#include <Predictor.h>

Inheritance diagram for Tracking::Predictor:

Public Member Functions
virtual void	movePredicted (const TomGine::tgPose &poseIn, TomGine::tgPose &poseOut, float &c)
	Interface for Prediction/Motion systems.
	Predictor ()
virtual void	resample (Distribution &d, int num_particles, Particle variance, bool useMotion=true)
	Resample particles according to current likelihood distribution (move particles)
virtual void	sample (Distribution &d, int num_particles, Particle mean, Particle variance)
	Sample new distribution.
virtual void	sampleFromGaussian (Distribution &d, int num_particles, Particle mean, Particle variance, float sigma=1.0)
	Adds samples/particles to a distribution d by adding gaussian noise.
void	setCamViewVector (TomGine::tgVector3 v)
	Set vector pointing from camera to object mean, to enable zooming.
void	setNoConvergence (float v)
	Set the number of particles (in percent) which are voting for no convergence (for capturing fast movement)
virtual void	updateTime (double dTime)
	Updates time of prediction system for higher order motion model.
Protected Member Functions
Particle	genNoise (float sigma, Particle pConstraint, unsigned int type=GAUSS)
float	noise (float sigma, unsigned int type=GAUSS)
Protected Attributes
float	c_pred
TomGine::tgVector3	m_cam_view
double	m_dTime
float	m_noConvergence
float	m_powTime
float	m_powTimeSteps

Detailed Description

Definition at line 24 of file Predictor.h.

Constructor & Destructor Documentation

Predictor::Predictor ( )

Definition at line 8 of file Predictor.cpp.

Particle Predictor::genNoise	(	float	sigma,
		Particle	pConstraint,
		unsigned int	type = `GAUSS`
	)		`[protected]`

Definition at line 37 of file Predictor.cpp.

void Predictor::movePredicted	(	const TomGine::tgPose &	poseIn,
		TomGine::tgPose &	poseOut,
		float &	c
	)		`[virtual]`

Interface for Prediction/Motion systems.

Parameters:

poseIn	actual pose
poseOut	predicted/moved pose
c	confidence about prediction/movement ( 1.0 = high confidence; 0.0 = no confidence )

Definition at line 88 of file Predictor.cpp.

Definition at line 15 of file Predictor.cpp.

Resample particles according to current likelihood distribution (move particles)

Parameters:

d	pointer to distribution
num_particles	number of particles of resampled likelihood distribution
variance	variance of sampling in each degree of freedom (represented as particle)

Definition at line 95 of file Predictor.cpp.

Sample new distribution.

Parameters:

d	pointer to distribution
num_particles	number of particles to represent likelihood distribution
mean	mean particle of distribution
variance	variance of sampling in each degree of freedom (represented as particle)

Definition at line 184 of file Predictor.cpp.

void Predictor::sampleFromGaussian	(	Distribution &	d,
		int	num_particles,
		Particle	mean,
		Particle	variance,
		float	sigma = `1.0`
	)		`[virtual]`

Adds samples/particles to a distribution d by adding gaussian noise.

Parameters:

d	particle distribution
num_particles	number of particles of resampled likelihood distribution
mean	mean particle of distribution
variance	variance of sampling in each degree of freedom (represented as particle)
sigma	standard variation of sampling (dependent on confidence value -> sigma(c))