jafar::icp::GICP Class Reference

Class GICP stands for Generalized ICP It is implementation of algorithm described in Generalized-ICP, Aleksandr V. More...

---

Detailed Description

Class GICP stands for Generalized ICP It is implementation of algorithm described in Generalized-ICP, Aleksandr V.

Segal, Dirk Haehnel, and Sebastian Thrun. In Robotics: Science and Systems 2009.

It is aim is to align two 3D [2D] points cloud computing iteratively the best t3D that can perform it.

Definition at line 21 of file gicp.hpp.

#include <gicp.hpp>

List of all members.

Public Member Functions
	GICP (size_t max_nb_points=1024, double epsilon=1e-3, double max_dist=5., int K=20, int max_iter=100, int inner_max_iter=8, double ftol=0.01, double gtol=0.01)
	GICP (const jblas::mat &model, jblas::mat &scene, double epsilon=1e-3, double max_dist=5., int K=20, int max_iter=100, int inner_max_iter=8, double ftol=0.01, double gtol=0.01)
void	addModelEntry (const jblas::vec3 &entry)
void	addSceneEntry (const jblas::vec3 &entry)
void	transform (const geom::T3D *t, jblas::vec3 &pt)
void	computeCovarianceMatrices (const ANNpointArray &points, size_t nb_points, ANNkd_tree *points_tree, covariances &points_covariances)
	coputes normal per point within a neighborhood of K
int	align (const ANNpointArray &model_points, unsigned int nb_model_points, const covariances &model_covariances, const ANNpointArray &scene_points, ANNkd_tree *scene_tree, const covariances &scene_covariances, const geom::T3D *guess, geom::T3D *t)
	align the scene with the model
int	align_gsl (const ANNpointArray &model, unsigned int nb_model_points, const covariances &model_covariances, const ANNpointArray &scene, ANNkd_tree *scene_tree, const covariances &scene_covariances, const geom::T3D *guess, geom::T3D *t)
void	run (geom::T3DEuler *guess=NULL)
	run gicpon the model and the scene points
double	f (const gsl_vector *x, void *params)
void	df (const gsl_vector *x, void *params, gsl_vector *g)
void	fdf (const gsl_vector *x, void *params, double *f, gsl_vector *g)
geom::T3D *	get_final_t3d ()
Private Types
typedef std::vector < jblas::mat_column_major >	covariances
	a vector to store covariance matrixes
Private Member Functions
double	evaluate (const double *x, double *g, const int n, const double step)
int	progress (const double *x, const double *g, const double fx, const double xnorm, const double gnorm, const double step, int n, int k, int ls)
double	trace_at_b (const jblas::mat &a, const jblas::mat &b)
Static Private Member Functions
static double	_evaluate (void *instance, const double *x, double *g, const int n, const double step)
static int	_progress (void *instance, const double *x, const double *g, const double fx, const double xnorm, const double gnorm, const double step, int n, int k, int ls)
Private Attributes
ANNpointArray	m_model_points
	model points cloud arranged in a natrix
ANNpointArray	m_scene_points
	scene points cloud arranged in a natrix
ANNkd_tree *	m_model_tree
	the indexed points
ANNkd_tree *	m_scene_tree
unsigned int	m_model_nb_points
	number of points of the model
unsigned int	m_scene_nb_points
	number of points of the scene
jblas::mat	R
jblas::mat	t
int	m_K
double	m_epsilon
double	m_max_dist
covariances	m_scene_covs
	covariances of the scene points
covariances	m_model_covs
	covariances of the model points
int	m_max_iter
	maximum number of iterations to be performed
int	m_lbfgs_max_iter
	maxmum number of iteration for the lbfgs algorithm
lbfgs_parameter_t	m_lbfgs_param
	lbfgs algorithm parameters
double	m_lbfgs_ftol
	lbfgs algorithm tolerance on f
double	m_lbfgs_gtol
	lbfgs algorithm tolerance on the gradient g
double *	m_x
	current x for lbfgs algorithm
double	m_fx
	the value of f(x) for lbfgs algorithm
int	nb_matches
geom::T3D *	initial_t3d
geom::T3D *	final_t3d
std::vector < jblas::mat_column_major >	mahalanobis
ANNidx *	indexes
std::vector< ANNdist >	distances

---

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

• modules/icp/include/icp/gicp.hpp