Package boofcv.alg.feature.associate

Class AssociateNearestNeighbor_ST<D>

java.lang.Object

boofcv.alg.feature.associate.AssociateNearestNeighbor<D>

boofcv.alg.feature.associate.AssociateNearestNeighbor_ST<D>

All Implemented Interfaces:

Associate<D>, AssociateDescription<D>

public class AssociateNearestNeighbor_ST<D>
extends AssociateNearestNeighbor<D>

Matches features using a NearestNeighbor search from DDogleg. The source features are processed as a lump using NearestNeighbor.setPoints(java.util.List, boolean) while destination features are matched one at time using NearestNeighbor.Search.findNearest(Object, double, org.ddogleg.nn.NnData). Typically the processing of source features is more expensive and should be minimized while looking up destination features is fast. Multiple matches for source features are possible while there will only be a unique match for each destination feature.

An optional ratio test inspired from [1] can be used. The ratio between the best and second best score is found. if the difference is significant enough then the match is accepted. This this is a ratio test, knowing if the score is squared is important. Please set the flag correctly. Almost always the score is Euclidean distance squared.

[1] Lowe, David G. "Distinctive image features from scale-invariant keypoints." International journal of computer vision 60.2 (2004): 91-110.

Field Summary

Fields inherited from class boofcv.alg.feature.associate.AssociateNearestNeighbor

matchesAll

Constructor Summary

Constructors

Constructor	Description
AssociateNearestNeighbor_ST(NearestNeighbor<D> alg, Class<D> descType)

Method Summary

Modifier and Type	Method	Description
void	associate()	Finds the best match for each item in the source list with an item in the destination list.
Class<D>	getDescriptionType()	Returns the type of object it can associate
void	setDestination(FastAccess<D> listDst)	Sets the list of destination features NOTE: A reference to the input list might be saved internally until the next call to this function.
void	setSource(FastAccess<D> listSrc)	Sets the list of source features.

Methods inherited from class boofcv.alg.feature.associate.AssociateNearestNeighbor

getMatches, getScoreRatioThreshold, getScoreType, getUnassociatedDestination, getUnassociatedSource, isRatioUsesSqrt, setMaxScoreThreshold, setRatioUsesSqrt, setScoreRatioThreshold, uniqueDestination, uniqueSource

Methods inherited from class java.lang.Object

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

Constructor Details

AssociateNearestNeighbor_ST

public AssociateNearestNeighbor_ST(NearestNeighbor<D> alg,
 Class<D> descType)

Method Details

setSource

public void setSource(FastAccess<D> listSrc)

Description copied from interface: AssociateDescription

Sets the list of source features. NOTE: A reference to the input list might be saved internally until the next call to this function.

Specified by:

setSource in interface AssociateDescription<D>

Overrides:

setSource in class AssociateNearestNeighbor<D>

Parameters:

listSrc - List of features

setDestination

public void setDestination(FastAccess<D> listDst)

Description copied from interface: AssociateDescription

Sets the list of destination features NOTE: A reference to the input list might be saved internally until the next call to this function.

Specified by:

setDestination in interface AssociateDescription<D>

Overrides:

setDestination in class AssociateNearestNeighbor<D>

Parameters:

listDst - List of features

associate

public void associate()

Description copied from interface: Associate

Finds the best match for each item in the source list with an item in the destination list.

getDescriptionType

public Class<D> getDescriptionType()

Description copied from interface: Associate

Returns the type of object it can associate