name: KalmanFilter : public AlgorithmBase


g++ [flags ...] file ... -l /isip/tools/lib/$ISIP_BINARY/lib_algo.a

#include <KalmanFilter.h>

KalmanFilter(ALGORITHM algorithm = DEF_ALGORITHM, IMPLEMENTATION implementation = DEF_IMPLEMENTATION, long order);
boolean eq(const KalmanFilter& arg);
boolean setAlgorithm(ALGORITHM algorithm);
boolean setOrder(long order);
boolean setNoise(double covar_nse_pro, double covar_nse_obs, double covar_err_est);
boolean setParameter(double alpha, double beta, double kappa);
quick start:

KalmanFilter kf;
VectorDouble input(L"1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15");
VectorDouble output;
kf.setNoise(0.03, 0.02, 0.0001);
kf.compute(output, input);

The KalmanFilter class is used to filter the speech signal in presence of noise. The speech signal is modeled by state-space model. The two reference documents were used to implement the algorithm for filtering of speech. These are the couple of references which will help in understanding of Kalman filter.
The unscented Kalman filter implementation is based on This class currently supports two algorithm choice: CONVENTIONAL and UNSCENTED. It is described in detail in the above two references. The state-space equations used to model the speech signals are shown below:

The filter order can be specified by the user using the setOrder() method. Other parameters that need to be specified are process noise covariance, observation noise covariance, and error estimate covariance by using the setParameter() method.


public constants:

error codes:

protected data:

required public methods:

class-specific public methods:

private methods: