Core implementation for Kalman sensor. More...

#include <Kalman.h>

Inheritance diagram for alvar::KalmanSensorCore:

Public Member Functions
int	get_m ()
	Accessor for m.
int	get_n ()
	Accessor for n.
	KalmanSensorCore (const KalmanSensorCore &k)
	Copy constructor.
	KalmanSensorCore (int _n, int _m)
	Constructor.
virtual void	update_x (CvMat x_pred, CvMat x)
	Method for updating the state estimate x This is called from predict_update() of Kalman. In KalmanSensorCore and in KalmanSensor this update is made linearly but KalmanSensorEkf will override this method to use unlinear estimation.
	~KalmanSensorCore ()
	Destructor.
Public Attributes
CvMat *	H
	The matrix (m*n) mapping Kalman state vector into this sensor's measurements vector.
CvMat *	K
	The matrix (nm) containing Kalman gain (something between 0 and H^-1). In this core-implementation we assume this to be precalculated. In KalmanSensor* this is updated using update_K .
CvMat *	z
	Latest measurement vector (m*1)
Protected Attributes
CvMat *	H_trans
int	m
int	n
CvMat *	x_gain
CvMat *	z_pred
CvMat *	z_residual
Friends
class	KalmanVisualize

Detailed Description

Core implementation for Kalman sensor.

Definition at line 39 of file Kalman.h.

Constructor & Destructor Documentation

alvar::KalmanSensorCore::KalmanSensorCore ( const KalmanSensorCore & k )

Copy constructor.

Definition at line 35 of file Kalman.cpp.

alvar::KalmanSensorCore::KalmanSensorCore	(	int	_n,
		int	_m
	)

Constructor.

Parameters:

_n	The number of items in the Kalman state vector
_m	The number of measurements given by this sensor

Definition at line 47 of file Kalman.cpp.

alvar::KalmanSensorCore::~KalmanSensorCore ( )

Destructor.

Definition at line 59 of file Kalman.cpp.

Member Function Documentation

int alvar::KalmanSensorCore::get_m ( ) [inline]

Accessor for m.

Definition at line 70 of file Kalman.h.

int alvar::KalmanSensorCore::get_n ( ) [inline]

Accessor for n.

Definition at line 68 of file Kalman.h.

void alvar::KalmanSensorCore::update_x	(	CvMat *	x_pred,
		CvMat *	x
	)		`[virtual]`

Method for updating the state estimate x This is called from predict_update() of Kalman. In KalmanSensorCore and in KalmanSensor this update is made linearly but KalmanSensorEkf will override this method to use unlinear estimation.

Reimplemented in alvar::KalmanSensorEkf.

Definition at line 69 of file Kalman.cpp.

Friends And Related Function Documentation

friend class KalmanVisualize [friend]

Definition at line 40 of file Kalman.h.

Member Data Documentation

CvMat* alvar::KalmanSensorCore::H

The matrix (m*n) mapping Kalman state vector into this sensor's measurements vector.

Definition at line 52 of file Kalman.h.

CvMat* alvar::KalmanSensorCore::H_trans [protected]

Definition at line 44 of file Kalman.h.

CvMat* alvar::KalmanSensorCore::K

The matrix (n*m) containing Kalman gain (something between 0 and H^-1). In this core-implementation we assume this to be precalculated. In KalmanSensor this is updated using update_K .

Definition at line 56 of file Kalman.h.

int alvar::KalmanSensorCore::m [protected]

Definition at line 43 of file Kalman.h.

int alvar::KalmanSensorCore::n [protected]

Definition at line 42 of file Kalman.h.

CvMat* alvar::KalmanSensorCore::x_gain [protected]

Definition at line 47 of file Kalman.h.

CvMat* alvar::KalmanSensorCore::z

Latest measurement vector (m*1)

Definition at line 50 of file Kalman.h.

CvMat* alvar::KalmanSensorCore::z_pred [protected]

Definition at line 45 of file Kalman.h.