# initcakf

Create constant-acceleration linear Kalman filter from detection report

## Description

creates and initializes a constant-acceleration linear Kalman
`filter`

= initcakf(`detection`

)`filter`

from information contained in a
`detection`

report. For more details, see Algorithms and `trackingKF`

.

## Examples

## Input Arguments

## Output Arguments

## Algorithms

The

`detection`

input must contain a 1-D, 2-D, or 3-D position measurement in Cartesian coordinates.For a 1-D position measurement, the function initializes a

`trackingKF`

with a 1-D constant acceleration model, in which the state is [`x; vx; ax`

]. The function sets the`MotionModel`

property of the filter as`"1D Constant Acceleration"`

.For a 2-D position measurement, the function initializes a

`trackingKF`

with a 2-D constant acceleration model, in which the state is [`x; vx; ax; y; vy; ay`

]. The function sets the`MotionModel`

property of the filter as`"2D Constant Acceleration"`

.For a 3-D position measurement, the function initializes a

`trackingKF`

with a 3-D constant acceleration model, in which the state is [*x*;*v*_{x};*a*_{x};*y*;*v*_{y};*a*_{y};*z*;*v*_{z};*a*_{z}]. The function sets the`MotionModel`

property of the filter as`"3D Constant Acceleration"`

.

*x*,*y*, and*z*are position coordinates. The function sets these position states same as those in the measurement of the`detection`

.*v*_{x},*v*_{y}, and*v*_{z}are the corresponding velocity states and the function sets them as 0.*a*_{x},*a*_{y}, and*a*_{z}are the corresponding acceleration states and the function sets these them as 0.The position components of the state error covariance matrix in the initialized

`trackingKF`

object are the same as those in the measurement covariance matrix contained in the detection. The velocity and acceleration components of the state error covariance matrix are set to 100 m^{2}/s^{2}and 100 m^{4}/s^{4}, respectively. The cross components of the state error covariance matrix are set to 0.The function computes the process noise matrix assuming a unit acceleration increment per step following the Weiner-sequence acceleration model.

The measurement noise matrix in the initialized filter is the same as that in the

`detection`

.You can use this function as the

`FilterInitializationFcn`

property of a`multiObjectTracker`

object.

## Extended Capabilities

## Version History

**Introduced in R2017a**