#include <vpRBSilhouetteCCDTracker.h>

Public Member Functions
	vpCCDParameters ()
	~vpCCDParameters ()=default

Public Attributes
double	gamma_1
double	gamma_2
double	gamma_3
double	gamma_4
double	alpha
double	beta
double	kappa
double	covarianceIterDecreaseFactor
int	h
int	delta_h
int	min_h
int	start_h
int	start_delta_h
int	iters_since_scale_change
int	min_iters_before_scale_change
int	phi_dim

Detailed Description

Examples: catchRBT.cpp.

Definition at line 65 of file vpRBSilhouetteCCDTracker.h.

Constructor & Destructor Documentation

◆ vpCCDParameters()

vpCCDParameters::vpCCDParameters ( )

inline

Definition at line 68 of file vpRBSilhouetteCCDTracker.h.

References alpha, beta, covarianceIterDecreaseFactor, delta_h, gamma_1, gamma_2, gamma_3, gamma_4, h, iters_since_scale_change, kappa, min_h, min_iters_before_scale_change, phi_dim, start_delta_h, and start_h.

◆ ~vpCCDParameters()

vpCCDParameters::~vpCCDParameters ( )

default

Member Data Documentation

◆ alpha

double vpCCDParameters::alpha

Definition at line 96 of file vpRBSilhouetteCCDTracker.h.

Referenced by vpCCDParameters().

◆ beta

double vpCCDParameters::beta

Definition at line 97 of file vpRBSilhouetteCCDTracker.h.

Referenced by vpCCDParameters().

◆ covarianceIterDecreaseFactor

double vpCCDParameters::covarianceIterDecreaseFactor

From the CCD paper: maximum decrease of the covariance within one iteration step. Between 0 and 1 If c2 is too high, the covariance declines slowly. Hence, a small number of iterations is necessary. If c2 is too small, the CCD algorithm may converge to a wrong solution. it is recommended to leave this value fixed.

Definition at line 114 of file vpRBSilhouetteCCDTracker.h.

Referenced by vpCCDParameters().

◆ delta_h

int vpCCDParameters::delta_h

Sample step when computing statistics and errors. Increase this value to decrease computation time, at the risk of obtaining inaccurate statistics.

Examples: catchRBT.cpp.

Definition at line 127 of file vpRBSilhouetteCCDTracker.h.

Referenced by vpCCDParameters().

◆ gamma_1

double vpCCDParameters::gamma_1

Curve uncertainty computation hyperparameter. Recommended to leave fixed.

Examples: catchRBT.cpp.

Definition at line 80 of file vpRBSilhouetteCCDTracker.h.

Referenced by vpCCDParameters().

◆ gamma_2

double vpCCDParameters::gamma_2

Curve uncertainty computation hyperparameter. Recommended to leave fixed.

Examples: catchRBT.cpp.

Definition at line 85 of file vpRBSilhouetteCCDTracker.h.

Referenced by vpCCDParameters().

◆ gamma_3

double vpCCDParameters::gamma_3

Curve uncertainty computation hyperparameter. Recommended to leave fixed.

Examples: catchRBT.cpp.

Definition at line 90 of file vpRBSilhouetteCCDTracker.h.

Referenced by vpCCDParameters().

◆ gamma_4

double vpCCDParameters::gamma_4

Curve uncertainty computation hyperparameter. Recommended to leave fixed.

Examples: catchRBT.cpp.

Definition at line 95 of file vpRBSilhouetteCCDTracker.h.

Referenced by vpCCDParameters().

◆ h

int vpCCDParameters::h

Size of the vicinity that is used to compute statistics and error. Length of the line along the normal (and the opposite direction). To subsample the line, set delta_h > 1. Number of pixels used is computed as 2 * floor(h/delta_h). If you expect large motions, set a large value. If you want to reduce computation time, decrease this value or increase delta_h Recommended value: 4 or above (this is dependent on image resolution).