This class is meant to filter observations from perception to try to reduce the presence of outliers. More...

#include <base_landmark_filter.hpp>

Inheritance diagram for LandmarkFilter:

Collaboration diagram for LandmarkFilter:

Public Member Functions
	LandmarkFilter ()=default

	LandmarkFilter (LandmarkFilterParameters params, std::shared_ptr< DataAssociationModel > data_association)

virtual	~LandmarkFilter ()=default

virtual Eigen::VectorXd	filter (const Eigen::VectorXd &observations, const Eigen::VectorXd &observation_confidences, Eigen::VectorXi &associations)=0
	This function receives a new set of observations and returns a filtered set of landmarks in global coordinates that are ready to be added to be added to the map.

virtual void	delete_landmarks (const Eigen::VectorXd &some_landmarks)=0
	Used by SLAM to signal to the filter that the landmarks are already in SLAM's map, and they should no longer be returned by the filter as new.

Protected Attributes
LandmarkFilterParameters	_params_

std::shared_ptr< DataAssociationModel >	_data_association_

Detailed Description

This class is meant to filter observations from perception to try to reduce the presence of outliers.

Definition at line 14 of file base_landmark_filter.hpp.

Constructor & Destructor Documentation

◆ LandmarkFilter() [1/2]

LandmarkFilter::LandmarkFilter ( )

default

◆ LandmarkFilter() [2/2]

LandmarkFilter::LandmarkFilter	(	LandmarkFilterParameters	params,
		std::shared_ptr< DataAssociationModel >	data_association
	)

inline

Definition at line 21 of file base_landmark_filter.hpp.

◆ ~LandmarkFilter()

virtual LandmarkFilter::~LandmarkFilter ( )

virtualdefault

Member Function Documentation

◆ delete_landmarks()

virtual void LandmarkFilter::delete_landmarks ( const Eigen::VectorXd & some_landmarks )

pure virtual

Used by SLAM to signal to the filter that the landmarks are already in SLAM's map, and they should no longer be returned by the filter as new.

Parameters

some_landmarks landmarks to be deleted in the form of [x1, y1, x2, y2, ...] in the global frame

Implemented in ConsecutiveCounterFilter.

◆ filter()

virtual Eigen::VectorXd LandmarkFilter::filter	(	const Eigen::VectorXd &	observations,
		const Eigen::VectorXd &	observation_confidences,
		Eigen::VectorXi &	associations
	)

pure virtual

This function receives a new set of observations and returns a filtered set of landmarks in global coordinates that are ready to be added to be added to the map.

Parameters

observations	Observations in the form of [x1, y1, x2, y2, ...] in the global frame
observation_confidences	Confidence in the observations in the same order as the observations
associations	Associations of observations to landmarks

Returns: Eigen::VectorXd the filtered observations in the form of [x1, y1, x2, y2, ...] in the global frame

Implemented in ConsecutiveCounterFilter.

Member Data Documentation

◆ _data_association_

std::shared_ptr<DataAssociationModel> LandmarkFilter::_data_association_

protected

Definition at line 17 of file base_landmark_filter.hpp.

◆ _params_

LandmarkFilterParameters LandmarkFilter::_params_