A parameter set for ICP matching and down sampling point clouds. More...

Public Member Functions
None	MatchingParameters (self, int max_scene_points=0, int max_model_points=0, int max_iterations=20, float tolerance=1e-9, float min_improvement=1e-5, float correspondence_threshold=0.8, float convergence_radius=1e-4, bool pre_align=False, int correspondence_method=cvb.match_3d.CorrespondenceType.ConsensusICP)
	Instantiates a MatchingParameters object.

Properties
	convergence_radius = property
	float: Gets or sets convergence radius to define when two points are considered as a converged match.

	correspondence_method = property
	int: Gets or sets correspondence type used (see cvb.match_3d.CorrespondenceType).

	correspondence_threshold = property
	float: Gets or sets parameter for shape sensitivity.

	max_iterations = property
	int: Gets or sets maximum number of iterations to execute.

	max_model_points = property
	int: Gets or sets maximum number of model points to use.

	max_scene_points = property
	int: Gets or sets maximum number of scene points to use.

	min_improvement = property
	float: Gets or sets threshold for minimum improvement for early stopping criteria.

	pre_align = property
	bool: Gets or sets pre-alignment configuration.

	tolerance = property
	float: Gets or sets tolerance for stopping criteria.

Detailed Description

A parameter set for ICP matching and down sampling point clouds.

Constructor & Destructor Documentation

◆ MatchingParameters()

None MatchingParameters	(		self,
		int	max_scene_points = 0,
		int	max_model_points = 0,
		int	max_iterations = 20,
		float	tolerance = 1e-9,
		float	min_improvement = 1e-5,
		float	correspondence_threshold = 0.8,
		float	convergence_radius = 1e-4,
		bool	pre_align = False,
		int	correspondence_method = cvb.match_3d.CorrespondenceType.ConsensusICP )

Instantiates a MatchingParameters object.

Parameters

max_scene_points : int Maximum number of scene points to use. Points are selected randomly. If zero, all points are used.

max_model_points : int Maximum number of model points to use. Points are selected randomly. If zero, all points are used.

max_iterations : int Maximum number of iterations to execute (optional, default value = 20).

tolerance : float Iteration stops if mean distance between aligned point clouds is equal/below this value (optional, default value = 1e-9).

min_improvement : float The mean distance between aligned point clouds has to differ more than this value from the results of the previous iteration. Otherwise iteration stops (optional, default value = 1e-5).

correspondence_threshold : float Parameter for shape sensitivity. It is defined by a value between 0 and 1. A higher value means higher sensitivity. A value of 0 corresponds to the classical ICP without any shape sensitivity. A value between 0.8 and 0.9 is recommended as default value when aligning point clouds (optional, default value = 0.8).

convergence_radius : float Convergence radius to define when two points are considered as a converged match (optional, default value = 1e-4).

pre_align : bool True, if pre-alignment of point clouds should be done using the center of gravity (optional, default value = False).

correspondence_method : int Correspondence type used (see cvb.match_3d.CorrespondenceType) (optional, default value = cvb.match_3d.CorrespondenceType.ConsensusICP).

Property Documentation

◆ max_model_points

max_model_points = property

static

int: Gets or sets maximum number of model points to use.

If zero, all points are used.

◆ max_scene_points

max_scene_points = property