6 0.6], 'Edgecolor','none','FaceLighting','gouraud','AmbientStrength', 0.15, 'MarkerEdgeColor',"r")

view([180, -1])

camlight('headlight');

material('dull');

But when I try to change the code to Python, the output is not what I want. The output plot I'm looking for is something like what is attached below: Spine ouput from matlab

I tried using functions such as mesh.Mesh.from_file to get the data as an equivalent of the Matlab function stlread() and plot_trisurf as an equivalent of the Matlab function trisurf(). The code I tried is:

fig = plt.figure()

ax = fig.gca(projection='3d')

stl_data = mesh.Mesh.from_file('Spine_PhaconModel.stl')

points = stl_data.points.reshape([-1, 3])

x = points[:,0]

y = points[:,1]

z = points[:,2]

collec = ax.plot_trisurf(x,y,z,linewidth=0.2)

However, I don't know how to make it look visually the same as the first attached image. What I get is this Spine Phacon output using Python

I would really appreciate the help, thanks so much!

Trenton McKinney
  • 56,955
  • 33
  • 144
  • 158
Maria Smith
  • 11
  • 1

2 Answers2

2

from the documentation it seems you need to plot it as 3d polygons, the answer below shows how to plot it, the following is a modified example because the docs are outdated, and matplotlib seems to have changed since then.

from stl import mesh
from mpl_toolkits import mplot3d
import matplotlib.pyplot as plt

# Create a new plot
figure = plt.figure()
axes = figure.add_subplot(projection='3d')

# Load the STL files and add the vectors to the plot
your_mesh = mesh.Mesh.from_file(r'your_mesh_path.stl')
poly_collection = mplot3d.art3d.Poly3DCollection(your_mesh.vectors)
poly_collection.set_color((0.7,0.7,0.7))  # play with color
axes.add_collection3d(poly_collection)

# Show the plot to the screen
plt.show()

if you want something even better than what matlab draws then you should use vtkplotlib as it draws using your GPU as opposed to matplotlib CPU rasterization, an example is in this answer https://stackoverflow.com/a/57501486/15649230

Edit: I have modified the code above to remove autoscale, now to scale it yourself, you rotate with left mouse button and move with middle mouse button and zoom with right mouse button, and this answer specifies how to do it in code https://stackoverflow.com/a/65017422/15649230

Trenton McKinney
  • 56,955
  • 33
  • 144
  • 158
Ahmed AEK
  • 8,584
  • 2
  • 7
  • 23
  • Hello Ahmed, thank you so much for your answer! The problem I have with the first code is that it appears really small in my python output and I can't find a way to zoom it in so that it looks like the matlab image. With the second thing you suggested (the vtkplotlib ) I get a very good visual but it doesn't have axis. Would you be able to tell me how to solve any of the 2 issues? Thanks so much @AhmedAEK – Maria Smith Dec 02 '22 at 09:46
  • @MariaSmith i have modified the answer for the matplotlib code part, and an edit to help with zooming, the axis on vtk isn't really something that they support, so not easily possible. – Ahmed AEK Dec 02 '22 at 10:51
  • Ahmed thanks so much for your help! I get a much nicer 3D plot now! The only last doubt I have is how can I include lighting so that I get a 3D sensation of the bone! With Matlab it is the function camlight('headlight') what does the trick. Do you know how to implement this in Python? Thanks so much again for everything – Maria Smith Dec 02 '22 at 13:15
  • @MariaSmith matplotlib doesn't implement this ... it can be done manually though https://stackoverflow.com/questions/56864378/how-to-light-and-shade-a-poly3dcollection – Ahmed AEK Dec 02 '22 at 13:19
  • Thanks @AhmedAEK! I had seen that but hadn't know how to implement this line of code rgb = ls.blend_hsv(rgb=ls.shade(facearray, plt.cm.Reds), intensity=ls.shade_normals(normals, fraction=0.25)) with what I have. I think facearray requires a 2D array which I don't know how to get. – Maria Smith Dec 02 '22 at 13:37
0

A better approach would be to use the open3d package to 3D visualize a .stl file. Here's how you can do it:

  • Installation
pip install open3d
  • Usage
import open3d as o3d

mesh = o3d.io.read_triangle_mesh("Body_Kylo_Ren_fixed.stl")
mesh = mesh.compute_vertex_normals()
o3d.visualization.draw_geometries([mesh], window_name="STL", left=1000, top=200, width=800, height=650)

Output: enter image description here You will see a completely interactable 3D stl file.

Check out open3d-documentation to know more.

Musabbir Arrafi
  • 744
  • 4
  • 18