Class MLSPointCloudInterpolation

Defined in File interpolation_base.h

Inheritance Relationships

Base Type

public pcms::InterpolationBase (Class InterpolationBase)

Class Documentation

class MLSPointCloudInterpolation : public pcms::InterpolationBase 

Meshless Point-Cloud Based Moving Least Square (MLS) Interpolation.

If two point clouds are provided, mesh based accelerated adjacency search cannot be used. Instead, each target points search the whole source point cloud for neighbors within a given radius.

Public Functions

template<typename SourceType, typename TargetType> inline MLSPointCloudInterpolation(pcms::Rank1View<SourceType, pcms::HostMemorySpace> source_points, pcms::Rank1View<TargetType, pcms::HostMemorySpace> target_points, int dim, double radius, unsigned min_req_supports = 10, unsigned degree = 3, bool adapt_radius = true, double lambda = 0.0, double decay_factor = 5.0)

Constructor for point-cloud based MLS interpolation.

For more details about MLS interpolation parameters, refer to the documentation of mls_interpolation and RadialBasisFunction.

Source and target point coordinates are expected to be in a flattened array format. For example, for 3D points, the coordinates should be provided as [x1, y1, z1, x2, y2, z2, …, xn, yn, zn] with n being the number of points.
It constructors the support structure by performing \((N_{targets} \times N_{sources})\) searches.