jafar::icp::OICP Class Reference

Class OICP stands for Original ICP It is an implementation of Besl original ICP algorithm described in A method for icp of 3-D shapes, Besl and McKay in IEEE PAMI, 14(2):239 - 256, February 1992 It is aim is to align two 3D [later 2D] points cloud computing iteratively the best T3D that can perform it. More...

---

Detailed Description

Class OICP stands for Original ICP It is an implementation of Besl original ICP algorithm described in A method for icp of 3-D shapes, Besl and McKay in IEEE PAMI, 14(2):239 - 256, February 1992 It is aim is to align two 3D [later 2D] points cloud computing iteratively the best T3D that can perform it.

Definition at line 21 of file oicp.hpp.

#include <oicp.hpp>

List of all members.

Public Member Functions
	OICP (size_t max_nb_points=1024, double epsilon=1e-4, double max_dist=5., int max_iter=100)
	OICP (const jblas::mat &model, jblas::mat &scene, double epsilon=1e-4, double max_dist=5., int max_iter=100)
void	add_model_entry (const jblas::vec &entry)
void	add_scene_entry (const jblas::vec &entry)
void	compute_RT (const jblas::mat &model, const jblas::mat &scene, jblas::mat &R, jblas::vec &T)
	computes optimal rotation and translation for current
void	run ()
Private Member Functions
void	barycenter (const jblas::mat &points, jblas::vec &center)
Private Attributes
double	m_epsilon
	error
double	m_max_dist
	max distance
int	m_max_iter
	maximum number of iterations to be performed
unsigned int	m_model_nb_points
	how many points in model
unsigned int	m_scene_nb_points
	how many points in model
ublas::matrix< double >	m_model_points
	model points cloud arranged in a matrix
ublas::matrix< double >	m_scene_points
	scene points cloud arranged in a matrix
jann::KD_tree_index< flann::L2 < double > > *	m_model_index
	indexed points
size_t	m_K
	number of neighbours
jblas::mat	R
jblas::vec	T

---

Member Function Documentation

void jafar::icp::OICP::barycenter	(	const jblas::mat &	points,
		jblas::vec &	center
	)		[inline, private]

Returns:

barycenter of N points of dimension dim stored in a Nxdim matrix

Remarks:

: the input/output is set to zero vector to begin

---

The documentation for this class was generated from the following file:

• modules/icp/include/icp/oicp.hpp