Main Content

setTrackFilterProperties

Set filter properties of track from multi-object tracker

Description

example

setTrackFilterProperties(tracker,trackID,property,value) sets the specified tracking filter property to the indicated value for a specific track within the multi-object tracker. trackID is the ID of that specific track.

setTrackFilterProperties(tracker,trackID,property1,value1,...,propertyN,valueN) sets multiple property values. You can specify the property-value pairs in any order.

Examples

collapse all

Create a multiObjectTracker System object™ using a constant-acceleration, linear Kalman filter for all tracks.

tracker = multiObjectTracker('FilterInitializationFcn',@initcakf, ...
    'ConfirmationParameters',[4 5],'DeletionThreshold',[9 9]);

Create two detections and generate tracks for these detections.

detection1 = objectDetection(1.0,[10; 10]);
detection2 = objectDetection(1.0,[1000; 1000]);
[~,tracks] = tracker([detection1 detection2],1.1)
tracks=2×1 objectTrack array with properties:
    TrackID
    BranchID
    SourceIndex
    UpdateTime
    Age
    State
    StateCovariance
    StateParameters
    ObjectClassID
    ObjectClassProbabilities
    TrackLogic
    TrackLogicState
    IsConfirmed
    IsCoasted
    IsSelfReported
    ObjectAttributes

Get filter property values for the first track. Display the process noise values.

values = getTrackFilterProperties(tracker,1,'MeasurementNoise','ProcessNoise','MotionModel');
values{2}
ans = 2×2

     1     0
     0     1

Set new values for this property by doubling the process noise for the first track. Display the updated process noise values.

setTrackFilterProperties(tracker,1,'ProcessNoise',2*values{2});
values = getTrackFilterProperties(tracker,1,'ProcessNoise');
values{1}
ans = 2×2

     2     0
     0     2

Input Arguments

collapse all

Multi-object tracker, specified as a multiObjectTracker System object.

Track ID, specified as a positive integer. trackID must be a valid track in tracker.

Tracking filter property to set values for, specified as a character vector or string scalar. property must be a valid property of the tracking filter used by tracker. Valid tracking filters are trackingKF, trackingEKF, and trackingUKF.

You can specify additional property-value pairs in any order.

Example: 'MeasurementNoise',eye(2,2),'MotionModel','2D Constant Acceleration'

Data Types: char | string

Value to set the corresponding tracking filter property to, specified as a MATLAB expression. value must be a valid value of the corresponding property.

You can specify additional property-value pairs in any order.

Example: 'MeasurementNoise',eye(2,2),'MotionModel','2D Constant Acceleration'

Version History

Introduced in R2017a