Feature detection and tracking

Detailed Description

Collaboration diagram for Feature detection and tracking:

Classes
struct	mrpt::vision::CGenericFeatureTracker
	A virtual interface for all feature trackers, implementing the part of feature tracking that is common to any specific tracker implementation. More...
struct	mrpt::vision::CFeatureTracker_KL
	Track a set of features from old_img -> new_img using sparse optimal flow (classic KL method). More...

Typedefs
using	mrpt::vision::CGenericFeatureTrackerAutoPtr = std::unique_ptr< CGenericFeatureTracker >

Functions
void	mrpt::vision::checkTrackedFeatures (CFeatureList &leftList, CFeatureList &rightList, vision::TMatchingOptions options)
	Search for correspondences which are not in the same row and deletes them ... More...
void	mrpt::vision::filterBadCorrsByDistance (mrpt::tfest::TMatchingPairList &list, unsigned int numberOfSigmas)
	Filter bad correspondences by distance ... More...

Typedef Documentation

CGenericFeatureTrackerAutoPtr

using mrpt::vision::CGenericFeatureTrackerAutoPtr = typedef std::unique_ptr<CGenericFeatureTracker>

Definition at line 293 of file tracking.h.

Function Documentation

checkTrackedFeatures()

void mrpt::vision::checkTrackedFeatures	(	CFeatureList &	leftList,
		CFeatureList &	rightList,
		vision::TMatchingOptions	options
	)

Search for correspondences which are not in the same row and deletes them ...

Definition at line 815 of file tracking.cpp.

References ASSERT_, mrpt::vision::CFeatureList::begin(), mrpt::vision::CFeatureList::end(), mrpt::vision::TMatchingOptions::epipolar_TH, mrpt::vision::CFeatureList::erase(), mrpt::vision::TMatchingOptions::minCC_TH, mrpt::vision::openCV_cross_correlation(), and mrpt::vision::CFeatureList::size().

filterBadCorrsByDistance()

void mrpt::vision::filterBadCorrsByDistance	(	mrpt::tfest::TMatchingPairList &	list,
		unsigned int	numberOfSigmas
	)

Filter bad correspondences by distance ...

Definition at line 881 of file tracking.cpp.

References ASSERT_, MRPT_END, MRPT_START, and mrpt::square().