cvflann::KNNUniqueResultSet< DistanceType > Class Template Reference

Class that holds the k NN neighbors Faster than KNNResultSet as it uses a binary heap and does not maintain two arrays. More...

#include <opencv2/flann/result_set.h>

Inheritance diagram for cvflann::KNNUniqueResultSet< DistanceType >:

Collaboration diagram for cvflann::KNNUniqueResultSet< DistanceType >:

Public Member Functions
	KNNUniqueResultSet (unsigned int capacity)
	Constructor. More...
void	addPoint (DistanceType dist, int index) CV_OVERRIDE
	Add a possible candidate to the best neighbors. More...
void	clear () CV_OVERRIDE
	Remove all elements in the set. More...
virtual void	copy (int *indices, DistanceType *dist, int n_neighbors=-1) const
	Copy the set to two C arrays. More...
bool	full () const CV_OVERRIDE
	Check the status of the set. More...
size_t	size () const
	The number of neighbors in the set. More...
virtual void	sortAndCopy (int *indices, DistanceType *dist, int n_neighbors=-1) const
	Copy the set to two C arrays but sort it according to the distance first. More...
DistanceType	worstDist () const CV_OVERRIDE
	The distance of the furthest neighbor If we don't have enough neighbors, it returns the max possible value. More...

Protected Types
typedef UniqueResultSet< DistanceType >::DistIndex	DistIndex

Protected Attributes
unsigned int	capacity_
	The number of neighbors to keep. More...
std::set< DistIndex >	dist_indices_
	The best candidates so far. More...
bool	is_full_
	Flag to say if the set is full. More...
DistanceType	worst_distance_
	The worst distance found so far. More...

Detailed Description

template<typename DistanceType>
class cvflann::KNNUniqueResultSet< DistanceType >

Class that holds the k NN neighbors Faster than KNNResultSet as it uses a binary heap and does not maintain two arrays.

Member Typedef Documentation

DistIndex

template<typename DistanceType>

typedef UniqueResultSet<DistanceType>::DistIndex cvflann::KNNUniqueResultSet< DistanceType >::DistIndex

protected

Constructor & Destructor Documentation

KNNUniqueResultSet()

template<typename DistanceType>

cvflann::KNNUniqueResultSet< DistanceType >::KNNUniqueResultSet ( unsigned int  capacity )

inline

Constructor.

Parameters

capacity

the number of neighbors to store at max

Member Function Documentation

addPoint()

template<typename DistanceType>

void cvflann::KNNUniqueResultSet< DistanceType >::addPoint ( DistanceType  dist, int  index  )

inlinevirtual

Add a possible candidate to the best neighbors.

Parameters

dist	distance for that neighbor
index	index of that neighbor

Implements cvflann::ResultSet< DistanceType >.

clear()

template<typename DistanceType>

void cvflann::KNNUniqueResultSet< DistanceType >::clear ( )

inlinevirtual

Remove all elements in the set.

Implements cvflann::UniqueResultSet< DistanceType >.

Reimplemented in cvflann::KNNRadiusUniqueResultSet< DistanceType >.

References cv::max().

Referenced by cvflann::NNIndex< cv::L1< ElementType > >::knnSearch(), and cvflann::LshIndex< Distance >::knnSearch().

Here is the call graph for this function:

Here is the caller graph for this function:

copy()

template<typename DistanceType >

virtual void cvflann::UniqueResultSet< DistanceType >::copy ( int *  indices, DistanceType *  dist, int  n_neighbors = -1  ) const

inlinevirtualinherited

Copy the set to two C arrays.

Parameters

indices	pointer to a C array of indices
dist	pointer to a C array of distances
n_neighbors	the number of neighbors to copy

Referenced by cvflann::NNIndex< cv::L1< ElementType > >::knnSearch(), cvflann::LshIndex< Distance >::knnSearch(), and cvflann::NNIndex< cv::L1< ElementType > >::radiusSearch().

Here is the caller graph for this function:

full()

template<typename DistanceType >

bool cvflann::UniqueResultSet< DistanceType >::full ( ) const

inlinevirtualinherited

Check the status of the set.

Returns

true if we have k NN

Implements cvflann::ResultSet< DistanceType >.

Reimplemented in cvflann::RadiusUniqueResultSet< DistanceType >.

size()

template<typename DistanceType >

size_t cvflann::UniqueResultSet< DistanceType >::size ( ) const

inlineinherited

The number of neighbors in the set.

Returns

Referenced by cvflann::NNIndex< cv::L1< ElementType > >::radiusSearch().

Here is the caller graph for this function:

sortAndCopy()

template<typename DistanceType >

virtual void cvflann::UniqueResultSet< DistanceType >::sortAndCopy ( int *  indices, DistanceType *  dist, int  n_neighbors = -1  ) const

inlinevirtualinherited

Copy the set to two C arrays but sort it according to the distance first.

Parameters

indices	pointer to a C array of indices
dist	pointer to a C array of distances
n_neighbors	the number of neighbors to copy

Referenced by cvflann::NNIndex< cv::L1< ElementType > >::knnSearch(), cvflann::LshIndex< Distance >::knnSearch(), and cvflann::NNIndex< cv::L1< ElementType > >::radiusSearch().

Here is the caller graph for this function:

worstDist()

template<typename DistanceType >

DistanceType cvflann::UniqueResultSet< DistanceType >::worstDist ( ) const

inlinevirtualinherited

The distance of the furthest neighbor If we don't have enough neighbors, it returns the max possible value.

Returns

Implements cvflann::ResultSet< DistanceType >.

Reimplemented in cvflann::RadiusUniqueResultSet< DistanceType >.

Member Data Documentation

capacity_

template<typename DistanceType>

unsigned int cvflann::KNNUniqueResultSet< DistanceType >::capacity_

protected

The number of neighbors to keep.

dist_indices_

template<typename DistanceType >

std::set<DistIndex> cvflann::UniqueResultSet< DistanceType >::dist_indices_

protectedinherited

The best candidates so far.

is_full_

template<typename DistanceType >

bool cvflann::UniqueResultSet< DistanceType >::is_full_

protectedinherited

Flag to say if the set is full.

worst_distance_

template<typename DistanceType >

DistanceType cvflann::UniqueResultSet< DistanceType >::worst_distance_

protectedinherited

The worst distance found so far.

---

The documentation for this class was generated from the following file:

• opencv2/flann/result_set.h