boofcv.alg.sfm.d2

Class ImageMotionPointTrackerKey<I extends ImageBase,IT extends InvertibleTransform>

java.lang.Object

boofcv.alg.sfm.d2.ImageMotionPointTrackerKey<I,IT>

Type Parameters:

I - Input image type

IT - Motion model data type

public class ImageMotionPointTrackerKey<I extends ImageBase,IT extends InvertibleTransform>
extends java.lang.Object

Computes the transform from the first image in a sequence to the current frame. Keyframe based algorithm. Whenever a new keyframe is selected by the user all tracks are dropped and new ones spawned. No logic is contained for selecting key frames and relies on the user for selecting them.

Constructor Summary

Constructors

Constructor and Description
ImageMotionPointTrackerKey(PointTracker<I> tracker, ModelMatcher<IT,AssociatedPair> modelMatcher, ModelFitter<IT,AssociatedPair> modelRefiner, IT model, int outlierPrune) Specify algorithms to use internally.

Method Summary

Modifier and Type	Method and Description
void	changeKeyFrame() Change the current frame into the keyframe.
IT	getKeyToCurr()
ModelMatcher<IT,AssociatedPair>	getModelMatcher()
java.lang.Class<IT>	getModelType()
int	getTotalFramesProcessed()
PointTracker<I>	getTracker()
IT	getWorldToCurr()
IT	getWorldToKey()
boolean	isKeyFrame()
boolean	process(I frame) Processes the next frame in the sequence.
void	reset() Makes the current frame the first frame and discards its past history
void	resetTransforms()

Methods inherited from class java.lang.Object

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

Constructor Detail

ImageMotionPointTrackerKey

public ImageMotionPointTrackerKey(PointTracker<I> tracker,
                                  ModelMatcher<IT,AssociatedPair> modelMatcher,
                                  ModelFitter<IT,AssociatedPair> modelRefiner,
                                  IT model,
                                  int outlierPrune)

Specify algorithms to use internally. Each of these classes must work with compatible data structures.

Parameters:

tracker - feature tracker

modelMatcher - Fits model to track data

modelRefiner - (Optional) Refines the found model using the entire inlier set. Can be null.

model - Motion model data structure

outlierPrune - If a track is an outlier for this many frames in a row they are pruned

Method Detail

reset

public void reset()

Makes the current frame the first frame and discards its past history

process

public boolean process(I frame)

Processes the next frame in the sequence.

Parameters:

frame - Next frame in the video sequence

Returns:

true if motion was estimated and false if no motion was estimated

changeKeyFrame

public void changeKeyFrame()

Change the current frame into the keyframe. p1 location of existing tracks is set to their current location and new tracks are spawned. Reference frame transformations are also updated

resetTransforms

public void resetTransforms()

getWorldToCurr

public IT getWorldToCurr()

getWorldToKey

public IT getWorldToKey()

getKeyToCurr

public IT getKeyToCurr()

getTracker

public PointTracker<I> getTracker()

getModelMatcher

public ModelMatcher<IT,AssociatedPair> getModelMatcher()

getTotalFramesProcessed

public int getTotalFramesProcessed()

isKeyFrame

public boolean isKeyFrame()

getModelType

public java.lang.Class<IT> getModelType()