A virtual kernel used for the ACRANSAC / LORANSAC framework. More...

#include <PointFittingRansacKernel.hpp>

Inheritance diagram for aliceVision::robustEstimation::PointFittingRansacKernel< SolverT_, ErrorT_, ModelT_, SolverLsT_ >:

Public Types
using	PFKernel = robustEstimation::PointFittingKernel< SolverT_, ErrorT_, ModelT_ >

Public Types inherited from aliceVision::robustEstimation::PointFittingKernel< SolverT_, ErrorT_, ModelT_ >
using	SolverT = SolverT_

using	ErrorT = ErrorT_

using	ModelT = ModelT_

Public Member Functions
	PointFittingRansacKernel (const Mat &x1, const Mat &x2)

std::size_t	getMinimumNbRequiredSamples () const override
	Return the minimum number of required samples for the solver. More...

std::size_t	getMinimumNbRequiredSamplesLS () const override

std::size_t	getMaximumNbModels () const override
	Return the maximum number of models for the solver. More...

void	fit (const std::vector< std::size_t > &samples, std::vector< ModelT_ > &models) const override
	This function is called to estimate the model from the minimum number of sample `minSample` (i.e. minimal problem solver). More...

void	fitLS (const std::vector< std::size_t > &inliers, std::vector< ModelT_ > &models, const std::vector< double > *weights=nullptr) const override
	This function is called to estimate the model using a least squared algorithm from a minimum of `minSampleLS`. More...

void	computeWeights (const ModelT_ &model, const std::vector< std::size_t > &inliers, std::vector< double > &weights, const double eps=0.001) const override
	Function used to estimate the weights, typically used by the least square algorithm. More...

double	error (std::size_t sample, const ModelT_ &model) const override
	Function that computes the estimation error for a given model and a given element. More...

void	errors (const ModelT_ &model, std::vector< double > &errors) const override
	Function that computes the estimation error for a given model and all the elements. More...

virtual void	unnormalize (ModelT_ &model) const =0
	Function used to unnormalize the model. More...

std::size_t	nbSamples () const override
	The number of elements in the data. More...

virtual double	logalpha0 () const =0
	Get logalpha0, Alpha0 is used to make the error adaptive to the image size. More...

virtual double	errorVectorDimension () const =0

virtual double	unormalizeError (double val) const =0

virtual Mat3	normalizer1 () const =0

virtual double	thresholdNormalizer () const =0

Public Member Functions inherited from aliceVision::robustEstimation::IRansacKernel< ModelT_ >
virtual std::size_t	getMinimumNbRequiredSamples () const=0
	Return the minimum number of required samples for the solver. More...

virtual std::size_t	getMinimumNbRequiredSamplesLS () const=0
	Return the minimum number of required samples for the solver Ls. More...

virtual std::size_t	getMaximumNbModels () const=0
	Return the maximum number of models for the solver. More...

virtual std::size_t	nbSamples () const=0
	The number of elements in the data. More...

virtual double	logalpha0 () const=0
	Get logalpha0, Alpha0 is used to make the error adaptive to the image size. More...

virtual double	errorVectorDimension () const=0

virtual Mat3	normalizer1 () const=0

virtual double	thresholdNormalizer () const=0

Public Member Functions inherited from aliceVision::robustEstimation::PointFittingKernel< SolverT_, ErrorT_, ModelT_ >
	PointFittingKernel (const Mat &x1, const Mat &x2)

std::size_t	getMinimumNbRequiredSamples () const
	Return the minimum number of required samples. More...

std::size_t	getMaximumNbModels () const
	Return the maximum number of models. More...

std::size_t	nbSamples () const
	get the number of putative points More...

Additional Inherited Members
Protected Attributes inherited from aliceVision::robustEstimation::PointFittingKernel< SolverT_, ErrorT_, ModelT_ >
const Mat &	_x1
	left corresponding data

const Mat &	_x2
	right corresponding data

const SolverT	_kernelSolver {}
	two view solver

const ErrorT	_errorEstimator {}
	solver error estimation

Detailed Description

template<typename SolverT_, typename ErrorT_, typename ModelT_, typename SolverLsT_ = robustEstimation::UndefinedSolver<ModelT_>>
class aliceVision::robustEstimation::PointFittingRansacKernel< SolverT_, ErrorT_, ModelT_, SolverLsT_ >

A virtual kernel used for the ACRANSAC / LORANSAC framework.

Template Parameters

SolverT_	The minimal solver able to find a solution from a minimum set of points.
ErrorT_	The functor computing the error for each data sample with respect to the estimated model.
UnnormalizerT_	The functor used to normalize the data before the estimation of the model.
ModelT_	= Mat34Model The type of the model to estimate.
SolverLsT_	= SolverT The least square solver that is used to find a solution from any set of data larger than the minimum required.

Note: Handle data normalization and compute the corresponding logalpha 0 that depends of the error model (point to line, or point to point)

Member Function Documentation

◆ computeWeights()

template<typename SolverT_ , typename ErrorT_ , typename ModelT_ , typename SolverLsT_ = robustEstimation::UndefinedSolver<ModelT_>>

void aliceVision::robustEstimation::PointFittingRansacKernel< SolverT_, ErrorT_, ModelT_, SolverLsT_ >::computeWeights	(	const ModelT_ &	model,
		const std::vector< std::size_t > &	inliers,
		std::vector< double > &	weights,
		const double	eps = `0.001`
	)		const

inlineoverridevirtual

Function used to estimate the weights, typically used by the least square algorithm.

Parameters

[in]	model	The model against which the weights are computed.
[in]	inliers	The array of the indices of the data to be used.
[out]	vec_weights	The array of weight of the same size as `inliers`.
[in]	eps	An optional threshold to max out the value of the threshold (typically to avoid division by zero or too small numbers).

Implements aliceVision::robustEstimation::IRansacKernel< ModelT_ >.

◆ error()

template<typename SolverT_ , typename ErrorT_ , typename ModelT_ , typename SolverLsT_ = robustEstimation::UndefinedSolver<ModelT_>>

double aliceVision::robustEstimation::PointFittingRansacKernel< SolverT_, ErrorT_, ModelT_, SolverLsT_ >::error	(	std::size_t	sample,
		const ModelT_ &	model
	)		const

inlineoverridevirtual

Function that computes the estimation error for a given model and a given element.

Parameters

[in]	sample	The index of the element for which the error is computed.
[in]	model	The model to consider.

Returns: The estimation error for the given element and the given model.

Reimplemented from aliceVision::robustEstimation::PointFittingKernel< SolverT_, ErrorT_, ModelT_ >.

◆ errors()

template<typename SolverT_ , typename ErrorT_ , typename ModelT_ , typename SolverLsT_ = robustEstimation::UndefinedSolver<ModelT_>>

void aliceVision::robustEstimation::PointFittingRansacKernel< SolverT_, ErrorT_, ModelT_, SolverLsT_ >::errors	(	const ModelT_ &	model,
		std::vector< double > &	errors
	)		const

inlineoverridevirtual

Function that computes the estimation error for a given model and all the elements.

Parameters

[in]	model	The model to consider.
[out]	vec_errors	The vector containing all the estimation errors for every element.

Reimplemented from aliceVision::robustEstimation::PointFittingKernel< SolverT_, ErrorT_, ModelT_ >.

◆ fit()

template<typename SolverT_ , typename ErrorT_ , typename ModelT_ , typename SolverLsT_ = robustEstimation::UndefinedSolver<ModelT_>>

void aliceVision::robustEstimation::PointFittingRansacKernel< SolverT_, ErrorT_, ModelT_, SolverLsT_ >::fit	(	const std::vector< std::size_t > &	samples,
		std::vector< ModelT_ > &	models
	)		const

inlineoverridevirtual

This function is called to estimate the model from the minimum number of sample minSample (i.e. minimal problem solver).

Parameters

[in]	samples	A vector containing the indices of the data to be used for the minimal estimation.
[out]	models	The model(s) estimated by the minimal solver.

Reimplemented from aliceVision::robustEstimation::PointFittingKernel< SolverT_, ErrorT_, ModelT_ >.

◆ fitLS()

template<typename SolverT_ , typename ErrorT_ , typename ModelT_ , typename SolverLsT_ = robustEstimation::UndefinedSolver<ModelT_>>

void aliceVision::robustEstimation::PointFittingRansacKernel< SolverT_, ErrorT_, ModelT_, SolverLsT_ >::fitLS	(	const std::vector< std::size_t > &	inliers,
		std::vector< ModelT_ > &	models,
		const std::vector< double > *	weights = `nullptr`
	)		const

inlineoverridevirtual

This function is called to estimate the model using a least squared algorithm from a minimum of minSampleLS.

Parameters

[in]	inliers	An array containing the indices of the data to use.
[out]	models	The model(s) estimated using the least squared algorithm.
[in]	weights	An optional array of weights, one for each sample

Implements aliceVision::robustEstimation::IRansacKernel< ModelT_ >.

◆ getMaximumNbModels()

template<typename SolverT_ , typename ErrorT_ , typename ModelT_ , typename SolverLsT_ = robustEstimation::UndefinedSolver<ModelT_>>

std::size_t aliceVision::robustEstimation::PointFittingRansacKernel< SolverT_, ErrorT_, ModelT_, SolverLsT_ >::getMaximumNbModels ( ) const

inlineoverride

Return the maximum number of models for the solver.

Returns: maximum number of models

◆ getMinimumNbRequiredSamples()

template<typename SolverT_ , typename ErrorT_ , typename ModelT_ , typename SolverLsT_ = robustEstimation::UndefinedSolver<ModelT_>>

std::size_t aliceVision::robustEstimation::PointFittingRansacKernel< SolverT_, ErrorT_, ModelT_, SolverLsT_ >::getMinimumNbRequiredSamples ( ) const

inlineoverride

Return the minimum number of required samples for the solver.

Returns: minimum number of required samples

◆ logalpha0()

template<typename SolverT_ , typename ErrorT_ , typename ModelT_ , typename SolverLsT_ = robustEstimation::UndefinedSolver<ModelT_>>

virtual double aliceVision::robustEstimation::PointFittingRansacKernel< SolverT_, ErrorT_, ModelT_, SolverLsT_ >::logalpha0 ( ) const

pure virtual

Get logalpha0, Alpha0 is used to make the error adaptive to the image size.

Returns: logalpha0

◆ nbSamples()

template<typename SolverT_ , typename ErrorT_ , typename ModelT_ , typename SolverLsT_ = robustEstimation::UndefinedSolver<ModelT_>>

std::size_t aliceVision::robustEstimation::PointFittingRansacKernel< SolverT_, ErrorT_, ModelT_, SolverLsT_ >::nbSamples ( ) const

inlineoverride

The number of elements in the data.

Returns: the number of elements in the data.

◆ unnormalize()

template<typename SolverT_ , typename ErrorT_ , typename ModelT_ , typename SolverLsT_ = robustEstimation::UndefinedSolver<ModelT_>>

virtual void aliceVision::robustEstimation::PointFittingRansacKernel< SolverT_, ErrorT_, ModelT_, SolverLsT_ >::unnormalize ( ModelT_ & model ) const

pure virtual

Function used to unnormalize the model.

Parameters

[in,out] model The model to unnormalize.

Implements aliceVision::robustEstimation::IRansacKernel< ModelT_ >.

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

/home/docs/checkouts/readthedocs.org/user_builds/alicevision/checkouts/latest/src/aliceVision/robustEstimation/PointFittingRansacKernel.hpp

Public Types

Public Member Functions

Additional Inherited Members

Detailed Description

template<typename SolverT_, typename ErrorT_, typename ModelT_, typename SolverLsT_ = robustEstimation::UndefinedSolver<ModelT_>> class aliceVision::robustEstimation::PointFittingRansacKernel< SolverT_, ErrorT_, ModelT_, SolverLsT_ >

Member Function Documentation

◆ computeWeights()

◆ error()

◆ errors()

◆ fit()

◆ fitLS()

◆ getMaximumNbModels()

◆ getMinimumNbRequiredSamples()

◆ logalpha0()

◆ nbSamples()

◆ unnormalize()

template<typename SolverT_, typename ErrorT_, typename ModelT_, typename SolverLsT_ = robustEstimation::UndefinedSolver<ModelT_>>
class aliceVision::robustEstimation::PointFittingRansacKernel< SolverT_, ErrorT_, ModelT_, SolverLsT_ >