Package edu.gsu.cs.dmlab.factory.interfaces

Interface ISTEventTrackingFactory

public interface ISTEventTrackingFactory

The public interface for classes that shall be used to create the objects used in event tracking.

Author:

Dustin Kempton, Data Mining Lab, Georgia State University

Method Summary

Modifier and Type	Method	Description
ISTAppearanceModel	getAppearanceModel(ImageDBWaveParamPair[] params)	Gets a model for producing probability values to indicate how visually similar two tracks are at their point of possible joining.
ISTAssociationProblem	getAssociationProblem(List<ISTTrackingTrajectory> tracks, ISTTrackingEventIndexer evntsIdxr, int stage)	Produces a new Graph that depicts the possible paths tracks can take through multiple shorter track fragments.
ISTLocationProbCal	getEnterModel()	Gets a model for producing probability values based on the input objects location and size to represent how likely the input object is to be the beginning of a track.
ISTLocationProbCal	getExitModel()	Gets a model for producing probability values based on the input objects location and size to represent how likely the input object is to be the end of a track.
ISTMotionModel	getMotionModel()	Gets a model for producing probability values based on how similar the movement of two input tracks is.
ISTObsModel	getObservationModel(ISTTrackingEventIndexer evntsIdxr)	Gets a model for producing probability values to indicate how likely an object is a true detection of an event or is a false detection.
ISTFrameSkipModel	getSkipModel()	Gets a model for producing probability values based on the number of skipped frames after a particular track another track starts.
ISTProcessingStage	getStage1(ISTTrackingEventIndexer eventIndexer)
ISTProcessingStage	getStage2(ISTTrackingEventIndexer eventIndexer, ISTTrackingTrajectoryIndexer tracksIdxr, int maxFrameSkip)
ISTProcessingStage	getStage3(ISTTrackingEventIndexer eventIndexer, ISTTrackingTrajectoryIndexer tracksIdxr, int maxFrameSkip)
IStatProducer	getStatProducer(List<ISTTrackingTrajectory> tracks)
ISTTrackingTrajectory	getTrack(ISTTrackingEvent event)	Creates a new track containing the passed in IEvent object.

Method Detail

getTrack

ISTTrackingTrajectory getTrack(ISTTrackingEvent event)

Creates a new track containing the passed in IEvent object.

Parameters:

event - The IEvent for the track to contain.

Returns:

The track containing the passed in IEvent.

getAssociationProblem

ISTAssociationProblem getAssociationProblem(List<ISTTrackingTrajectory> tracks,
                                            ISTTrackingEventIndexer evntsIdxr,
                                            int stage)

Produces a new Graph that depicts the possible paths tracks can take through multiple shorter track fragments.

Parameters:

tracks - List of track fragments to build the association graph from.

evntsIdxr - The indexer of the events that are contained in all the tracks passed in. Which is used to calculate the expected change in the number of detections over a period of time.

stage - The stage which the association problem shall be used. This matters because we calculate the weights differently based on what stage we are in.

Returns:

The graph of possible association paths.

getSkipModel

ISTFrameSkipModel getSkipModel()

Gets a model for producing probability values based on the number of skipped frames after a particular track another track starts.

Returns:

The fame skip model.

getEnterModel

ISTLocationProbCal getEnterModel()

Gets a model for producing probability values based on the input objects location and size to represent how likely the input object is to be the beginning of a track.

Returns:

The enter model object.

getExitModel

ISTLocationProbCal getExitModel()

Gets a model for producing probability values based on the input objects location and size to represent how likely the input object is to be the end of a track.

Returns:

The exit model object.

getMotionModel

ISTMotionModel getMotionModel()

Gets a model for producing probability values based on how similar the movement of two input tracks is.

Returns:

The motion model object.

getObservationModel

ISTObsModel getObservationModel(ISTTrackingEventIndexer evntsIdxr)

Gets a model for producing probability values to indicate how likely an object is a true detection of an event or is a false detection.

Parameters:

evntsIdxr - An indexer object that contains event count per frame data for event reports in the temporal vicinity of an input object

Returns:

The observation model

getAppearanceModel

ISTAppearanceModel getAppearanceModel(ImageDBWaveParamPair[] params)

Gets a model for producing probability values to indicate how visually similar two tracks are at their point of possible joining.

Parameters:

params -

Returns:

getStatProducer

IStatProducer getStatProducer(List<ISTTrackingTrajectory> tracks)

getStage1

ISTProcessingStage getStage1(ISTTrackingEventIndexer eventIndexer)

getStage2

ISTProcessingStage getStage2(ISTTrackingEventIndexer eventIndexer,
                             ISTTrackingTrajectoryIndexer tracksIdxr,
                             int maxFrameSkip)

getStage3

ISTProcessingStage getStage3(ISTTrackingEventIndexer eventIndexer,
                             ISTTrackingTrajectoryIndexer tracksIdxr,
                             int maxFrameSkip)