Matches features between a reference and a target image. More...

Public Member Functions
	FeatureMatcher ()
void	matchFeatures (PyramidLevel ref_level, PyramidLevel target_level, const std::vector< std::vector< int > > &candidates, FeatureMatch matches, int num_matches)
virtual	~FeatureMatcher ()
Private Member Functions
	FeatureMatcher (const FeatureMatcher &other)
FeatureMatcher &	operator= (const FeatureMatcher &other)
Private Attributes
int	_ref_feature_capacity
int32_t *	_ref_to_target_indices
int32_t *	_ref_to_target_scores
int	_target_feature_capacity
int32_t *	_target_to_ref_indices
int32_t *	_target_to_ref_scores

Detailed Description

Matches features between a reference and a target image.

Definition at line 12 of file feature_matcher.hpp.

Constructor & Destructor Documentation

fovis::FeatureMatcher::FeatureMatcher ( )

Definition at line 10 of file feature_matcher.cpp.

fovis::FeatureMatcher::~FeatureMatcher ( ) [virtual]

Definition at line 32 of file feature_matcher.cpp.

fovis::FeatureMatcher::FeatureMatcher ( const FeatureMatcher & other ) [private]

void fovis::FeatureMatcher::matchFeatures	(	PyramidLevel *	ref_level,
		PyramidLevel *	target_level,
		const std::vector< std::vector< int > > &	candidates,
		FeatureMatch *	matches,
		int *	num_matches
	)

Feature matching using sum of absolute differences (SAD).

Parameters:

ref_level	features in the reference image.
target_level	features in the target image.
candidates	identifies potential match candidates for each feature in the reference image. For every reference feature, there is a vector of target feature indices that is a potential match.
matches	output array of matches. This should be pre-allocated and of size at least min(num features in `ref_level`, num features in `target_level`)
num_matches	output parameter, is set to the number of features matched.