Inheritance diagram for statismo::SpatiallyVaryingKernel< T >:

Public Types
typedef Representer< T >	RepresenterType

typedef RepresenterType::PointType	PointType

Public Member Functions
	SpatiallyVaryingKernel (const RepresenterType *representer, const MatrixValuedKernel< PointType > &kernel, const TemperingFunction< PointType > &eta, unsigned numEigenfunctions, unsigned numberOfPointsForApproximation=0, bool cacheValues=true)
	Make a given kernel spatially varying according to the given tempering function. More...

MatrixType	operator() (const PointType &x, const PointType &y) const

std::string	GetKernelInfo () const

Public Member Functions inherited from statismo::MatrixValuedKernel< Representer< T >::PointType >
	MatrixValuedKernel (unsigned dim)

virtual MatrixType	operator() (const Representer< T >::PointType &x, const Representer< T >::PointType &y) const =0

virtual unsigned	GetDimension () const

Additional Inherited Members
Protected Attributes inherited from statismo::MatrixValuedKernel< Representer< T >::PointType >
unsigned	m_dimension

Detailed Description

spatially-varing kernel, as described in the paper:

T. Gerig, K. Shahim, M. Reyes, T. Vetter, M. Luethi Spatially varying registration using gaussian processes Miccai 2014

Constructor & Destructor Documentation

template<class T >

statismo::SpatiallyVaryingKernel< T >::SpatiallyVaryingKernel	(	const RepresenterType *	representer,
		const MatrixValuedKernel< PointType > &	kernel,
		const TemperingFunction< PointType > &	eta,
		unsigned	numEigenfunctions,
		unsigned	numberOfPointsForApproximation = `0`,
		bool	cacheValues = `true`
	)

inline

Make a given kernel spatially varying according to the given tempering function.

Parameters

representer,A	representer which defines the domain over which the approximation is done
kernel	The kernel that is made spatially adaptive
eta	The tempering function that defines the amount of tempering for each point in the domain
numEigenfunctions	The number of eigenfunctions to be used for the approximation
numberOfPointsForApproximation	The number of points used for the nystrom approximation
cacheValues	Cache result of eigenfunction computations. Greatly speeds up the computation.

template<class T >

std::string statismo::SpatiallyVaryingKernel< T >::GetKernelInfo ( ) const

inlinevirtual

Return a description of this kernel.

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