mrpt::vision::TSIFTDescriptorsKDTreeIndex< distance_t, metric_t > Struct Template Reference

Detailed Description

template<typename distance_t, class metric_t = nanoflann::L2_Simple_Adaptor< uint8_t , detail::TSIFTDesc2KDTree_Adaptor<distance_t>, distance_t>>
struct mrpt::vision::TSIFTDescriptorsKDTreeIndex< distance_t, metric_t >

A kd-tree builder for sets of features with SIFT descriptors.

Example of usage:

TSIFTDescriptorsKDTreeIndex<double>  feats_kdtree(feats);
feats_kdtree.get_kdtree().knnSearch( ... );

See also

CFeatureList, mrpt::vision::find_descriptor_pairings

Definition at line 49 of file descriptor_kdtrees.h.

#include <mrpt/vision/descriptor_kdtrees.h>

Public Types
using	kdtree_t = typename nanoflann::KDTreeSingleIndexAdaptor< metric_t, detail::TSIFTDesc2KDTree_Adaptor< distance_t > >

Public Member Functions
	TSIFTDescriptorsKDTreeIndex (const CFeatureList &feats)
	Constructor from a list of SIFT features. More...
void	regenerate_kdtreee ()
	Re-creates the kd-tree, which must be done whenever the data source (the CFeatureList) changes. More...
kdtree_t &	get_kdtree ()
	Access to the kd-tree object. More...
const kdtree_t &	get_kdtree () const
	~TSIFTDescriptorsKDTreeIndex ()

Private Attributes
detail::TSIFTDesc2KDTree_Adaptor< distance_t >	m_adaptor
kdtree_t *	m_kdtree
const CFeatureList &	m_feats

Member Typedef Documentation

kdtree_t

template<typename distance_t , class metric_t = nanoflann::L2_Simple_Adaptor< uint8_t , detail::TSIFTDesc2KDTree_Adaptor<distance_t>, distance_t>>

using mrpt::vision::TSIFTDescriptorsKDTreeIndex< distance_t, metric_t >::kdtree_t = typename nanoflann::KDTreeSingleIndexAdaptor< metric_t, detail::TSIFTDesc2KDTree_Adaptor<distance_t> >

Definition at line 53 of file descriptor_kdtrees.h.

Constructor & Destructor Documentation

TSIFTDescriptorsKDTreeIndex()

template<typename distance_t , class metric_t = nanoflann::L2_Simple_Adaptor< uint8_t , detail::TSIFTDesc2KDTree_Adaptor<distance_t>, distance_t>>

mrpt::vision::TSIFTDescriptorsKDTreeIndex< distance_t, metric_t >::TSIFTDescriptorsKDTreeIndex ( const CFeatureList &  feats )

inline

Constructor from a list of SIFT features.

Automatically build the KD-tree index. The list of features must NOT be empty or an exception will be raised.

Definition at line 59 of file descriptor_kdtrees.h.

References ASSERT_, mrpt::vision::CFeatureList::empty(), and mrpt::vision::TSIFTDescriptorsKDTreeIndex< distance_t, metric_t >::regenerate_kdtreee().

Here is the call graph for this function:

~TSIFTDescriptorsKDTreeIndex()

template<typename distance_t , class metric_t = nanoflann::L2_Simple_Adaptor< uint8_t , detail::TSIFTDesc2KDTree_Adaptor<distance_t>, distance_t>>

mrpt::vision::TSIFTDescriptorsKDTreeIndex< distance_t, metric_t >::~TSIFTDescriptorsKDTreeIndex ( )

inline

Definition at line 81 of file descriptor_kdtrees.h.

References mrpt::vision::TSIFTDescriptorsKDTreeIndex< distance_t, metric_t >::m_kdtree.

Member Function Documentation

get_kdtree() [1/2]

template<typename distance_t , class metric_t = nanoflann::L2_Simple_Adaptor< uint8_t , detail::TSIFTDesc2KDTree_Adaptor<distance_t>, distance_t>>

kdtree_t& mrpt::vision::TSIFTDescriptorsKDTreeIndex< distance_t, metric_t >::get_kdtree ( )

inline

Access to the kd-tree object.

Definition at line 79 of file descriptor_kdtrees.h.

References mrpt::vision::TSIFTDescriptorsKDTreeIndex< distance_t, metric_t >::m_kdtree.

get_kdtree() [2/2]

template<typename distance_t , class metric_t = nanoflann::L2_Simple_Adaptor< uint8_t , detail::TSIFTDesc2KDTree_Adaptor<distance_t>, distance_t>>

const kdtree_t& mrpt::vision::TSIFTDescriptorsKDTreeIndex< distance_t, metric_t >::get_kdtree ( ) const

inline

Definition at line 80 of file descriptor_kdtrees.h.

References mrpt::vision::TSIFTDescriptorsKDTreeIndex< distance_t, metric_t >::m_kdtree.

regenerate_kdtreee()

template<typename distance_t , class metric_t = nanoflann::L2_Simple_Adaptor< uint8_t , detail::TSIFTDesc2KDTree_Adaptor<distance_t>, distance_t>>

void mrpt::vision::TSIFTDescriptorsKDTreeIndex< distance_t, metric_t >::regenerate_kdtreee ( )

inline

Re-creates the kd-tree, which must be done whenever the data source (the CFeatureList) changes.

Definition at line 68 of file descriptor_kdtrees.h.

References mrpt::vision::TSIFTDescriptorsKDTreeIndex< distance_t, metric_t >::m_adaptor, mrpt::vision::TSIFTDescriptorsKDTreeIndex< distance_t, metric_t >::m_feats, mrpt::vision::TSIFTDescriptorsKDTreeIndex< distance_t, metric_t >::m_kdtree, params, and mrpt::vision::CFeatureList::size().

Referenced by mrpt::vision::TSIFTDescriptorsKDTreeIndex< distance_t, metric_t >::TSIFTDescriptorsKDTreeIndex().

Here is the call graph for this function:

Here is the caller graph for this function:

Member Data Documentation

m_adaptor

template<typename distance_t , class metric_t = nanoflann::L2_Simple_Adaptor< uint8_t , detail::TSIFTDesc2KDTree_Adaptor<distance_t>, distance_t>>

detail::TSIFTDesc2KDTree_Adaptor<distance_t> mrpt::vision::TSIFTDescriptorsKDTreeIndex< distance_t, metric_t >::m_adaptor

private

Definition at line 88 of file descriptor_kdtrees.h.

Referenced by mrpt::vision::TSIFTDescriptorsKDTreeIndex< distance_t, metric_t >::regenerate_kdtreee().

m_feats

template<typename distance_t , class metric_t = nanoflann::L2_Simple_Adaptor< uint8_t , detail::TSIFTDesc2KDTree_Adaptor<distance_t>, distance_t>>

const CFeatureList& mrpt::vision::TSIFTDescriptorsKDTreeIndex< distance_t, metric_t >::m_feats

private

Definition at line 91 of file descriptor_kdtrees.h.

Referenced by mrpt::vision::TSIFTDescriptorsKDTreeIndex< distance_t, metric_t >::regenerate_kdtreee().

m_kdtree

template<typename distance_t , class metric_t = nanoflann::L2_Simple_Adaptor< uint8_t , detail::TSIFTDesc2KDTree_Adaptor<distance_t>, distance_t>>

kdtree_t* mrpt::vision::TSIFTDescriptorsKDTreeIndex< distance_t, metric_t >::m_kdtree

private

Definition at line 89 of file descriptor_kdtrees.h.

Referenced by mrpt::vision::TSIFTDescriptorsKDTreeIndex< distance_t, metric_t >::get_kdtree(), mrpt::vision::TSIFTDescriptorsKDTreeIndex< distance_t, metric_t >::regenerate_kdtreee(), and mrpt::vision::TSIFTDescriptorsKDTreeIndex< distance_t, metric_t >::~TSIFTDescriptorsKDTreeIndex().