Crate mlsnd

An implementation of the moving least squares point deformation algorithm (Schaefer 2006).

The algorithm takes a set of control points in the origin space, and where those same points should end up in the target (deformed) space. Transformations for new points are calculated based on how close each control point is to the query point.

Re-exports

pub use nalgebra;

Structs

PointMLS

Struct which holds all the information it needs to transform a point, in either direction.

Enums

ConstructionError

PointMLS could not be built because of inappropriate control points (mismatched lengths or empty).

MLSStrategy

Variant of the MLS algorithm. Only Affine has a dimension-agnostic implementation in the paper, so that’s all that’s implemented here.

Traits

Float

Trait covering the necessary features of f32 and f64 for use here

Functions

deform_affine

Free function for transforming a single point.