Package edu.gsu.cs.dmlab.tracking

Class BinomialFrameSkipModel

java.lang.Object

edu.gsu.cs.dmlab.tracking.BinomialFrameSkipModel

All Implemented Interfaces:

ISTFrameSkipModel

public class BinomialFrameSkipModel
extends Object
implements ISTFrameSkipModel

Model of frame skipping that return the probability of two track fragments being from the same object, given the number of skipped frames between them. This model uses the exit probability as a means to calculate this value by modeling the probability as a binomial distribution.

Author:

Dustin Kempton, Data Mining Lab, Georgia State University Michael Tinglof, Data Mining Lab, Georgia State University

Constructor Summary

Constructors

Constructor	Description
BinomialFrameSkipModel(ISTLocationProbCal exitProbCalculator)	Constructor with the exit probability calculator passed in.

Method Summary

Modifier and Type	Method	Description
void	finalize()
double	getSkipProb(ISTTrackingTrajectory leftTrack, ISTTrackingTrajectory rightTrack)	Returns the probability of two track fragments being from the same object, given the number of skipped frames between them.

Methods inherited from class java.lang.Object

clone, equals, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

BinomialFrameSkipModel

public BinomialFrameSkipModel(ISTLocationProbCal exitProbCalculator)

Constructor with the exit probability calculator passed in.

Parameters:

exitProbCalculator - The exit probability calculator used in this object.

Method Detail

finalize

public void finalize()
              throws Throwable

Overrides:

finalize in class Object

Throws:

Throwable

getSkipProb

public double getSkipProb(ISTTrackingTrajectory leftTrack,
                          ISTTrackingTrajectory rightTrack)

Description copied from interface: ISTFrameSkipModel

Returns the probability of two track fragments being from the same object, given the number of skipped frames between them.

Specified by:

getSkipProb in interface ISTFrameSkipModel

Parameters:

leftTrack - The track fragment that is earlier in time.

rightTrack - The track fragment that is later in time.

Returns:

The probability of the two fragments being the same object only based on the number of skipped frames between them.