OpenCV  3.1.0
Open Source Computer Vision
Public Member Functions | Public Attributes | List of all members
cv::tracking::UnscentedKalmanFilterParams Class Reference

Unscented Kalman filter parameters. The class for initialization parameters of Unscented Kalman filter. More...

#include "kalman_filters.hpp"

Inheritance diagram for cv::tracking::UnscentedKalmanFilterParams:
cv::tracking::AugmentedUnscentedKalmanFilterParams

Public Member Functions

 UnscentedKalmanFilterParams ()
 
 UnscentedKalmanFilterParams (int dp, int mp, int cp, double processNoiseCovDiag, double measurementNoiseCovDiag, Ptr< UkfSystemModel > dynamicalSystem, int type=CV_64F)
 
void init (int dp, int mp, int cp, double processNoiseCovDiag, double measurementNoiseCovDiag, Ptr< UkfSystemModel > dynamicalSystem, int type=CV_64F)
 

Public Attributes

double alpha
 Default is 1e-3. More...
 
double beta
 Default is 2.0. More...
 
int CP
 Dimensionality of the control vector. More...
 
int dataType
 Type of elements of vectors and matrices, default is CV_64F. More...
 
int DP
 Dimensionality of the state vector. More...
 
Mat errorCovInit
 State estimate cross-covariance matrix, DP x DP, default is identity. More...
 
double k
 Default is 0. More...
 
Mat measurementNoiseCov
 Measurement noise cross-covariance matrix, MP x MP. More...
 
Ptr< UkfSystemModelmodel
 Object of the class containing functions for computing the next state and the measurement. More...
 
int MP
 Dimensionality of the measurement vector. More...
 
Mat processNoiseCov
 Process noise cross-covariance matrix, DP x DP. More...
 
Mat stateInit
 Initial state, DP x 1, default is zero. More...
 

Detailed Description

Unscented Kalman filter parameters. The class for initialization parameters of Unscented Kalman filter.

Constructor & Destructor Documentation

cv::tracking::UnscentedKalmanFilterParams::UnscentedKalmanFilterParams ( )
inline

The constructors.

cv::tracking::UnscentedKalmanFilterParams::UnscentedKalmanFilterParams ( int  dp,
int  mp,
int  cp,
double  processNoiseCovDiag,
double  measurementNoiseCovDiag,
Ptr< UkfSystemModel dynamicalSystem,
int  type = CV_64F 
)
Parameters
dp- dimensionality of the state vector,
mp- dimensionality of the measurement vector,
cp- dimensionality of the control vector,
processNoiseCovDiag- value of elements on main diagonal process noise cross-covariance matrix,
measurementNoiseCovDiag- value of elements on main diagonal measurement noise cross-covariance matrix,
dynamicalSystem- ptr to object of the class containing functions for computing the next state and the measurement,
type- type of the created matrices that should be CV_32F or CV_64F.

Member Function Documentation

void cv::tracking::UnscentedKalmanFilterParams::init ( int  dp,
int  mp,
int  cp,
double  processNoiseCovDiag,
double  measurementNoiseCovDiag,
Ptr< UkfSystemModel dynamicalSystem,
int  type = CV_64F 
)

The function for initialization of Unscented Kalman filter

Parameters
dp- dimensionality of the state vector,
mp- dimensionality of the measurement vector,
cp- dimensionality of the control vector,
processNoiseCovDiag- value of elements on main diagonal process noise cross-covariance matrix,
measurementNoiseCovDiag- value of elements on main diagonal measurement noise cross-covariance matrix,
dynamicalSystem- ptr to object of the class containing functions for computing the next state and the measurement,
type- type of the created matrices that should be CV_32F or CV_64F.

Member Data Documentation

double cv::tracking::UnscentedKalmanFilterParams::alpha

Default is 1e-3.

double cv::tracking::UnscentedKalmanFilterParams::beta

Default is 2.0.

int cv::tracking::UnscentedKalmanFilterParams::CP

Dimensionality of the control vector.

int cv::tracking::UnscentedKalmanFilterParams::dataType

Type of elements of vectors and matrices, default is CV_64F.

int cv::tracking::UnscentedKalmanFilterParams::DP

Dimensionality of the state vector.

Mat cv::tracking::UnscentedKalmanFilterParams::errorCovInit

State estimate cross-covariance matrix, DP x DP, default is identity.

double cv::tracking::UnscentedKalmanFilterParams::k

Default is 0.

Mat cv::tracking::UnscentedKalmanFilterParams::measurementNoiseCov

Measurement noise cross-covariance matrix, MP x MP.

Ptr<UkfSystemModel> cv::tracking::UnscentedKalmanFilterParams::model

Object of the class containing functions for computing the next state and the measurement.

int cv::tracking::UnscentedKalmanFilterParams::MP

Dimensionality of the measurement vector.

Mat cv::tracking::UnscentedKalmanFilterParams::processNoiseCov

Process noise cross-covariance matrix, DP x DP.

Mat cv::tracking::UnscentedKalmanFilterParams::stateInit

Initial state, DP x 1, default is zero.


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