Poisson Surface Reconstruction

This example demonstrates how to perform Poisson surface reconstruction from a point cloud with normals using COMPAS CGAL.

Key Features:

• Loading point clouds with normals from XYZ files

• Poisson surface reconstruction to create a mesh

• Mesh transformation and visualization

• Side-by-side display of input points and reconstructed mesh

import math
from pathlib import Path

from compas.datastructures import Mesh
from compas.geometry import Pointcloud
from compas.geometry import Rotation
from compas.geometry import Scale
from compas_viewer import Viewer
from line_profiler import profile

from compas_cgal.reconstruction import poisson_surface_reconstruction


@profile
def reconstruction_poisson_surface_reconstruction():
    FILE = Path(__file__).parent.parent.parent / "data" / "oni.xyz"

    points = []
    normals = []
    with open(FILE, "r") as f:
        for line in f:
            x, y, z, nx, ny, nz = line.strip().split()
            points.append([float(x), float(y), float(z)])
            normals.append([float(nx), float(ny), float(nz)])

    V, F = poisson_surface_reconstruction(points, normals)
    mesh = Mesh.from_vertices_and_faces(V, F)

    c = Pointcloud(V)

    return c, mesh


points, mesh = reconstruction_poisson_surface_reconstruction()

# ==============================================================================
# Visualize
# ==============================================================================

viewer = Viewer()

viewer.renderer.camera.target = [0, 0, 0]
viewer.renderer.camera.position = [0, -0.2, 2.0]

viewer.scene.add(points, pointsize=10, pointcolor=(255, 0, 0))
viewer.scene.add(mesh, show_points=False)

viewer.show()