CGAL::Shape_detection::Efficient_RANSAC< Traits >::Parameters Struct Reference

#include <CGAL/Shape_detection/Efficient_RANSAC/Efficient_RANSAC.h>

Definition

template<class Traits>
struct CGAL::Shape_detection::Efficient_RANSAC< Traits >::Parameters

Parameters for the shape detection algorithm.

They are explained in detail in Section Parameters of the User Manual.

Public Attributes
FT	probability
	Probability to control search endurance.
std::size_t	min_points
	Minimum number of points in a shape.
FT	epsilon
	Maximum acceptable Euclidean distance between a point and a shape.
FT	normal_threshold
	Maximum threshold on the dot product between the estimated shape's normal and the point's normal, that is the cosine of the angle (cos(25°) = 0.9).
FT	cluster_epsilon
	Maximum acceptable Euclidean distance between points, which are assumed to be neighbors.

Member Data Documentation

cluster_epsilon

template<class Traits >

FT CGAL::Shape_detection::Efficient_RANSAC< Traits >::Parameters::cluster_epsilon

Maximum acceptable Euclidean distance between points, which are assumed to be neighbors.

Default value is 1% of the bounding box diagonal.

It must belong to the interval [0, +inf).

epsilon

template<class Traits >

FT CGAL::Shape_detection::Efficient_RANSAC< Traits >::Parameters::epsilon

Maximum acceptable Euclidean distance between a point and a shape.

Default value is 1% of the bounding box diagonal.

It must belong to the interval [0, +inf).

min_points

template<class Traits >

std::size_t CGAL::Shape_detection::Efficient_RANSAC< Traits >::Parameters::min_points

Minimum number of points in a shape.

Default value is 1% of total number of input points.

It must belong to the interval [0, +inf).

normal_threshold

template<class Traits >

FT CGAL::Shape_detection::Efficient_RANSAC< Traits >::Parameters::normal_threshold

Maximum threshold on the dot product between the estimated shape's normal and the point's normal, that is the cosine of the angle (cos(25°) = 0.9).

Default value is 0.9 (around 25 degrees).

It must belong to the interval [0, 1].

probability

template<class Traits >

FT CGAL::Shape_detection::Efficient_RANSAC< Traits >::Parameters::probability

Probability to control search endurance.

Default value is 0.05.

A lower probability provides a higher reliability and determinism at the cost of longer running time due to a higher search endurance.

It must belong to the interval [0, 1].