Home || Architecture || Video Search || Visual Search || Scripts || Applications || Important Messages || OGL || Src

Impala::Core::Array::Trait::FuncKalman Class Reference

functor to do Kalman filtering on all pixels using the same Kalman filter parameters for all pixels. More...

#include <FunctorKalman.h>

Collaboration diagram for Impala::Core::Array::Trait::FuncKalman:

Collaboration graph
[legend]
List of all members.

Public Types

typedef Pattern::TagTransVar TransVarianceCategory
typedef Pattern::TagCallValue CallCategory
typedef Vec3Int32 DstArithT
typedef Vec3Int32 SrcArithT

Public Member Functions

 FuncKalman (int w, int h)
 the constructor needs the width and height of the template to allocate the outlier 'graphic'
void Reset (int w, int h)
 this function will make the functor reinitialize it, i.e.
Vec3Real64 doIt (const Vec3Real64 &p1, const Vec3Real64 &p2, int x, int y)
 this function blends the old template with a new measurement using eqation 12 from the paper.
void InitialiseSigma ()
 this function sets sigmas directly after the first time the average residual is computed
void UpdateKalman ()
 this function does reestimation of the sigmas using the residual
void ComputeResiduals ()

Public Attributes

double mSigmaI
double mSigmaW
double mSigmaG
double mAlpha
double mBeta
double mGamma
double mN_omax
double mK
double mPixelCount
int mOutliers
double mResidualThisFrame
double mResidualAvrg
bool mInitialised
bool mOcclusion

Private Attributes

Array2dScalarUInt8mOutCounter
std::deque< double > mPastResiduals
double mOutlierTreshold

Detailed Description

functor to do Kalman filtering on all pixels using the same Kalman filter parameters for all pixels.

Handles 'outliers' as described in the paper of Hieu

debug visualisations:

Definition at line 40 of file FunctorKalman.h.


The documentation for this class was generated from the following file:
Generated on Fri Mar 19 11:05:51 2010 for ImpalaSrc by  doxygen 1.5.1