Point Cloud Measure geometry node

Measures unorganized points using the characteristics of each point’s neighborhood, and puts the results in attributes.

On this page	Inputs Parameters
Since	20.5

This node can be used to interpret the geometric signatures of unorganized points without needing to convert them into a surface or volume.

Inputs ¶

Points to Measure

The point geometry to compute measurements for.

Optional Points to Sample

Optional input point geometry to sample from when computing measurments.

Parameters ¶

Group

The subset of points in the input geometry to compute measurements on.

Proximity Radius

Determines the maximum extent of each point’s neighborhood. That is, the maximum scale of the features being measured.

Max Neighbors

Determines the maximum number of neighbors each point is permitted to have. Large numbers of neighbors may reduce performance.

Curvature

Measure an estimate of curvature ranging from 0 to 1, taken as the deviation of the point’s neighborhood from a best-fitting tangent plane.

Linear Proportion

Measure the proportion of a linear (1D) structure in the point neighborhood, ranging from 0 to 1.

Planar Proportion

Measure the proportion of a planar (2D) structure in the point neighborhood, ranging from 0 to 1.

Spherical Proportion

Measure the proportion of a spherical (3D) structure in the point neighborhood, ranging from 0 to 1.

Linear Density

Measure the number of points per unit length of the neighborhood’s radius.

Area Density

Measure the number of points per unit area of the neighborhood’s disk.

Volume Density

Measure the number of points per unit volume of the neighborhood’s bounding sphere.

Eigenvalues

Output a vector attribute containing the amount of variance in each point neighborhood’s principal directions. If imagining the principal components as the directions of length, width, and height of the best-fitting ellipsoid, the eigenvalues are the scale of the ellipsoid in each of the principal component directions. The eigenvalues are positive and ordered from greatest to least.

Eigenvectors

Output a matrix attribute containing the three orthogonal principal component unit vectors for each point’s neighborhood. The principal components are ordered by decreasing variance. The principal components can be thought of as estimating the ellipsoid that best fits the point neighborhood, where the first, second, and third eigenvector correspond to the ellipsoid’s direction of length, width, and height, respectively.

Note

If the point cloud corresponds to surface geometry, the principal components can be thought of as estimating the best-fitting tangent plane to the points, where the first and second eigenvectors correspond to the tangents, and the third eigenvector corresponds to the normal vector. Note that the orientation of the normal is ambiguous; point cloud normals with consistent orientation can be built using the Point Cloud Normal node.

Inputs ¶

Parameters ¶

Geometry nodes