mirtk::Transformation Class Reference

#include <Transformation.h>

Inheritance diagram for mirtk::Transformation:

Collaboration diagram for mirtk::Transformation:

Public Types
typedef Status	DOFStatus
	Type of transforamtion parameter status.
typedef double	DOFValue
	Type of transformation parameter value.

Public Member Functions
virtual void	Add (const DOFValue *)
	Add change to transformation parameters.
virtual double	Approximate (const ImageAttributes &, const Transformation *, int=1, double=.0)
	Approximate another transformation and return approximation error.
virtual double	Approximate (GenericImage< double > &, int=1, double=.0)
virtual double	Approximate (const ImageAttributes &, double *, double *, double *, int=1, double=.0)
virtual double	Approximate (const double *, const double *, const double *, double *, double *, double *, int, int=1, double=.0)
virtual double	Approximate (const double *, const double *, const double *, const double *, double *, double *, double *, int, int=1, double=.0)
virtual double	ApproximateAsNew (const ImageAttributes &, const Transformation *, int=1, double=.0)
	Approximate another transformation and return approximation error.
virtual double	ApproximateAsNew (GenericImage< double > &, int=1, double=.0)
virtual double	ApproximateAsNew (const ImageAttributes &, double *, double *, double *, int=1, double=.0)
virtual double	ApproximateAsNew (const double *, const double *, const double *, double *, double *, double *, int, int=1, double=.0)
virtual double	ApproximateAsNew (const double *, const double *, const double *, const double *, double *, double *, double *, int, int=1, double=.0)
virtual void	ApproximateDOFs (const double *, const double *, const double *, const double *, const double *, const double *, const double *, int)
virtual void	ApproximateDOFsGradient (const double *, const double *, const double *, const double *, const double *, const double *, const double *, int, double *, double=1.0) const
virtual void	ApproximateGradient (const ImageAttributes &, const double *, const double *, const double *, double *, double=1.0) const
virtual void	ApproximateGradient (const double *, const double *, const double *, const double *, const double *, const double *, int, double *, double=1.0) const
virtual void	ApproximateGradient (const double *, const double *, const double *, const double *, const double *, const double *, const double *, int, double *, double=1.0) const
virtual bool	CanModifyDisplacement (int=-1) const
virtual bool	CanRead (TransformationType) const
	Whether this transformation can read a file of specified type (i.e. format)
virtual bool	CopyFrom (const Transformation *)
virtual void	DeriveJacobianWrtDOF (Matrix &, int, double, double, double, double=0, double=NaN) const
	Calculates the derivative of the Jacobian of the transformation (w.r.t. world coordinates) w.r.t. a transformation parameter.
virtual void	Displacement (double &, double &, double &, double=0, double=NaN) const
	Calculates the displacement of a single point.
virtual void	Displacement (const ImageAttributes &, double *, double *, double *) const
virtual void	Displacement (GenericImage< double > &, double=NaN, const WorldCoordsImage *=NULL) const
	Calculates the displacement vectors for a whole image domain.
virtual void	Displacement (GenericImage< float > &, double=NaN, const WorldCoordsImage *=NULL) const
	Calculates the displacement vectors for a whole image domain.
virtual void	Displacement (GenericImage< double > &, double, double, const WorldCoordsImage *=NULL) const
virtual void	Displacement (GenericImage< float > &, double, double, const WorldCoordsImage *=NULL) const
virtual void	DisplacementAfterDOFChange (int dof, double dv, GenericImage< double > &dx, double t, double t0=NaN, const WorldCoordsImage *i2w=NULL) const
virtual bool	DOFBoundingBox (const Image *, int, int &, int &, int &, int &, int &, int &, double=1) const
virtual double	DOFGradientNorm (const double *) const
	Get norm of the gradient vector.
double	EvaluateRMSError (const ImageAttributes &, const Transformation *) const
	Evaluates RMS error of transformation compared to another.
double	EvaluateRMSError (const ImageAttributes &, double *, double *) const
double	EvaluateRMSError (const ImageAttributes &, double *, double *, double *) const
double	EvaluateRMSError (const double *, const double *, const double *, double, double *, double *, double *, int no) const
	Evaluates RMS error of transformation compared to displacement field.
double	EvaluateRMSError (const double *, const double *, const double *, const double *, double *, double *, double *, int no) const
	Evaluates RMS error of transformation compared to given displacement field.
virtual double	Get (int) const
	Get value of transformation parameter.
virtual void	Get (DOFValue *) const
	Get values of transformation parameters.
virtual DOFStatus	GetStatus (int) const
	Get status of transformation parameter.
virtual void	GlobalDisplacement (double &, double &, double &, double=0, double=NaN) const
	Calculates the displacement of a single point using the global transformation component only.
virtual void	GlobalHessian (Matrix [3], double, double, double, double=0, double=NaN) const
	Calculates the Hessian for each component of the global transformation w.r.t world coordinates.
virtual void	GlobalInverse (double &, double &, double &, double=0, double=NaN) const
	Transforms a single point using the inverse of the global transformation only.
virtual void	GlobalInverseDisplacement (double &, double &, double &, double=0, double=NaN) const
	Calculates the displacement of a single point using the inverse of the global transformation only.
virtual void	GlobalJacobian (Matrix &, double, double, double, double=0, double=NaN) const
	Calculates the Jacobian of the global transformation w.r.t world coordinates.
virtual double	GlobalJacobian (double, double, double, double=0, double=NaN) const
	Calculates the determinant of the Jacobian of the global transformation w.r.t world coordinates.
virtual void	GlobalTransform (double &, double &, double &, double=0, double=NaN) const =0
	Transforms a single point using the global transformation component only.
virtual bool	HasSameDOFsAs (const Transformation *) const
	Checks whether transformation depends on the same vector of parameters.
virtual void	Hessian (Matrix [3], double, double, double, double=0, double=NaN) const
	Calculates the Hessian for each component of the transformation w.r.t world coordinates.
virtual bool	Inverse (double &, double &, double &, double=0, double=NaN) const
	Transforms a single point using the inverse of the transformation.
virtual bool	Inverse (Point &, double=0, double=NaN) const
	Transforms a single point using the inverse of the transformation.
virtual int	Inverse (PointSet &, double=0, double=NaN) const
virtual bool	InverseDisplacement (double &, double &, double &, double=0, double=NaN) const
	Calculates the displacement of a single point using the inverse of the transformation.
virtual int	InverseDisplacement (const ImageAttributes &, double *, double *, double *) const
virtual int	InverseDisplacement (GenericImage< double > &, double=NaN, const WorldCoordsImage *=NULL) const
virtual int	InverseDisplacement (GenericImage< float > &, double=NaN, const WorldCoordsImage *=NULL) const
virtual int	InverseDisplacement (GenericImage< double > &, double, double, const WorldCoordsImage *=NULL) const
virtual int	InverseDisplacement (GenericImage< float > &, double, double, const WorldCoordsImage *=NULL) const
virtual bool	IsIdentity () const
	Checks whether the transformation is an identity mapping.
virtual void	Jacobian (Matrix &, double, double, double, double=0, double=NaN) const
	Calculates the Jacobian of the transformation w.r.t world coordinates.
virtual double	Jacobian (double, double, double, double=0, double=NaN) const
	Calculates the determinant of the Jacobian of the transformation w.r.t world coordinates.
virtual void	JacobianDOFs (double [3], int, double, double, double, double=0, double=NaN) const
	Calculates the Jacobian of the transformation w.r.t a transformation parameter.
virtual void	LocalDisplacement (double &, double &, double &, double=0, double=NaN) const
	Calculates the displacement of a single point using the local transformation component only.
virtual void	LocalHessian (Matrix [3], double, double, double, double=0, double=NaN) const
	Calculates the Hessian for each component of the local transformation w.r.t world coordinates.
virtual bool	LocalInverse (double &, double &, double &, double=0, double=NaN) const
	Transforms a single point using the inverse of the local transformation only.
virtual bool	LocalInverseDisplacement (double &, double &, double &, double=0, double=NaN) const
	Calculates the displacement of a single point using the inverse of the local transformation only.
virtual void	LocalJacobian (Matrix &, double, double, double, double=0, double=NaN) const
	Calculates the Jacobian of the local transformation w.r.t world coordinates.
virtual double	LocalJacobian (double, double, double, double=0, double=NaN) const
	Calculates the determinant of the Jacobian of the local transformation w.r.t world coordinates.
virtual void	LocalTransform (double &, double &, double &, double=0, double=NaN) const =0
	Transforms a single point using the local transformation component only.
int	NumberOfActiveDOFs () const
	Get number of active transformation parameters.
virtual int	NumberOfDOFs () const
	Get number of transformation parameters.
int	NumberOfPassiveDOFs () const
	Get number of passive transformation parameters.
virtual ParameterList	Parameter () const
	Get (non-DoF) parameters as key/value as string map.
virtual void	ParametricGradient (const GenericImage< double > *, double *, const WorldCoordsImage *, const WorldCoordsImage *, double=NaN, double=1) const
void	ParametricGradient (const GenericImage< double > *, double *, const WorldCoordsImage *, double=NaN, double=1) const
void	ParametricGradient (const GenericImage< double > *, double *, double=NaN, double=1) const
virtual void	ParametricGradient (const GenericImage< double > **, int, double *, const WorldCoordsImage *, const WorldCoordsImage *, const double *=NULL, double=1) const
void	ParametricGradient (const GenericImage< double > **, int, double *, const WorldCoordsImage *, const double *=NULL, double=1) const
void	ParametricGradient (const GenericImage< double > **, int, double *, const double *=NULL, double=1) const
virtual void	ParametricGradient (const PointSet &, const Vector3D< double > *, double *, double=0, double=NaN, double=1) const
void	Print (Indent=0) const
	Prints information about the transformation.
virtual void	Print (ostream &os, Indent=0) const =0
	Prints information about the transformation.
virtual void	Put (int, double)
	Put value of transformation parameter.
virtual void	Put (const DOFValue *)
	Put values of transformation parameters.
virtual void	PutStatus (int, DOFStatus)
	Put status of transformation parameter.
virtual void	Read (const char *)
	Reads a transformation from a file.
virtual Cifstream &	Read (Cifstream &)
	Reads a transformation from a file stream.
virtual bool	RequiresCachingOfDisplacements () const
virtual void	Reset ()
	Reset transformation.
virtual bool	Set (const char *, const char *)
	Set named (non-DoF) parameter from value as string.
virtual void	Transform (double &, double &, double &, double=0, double=NaN) const =0
	Transforms a single point.
virtual void	Transform (Point &, double=0, double=NaN) const
	Transforms a single point.
virtual void	Transform (PointSet &, double=0, double=NaN) const
	Transforms a set of points.
virtual void	Transform (int, double *, double *, double *, double=0, double=NaN) const
	Transforms a set of points.
virtual void	Transform (int, double *, double *, double *, const double *, double=NaN) const
	Transforms a set of points.
virtual void	Transform (WorldCoordsImage &, double=NaN) const
	Transforms world coordinates of image voxels.
virtual TransformationType	TypeOfClass () const
	Returns type ID of the instantiated transformation class.
virtual double	Update (const DOFValue *)
	Update transformation parameters given parametric gradient.
virtual void	Verify ()
virtual void	Write (const char *) const
	Writes a transformation to a file.
virtual Cofstream &	Write (Cofstream &) const
	Writes a transformation to a file stream.
virtual	~Transformation ()
	Default destructor.
Public Member Functions inherited from mirtk::Observable
void	AddObserver (Observer &)
	Add observer.
void	Broadcast (Event, const void *=NULL)
	Broadcast event to observers.
void	ClearObservers ()
	Delete all observers.
void	DeleteObserver (Observer &)
	Delete observer.
void	NotifyObservers (Event, const void *=NULL)
	Notify all observers about given event if this object has changed.
int	NumberOfObservers () const
	Number of current observers.
virtual	~Observable ()
	Destructor.
Public Member Functions inherited from mirtk::Object
virtual const char *	NameOfClass () const =0
	Get name of class, which this object is an instance of.
bool	Parameter (const ParameterList &)
	Set parameters from name/value pairs.
virtual	~Object ()
	Destructor.

Static Public Member Functions
static bool	CheckHeader (const char *)
	Whether magic number in header of given file.
static Transformation *	New (TransformationType)
static Transformation *	New (const Transformation *)
static Transformation *	New (const char *)
static TransformationType	TypeOfClass (const char *)
	Returns type ID corresponding to transformations of the named class.
Static Public Member Functions inherited from mirtk::Object
static const char *	NameOfType ()
	Get name of this class type.

Protected Member Functions
void	InitializeDOFs (int)
	Initialize transformation parameters.
void	InitializeDOFs (const Transformation &, int=-1)
	Copy transformation parameters (DoFs) and their status.
virtual Cifstream &	ReadDOFs (Cifstream &, TransformationType)
	Reads transformation parameters from a file stream.
	Transformation (int=0)
	Default constructor.
	Transformation (const Transformation &)
	Copy constructor.
	Transformation (const Transformation &, int)
	Copy constructor.
virtual Cofstream &	WriteDOFs (Cofstream &) const
	Writes transformation parameters to a file stream.
Protected Member Functions inherited from mirtk::Observable
	Observable ()
	Default constructor.
	Observable (const Observable &)
	Copy constructor.
Observable &	operator= (const Observable &)
	Assignment operator.
Protected Member Functions inherited from mirtk::Object
template<typename... Args>
void	Throw (ErrorType err, const char *func, Args... args) const

Protected Attributes
int	_NumberOfDOFs
	Number of transformation parameters.
DOFValue *	_Param
	Value of each transformation parameter.
DOFStatus *	_Status
	Status of each transformation parameter (Active or Passive)

Additional Inherited Members
Static Protected Member Functions inherited from mirtk::Object
template<typename... Args>
static void	ThrowStatic (ErrorType err, const char *cls, const char *func, Args... args)

Detailed Description

Abstract base class for general transformations.

This is the abstract base class which defines a common interface for all transformations. Each derived class has to implement at least the abstract methods and some of the virtual ones. Most other methods call these virtual methods and should not be required to be overwritten in subclasses.

The second time argument to the interface methods corresponds to the time of the untransformed source image. It is only considered by some 3D+t transformations, in particular those which parameterize the transformation using a non-stationary velocity field.

Definition at line 62 of file Transformation.h.

Member Function Documentation

Approximate() [1/4]

virtual double mirtk::Transformation::Approximate ( GenericImage< double > &  , int  = 1, double  = .0  )

virtual

Approximate displacements: This function takes a set of points and a set of displacements and finds a transformation which approximates these displacements. After approximation, the displacements are replaced by the residual displacement errors at the points.

Approximate() [2/4]

virtual double mirtk::Transformation::Approximate ( const ImageAttributes &  , double *  , double *  , double *  , int  = 1, double  = .0  )

virtual

Approximate displacements: This function takes a set of points and a set of displacements and finds a transformation which approximates these displacements. After approximation, the displacements are replaced by the residual displacement errors at the points.

Reimplemented in mirtk::FreeFormTransformation, mirtk::FluidFreeFormTransformation, mirtk::HomogeneousTransformation, mirtk::MultiLevelTransformation, and mirtk::MultiLevelFreeFormTransformation.

Approximate() [3/4]

virtual double mirtk::Transformation::Approximate ( const double *  , const double *  , const double *  , double *  , double *  , double *  , int  , int  = 1, double  = .0  )

virtual

Approximate displacements: This function takes a set of points and a set of displacements and finds a transformation which approximates these displacements. After approximation, the displacements are replaced by the residual displacement errors at the points.

Reimplemented in mirtk::FreeFormTransformation, mirtk::FluidFreeFormTransformation, mirtk::HomogeneousTransformation, mirtk::MultiLevelTransformation, and mirtk::MultiLevelFreeFormTransformation.

Approximate() [4/4]

virtual double mirtk::Transformation::Approximate ( const double *  , const double *  , const double *  , const double *  , double *  , double *  , double *  , int  , int  = 1, double  = .0  )

virtual

Approximate displacements: This function takes a set of points and a set of displacements and finds a transformation which approximates these displacements. After approximation, the displacements are replaced by the residual displacement errors at the points.

Reimplemented in mirtk::FreeFormTransformation, mirtk::FluidFreeFormTransformation, mirtk::HomogeneousTransformation, mirtk::MultiLevelTransformation, and mirtk::MultiLevelFreeFormTransformation.

ApproximateAsNew() [1/4]

virtual double mirtk::Transformation::ApproximateAsNew ( GenericImage< double > &  , int  = 1, double  = .0  )

virtual

Approximate displacements: This function takes a set of points and a set of displacements and finds a !new! transformation which approximates these displacements. After approximation, the displacements are replaced by the residual displacement errors at the points.

Reimplemented in mirtk::BSplineFreeFormTransformationSV.

ApproximateAsNew() [2/4]

virtual double mirtk::Transformation::ApproximateAsNew ( const ImageAttributes &  , double *  , double *  , double *  , int  = 1, double  = .0  )

virtual

Approximate displacements: This function takes a set of points and a set of displacements and finds a !new! transformation which approximates these displacements. After approximation, the displacements are replaced by the residual displacement errors at the points.

Reimplemented in mirtk::FreeFormTransformation, mirtk::FluidFreeFormTransformation, mirtk::MultiLevelTransformation, and mirtk::MultiLevelFreeFormTransformation.

ApproximateAsNew() [3/4]

virtual double mirtk::Transformation::ApproximateAsNew ( const double *  , const double *  , const double *  , double *  , double *  , double *  , int  , int  = 1, double  = .0  )

virtual

Approximate displacements: This function takes a set of points and a set of displacements and finds a !new! transformation which approximates these displacements. After approximation, the displacements are replaced by the residual displacement errors at the points.

Reimplemented in mirtk::FluidFreeFormTransformation, mirtk::MultiLevelTransformation, and mirtk::MultiLevelFreeFormTransformation.

ApproximateAsNew() [4/4]

virtual double mirtk::Transformation::ApproximateAsNew ( const double *  , const double *  , const double *  , const double *  , double *  , double *  , double *  , int  , int  = 1, double  = .0  )

virtual

Approximate displacements: This function takes a set of points and a set of displacements and finds a !new! transformation which approximates these displacements. After approximation, the displacements are replaced by the residual displacement errors at the points.

Reimplemented in mirtk::FluidFreeFormTransformation, mirtk::MultiLevelTransformation, and mirtk::MultiLevelFreeFormTransformation.

ApproximateDOFs()

virtual void mirtk::Transformation::ApproximateDOFs ( const double *  , const double *  , const double *  , const double *  , const double *  , const double *  , const double *  , int    )

virtual

Approximate displacements: This function takes a set of points and a set of displacements and finds !new! parameters such that the resulting transformation approximates the displacements as good as possible.

Reimplemented in mirtk::FluidFreeFormTransformation, mirtk::MultiLevelTransformation, mirtk::MultiLevelFreeFormTransformation, mirtk::HomogeneousTransformation, mirtk::BSplineFreeFormTransformationSV, mirtk::BSplineFreeFormTransformation3D, mirtk::AffineTransformation, mirtk::LinearFreeFormTransformationTD, mirtk::BSplineFreeFormTransformationTD, mirtk::RigidTransformation, mirtk::LinearFreeFormTransformation3D, mirtk::BSplineFreeFormTransformation4D, mirtk::LinearFreeFormTransformation4D, mirtk::BSplineFreeFormTransformationStatistical, and mirtk::SimilarityTransformation.

ApproximateDOFsGradient()

virtual void mirtk::Transformation::ApproximateDOFsGradient ( const double *  , const double *  , const double *  , const double *  , const double *  , const double *  , const double *  , int  , double *  , double  = 1.0  ) const

virtual

Finds gradient of approximation error: This function takes a set of points and a set of errors. It finds a gradient w.r.t. the transformation parameters which minimizes the L2 norm of the approximation error and adds it to the input gradient with the given weight.

Reimplemented in mirtk::FluidFreeFormTransformation, mirtk::MultiLevelTransformation, mirtk::MultiLevelFreeFormTransformation, mirtk::BSplineFreeFormTransformationSV, mirtk::BSplineFreeFormTransformation3D, mirtk::LinearFreeFormTransformationTD, mirtk::BSplineFreeFormTransformationTD, mirtk::LinearFreeFormTransformation3D, mirtk::BSplineFreeFormTransformation4D, mirtk::LinearFreeFormTransformation4D, and mirtk::BSplineFreeFormTransformationStatistical.

ApproximateGradient() [1/3]

virtual void mirtk::Transformation::ApproximateGradient ( const ImageAttributes &  , const double *  , const double *  , const double *  , double *  , double  = 1.0  ) const

virtual

Finds gradient of approximation error: This function takes a set of points and a set of errors and finds a gradient to minimize the L2 norm of the error.

ApproximateGradient() [2/3]

virtual void mirtk::Transformation::ApproximateGradient ( const double *  , const double *  , const double *  , const double *  , const double *  , const double *  , int  , double *  , double  = 1.0  ) const

virtual

Finds gradient of approximation error: This function takes a set of points and a set of errors and finds a gradient to minimize the L2 norm of the error.

ApproximateGradient() [3/3]

virtual void mirtk::Transformation::ApproximateGradient ( const double *  , const double *  , const double *  , const double *  , const double *  , const double *  , const double *  , int  , double *  , double  = 1.0  ) const

virtual

Finds gradient of approximation error: This function takes a set of points and a set of errors and finds a gradient to minimize the L2 norm of the error.

CanModifyDisplacement()

virtual bool mirtk::Transformation::CanModifyDisplacement ( int  = -1 ) const

virtual

Whether this transformation implements a more efficient update of a given displacement field given the desired change of a transformation parameter

Reimplemented in mirtk::BSplineFreeFormTransformation3D, and mirtk::MultiLevelFreeFormTransformation.

CopyFrom()

virtual bool mirtk::Transformation::CopyFrom ( const Transformation *  )

virtual

Copy active transformation parameters (DoFs) from given transformation if possible and return false, otherwise

Reimplemented in mirtk::FreeFormTransformation, mirtk::FluidFreeFormTransformation, mirtk::MultiLevelTransformation, mirtk::HomogeneousTransformation, mirtk::AffineTransformation, and mirtk::SimilarityTransformation.

Displacement() [1/3]

virtual void mirtk::Transformation::Displacement ( const ImageAttributes &  , double *  , double *  , double *    ) const

virtual

Calculates the displacement at specified lattice points

Attention

The displacements are computed at the positions after applying the current displacements at each point. These displacements are then added to the current displacements. Therefore, set the input displacements to zero if only interested in the displacements of this transformation at the lattice points.

Displacement() [2/3]

virtual void mirtk::Transformation::Displacement ( GenericImage< double > &  , double  , double  , const WorldCoordsImage *  = NULL  ) const

virtual

Calculates the displacement vectors for a whole image domain

Attention

The displacements are computed at the positions after applying the current displacements at each voxel. These displacements are then added to the current displacements. Therefore, set the input displacements to zero if only interested in the displacements of this transformation at the voxel positions.

Reimplemented in mirtk::MultiLevelTransformation, mirtk::BSplineFreeFormTransformationSV, and mirtk::PartialBSplineFreeFormTransformationSV.

Displacement() [3/3]

virtual void mirtk::Transformation::Displacement ( GenericImage< float > &  , double  , double  , const WorldCoordsImage *  = NULL  ) const

virtual

Calculates the displacement vectors for a whole image domain

Attention

The displacements are computed at the positions after applying the current displacements at each voxel. These displacements are then added to the current displacements. Therefore, set the input displacements to zero if only interested in the displacements of this transformation at the voxel positions.

Reimplemented in mirtk::MultiLevelTransformation, mirtk::BSplineFreeFormTransformationSV, and mirtk::PartialBSplineFreeFormTransformationSV.

DisplacementAfterDOFChange()

virtual void mirtk::Transformation::DisplacementAfterDOFChange ( int  dof, double  dv, GenericImage< double > &  dx, double  t, double  t0 = NaN, const WorldCoordsImage *  i2w = NULL  ) const

virtual

Updates the displacement vectors for a whole image domain

Parameters

[in]	dof	Transformation parameter.
[in]	dv	Change of transformation parameter value.
[in,out]	dx	Displacement field to be updated.
[in]	t	Time point of start point.
[in]	t0	Time point of end point.
[in]	i2w	Pre-computed world coordinates.

Reimplemented in mirtk::BSplineFreeFormTransformation3D, and mirtk::MultiLevelFreeFormTransformation.

DOFBoundingBox()

bool mirtk::Transformation::DOFBoundingBox ( const Image *  image, int  , int &  i1, int &  j1, int &  k1, int &  i2, int &  j2, int &  k2, double  fraction = 1  ) const

inlinevirtual

Gets the spatial bounding box for a transformation parameter in image coordinates. The last parameter specifies what fraction of the bounding box to return. The default is 1 which equals 100% of the bounding box.

Reimplemented in mirtk::FreeFormTransformation, mirtk::FluidFreeFormTransformation, and mirtk::MultiLevelFreeFormTransformation.

Definition at line 780 of file Transformation.h.

EvaluateRMSError() [1/2]

double mirtk::Transformation::EvaluateRMSError	(	const ImageAttributes &	,
		double *	,
		double *
	)		const

Evaluates RMS error of transformation compared to given displacement field

This overloaded version of EvaluateRMSError is recommended for displacement fields defined on a regular lattice. It does not require memory for the explicit storage of the locations of each displacement vector. Moreover, if this transformation requires the caching of the displacements, the other overloads are either very slow or cannot be used.

EvaluateRMSError() [2/2]

double mirtk::Transformation::EvaluateRMSError	(	const ImageAttributes &	,
		double *	,
		double *	,
		double *
	)		const

Evaluates RMS error of transformation compared to given displacement field

This overloaded version of EvaluateRMSError is recommended for displacement fields defined on a regular lattice. It does not require memory for the explicit storage of the locations of each displacement vector. Moreover, if this transformation requires the caching of the displacements, the other overloads are either very slow or cannot be used.

Inverse()

int mirtk::Transformation::Inverse ( PointSet &  pset, double  t = 0, double  t0 = NaN  ) const

inlinevirtual

Transforms a set of points using the inverse of the transformation

Returns

Number of points at which transformation is non-invertible.

Definition at line 817 of file Transformation.h.

InverseDisplacement() [1/5]

virtual int mirtk::Transformation::InverseDisplacement ( const ImageAttributes &  , double *  , double *  , double *    ) const

virtual

Calculates the inverse displacement at specified lattice points

Attention

The displacements are computed at the positions after applying the current displacements at each point. These displacements are then added to the current displacements. Therefore, set the input displacements to zero if only interested in the displacements of this transformation at the lattice points.

Returns

Number of points at which transformation is non-invertible.

InverseDisplacement() [2/5]

virtual int mirtk::Transformation::InverseDisplacement ( GenericImage< double > &  , double  = NaN, const WorldCoordsImage *  = NULL  ) const

virtual

Calculates the inverse displacement vectors for a whole image domain

Returns

Number of points at which transformation is non-invertible.

InverseDisplacement() [3/5]

virtual int mirtk::Transformation::InverseDisplacement ( GenericImage< float > &  , double  = NaN, const WorldCoordsImage *  = NULL  ) const

virtual

Calculates the inverse displacement vectors for a whole image domain

Returns

Number of points at which transformation is non-invertible.

InverseDisplacement() [4/5]

virtual int mirtk::Transformation::InverseDisplacement ( GenericImage< double > &  , double  , double  , const WorldCoordsImage *  = NULL  ) const

virtual

Calculates the inverse displacement vectors for a whole image domain

Attention

The displacements are computed at the positions after applying the current displacements at each voxel. These displacements are then added to the current displacements. Therefore, set the input displacements to zero if only interested in the displacements of this transformation at the voxel positions.

Returns

Number of points at which transformation is non-invertible.

Reimplemented in mirtk::MultiLevelTransformation, mirtk::BSplineFreeFormTransformationSV, and mirtk::PartialBSplineFreeFormTransformationSV.

InverseDisplacement() [5/5]

virtual int mirtk::Transformation::InverseDisplacement ( GenericImage< float > &  , double  , double  , const WorldCoordsImage *  = NULL  ) const

virtual

Calculates the inverse displacement vectors for a whole image domain

Attention

The displacements are computed at the positions after applying the current displacements at each voxel. These displacements are then added to the current displacements. Therefore, set the input displacements to zero if only interested in the displacements of this transformation at the voxel positions.

Returns

Number of points at which transformation is non-invertible.

Reimplemented in mirtk::MultiLevelTransformation, mirtk::BSplineFreeFormTransformationSV, and mirtk::PartialBSplineFreeFormTransformationSV.

New() [1/3]

static Transformation* mirtk::Transformation::New ( TransformationType  )

static

Static constructor. This function returns a pointer to a concrete new transformation of the specified type (e.g., TRANSFORMATION_RIGID).

New() [2/3]

static Transformation* mirtk::Transformation::New ( const Transformation *  )

static

Static constructor. This function returns a pointer to a concrete transformation by copying the transformation passed to it.

New() [3/3]

static Transformation* mirtk::Transformation::New ( const char *  )

static

Static constructor. This function returns a pointer to a concrete transformation by reading the transformation parameters from a file and creating the appropriate transformation.

ParametricGradient() [1/7]

virtual void mirtk::Transformation::ParametricGradient ( const GenericImage< double > *  , double *  , const WorldCoordsImage *  , const WorldCoordsImage *  , double  = NaN, double  = 1  ) const

virtual

Applies the chain rule to convert spatial non-parametric gradient to a gradient w.r.t the parameters of this transformation.

If the transformation itself is non-parametric, the gradient will be passed through unchanged. The default implementation uses the full Jacobian matrix computed for each DoF separately (i.e., calls JacobianDOFs for each DoF).

For 4D transformations, the temporal coordinate t used for the computation of the Jacobian of the transformation w.r.t the transformation parameters for all spatial (x, y, z) voxel coordinates in world units, is assumed to correspond to the temporal origin of the given gradient image. For 4D transformations parameterized by velocities, a second time for the upper integration bound can be provided as last argument to this method. This last argument is ignored by transformations parameterized by displacements.

See also

ImageSimilarityMetric::EvaluateGradient

Reimplemented in mirtk::FreeFormTransformation, mirtk::BSplineFreeFormTransformationSV, mirtk::BSplineFreeFormTransformation3D, mirtk::FluidFreeFormTransformation, mirtk::MultiLevelStationaryVelocityTransformation, mirtk::MultiLevelFreeFormTransformation, mirtk::BSplineFreeFormTransformation4D, mirtk::PartialMultiLevelStationaryVelocityTransformation, mirtk::PartialBSplineFreeFormTransformationSV, mirtk::BSplineFreeFormTransformationTD, mirtk::BSplineFreeFormTransformationStatistical, mirtk::PartialAffineTransformation, mirtk::FreeFormTransformation3D, and mirtk::FreeFormTransformation4D.

ParametricGradient() [2/7]

void mirtk::Transformation::ParametricGradient ( const GenericImage< double > *  in, double *  out, const WorldCoordsImage *  i2w, double  t0 = NaN, double  w = 1  ) const

inline

Applies the chain rule to convert spatial non-parametric gradient to a gradient w.r.t the parameters of this transformation.

Definition at line 1000 of file Transformation.h.

ParametricGradient() [3/7]

void mirtk::Transformation::ParametricGradient ( const GenericImage< double > *  in, double *  out, double  t0 = NaN, double  w = 1  ) const

inline

Applies the chain rule to convert spatial non-parametric gradient to a gradient w.r.t the parameters of this transformation.

Definition at line 1008 of file Transformation.h.

ParametricGradient() [4/7]

void mirtk::Transformation::ParametricGradient ( const GenericImage< double > **  in, int  n, double *  out, const WorldCoordsImage *  i2w, const WorldCoordsImage *  wc, const double *  t0 = NULL, double  w = 1  ) const

inlinevirtual

Applies the chain rule to convert spatial non-parametric gradient to a gradient w.r.t the parameters of this transformation.

Reimplemented in mirtk::PartialMultiLevelStationaryVelocityTransformation, and mirtk::PartialBSplineFreeFormTransformationSV.

Definition at line 1015 of file Transformation.h.

ParametricGradient() [5/7]

void mirtk::Transformation::ParametricGradient ( const GenericImage< double > **  in, int  n, double *  out, const WorldCoordsImage *  i2w, const double *  t0 = NULL, double  w = 1  ) const

inline

Applies the chain rule to convert spatial non-parametric gradient to a gradient w.r.t the parameters of this transformation.

Definition at line 1026 of file Transformation.h.

ParametricGradient() [6/7]

void mirtk::Transformation::ParametricGradient ( const GenericImage< double > **  in, int  n, double *  out, const double *  t0 = NULL, double  w = 1  ) const

inline

Applies the chain rule to convert spatial non-parametric gradient to a gradient w.r.t the parameters of this transformation.

Definition at line 1034 of file Transformation.h.

ParametricGradient() [7/7]

virtual void mirtk::Transformation::ParametricGradient ( const PointSet &  , const Vector3D< double > *  , double *  , double  = 0, double  = NaN, double  = 1  ) const

virtual

Applies the chain rule to convert point-wise non-parametric gradient to a gradient w.r.t the parameters of this transformation.

Reimplemented in mirtk::FreeFormTransformation, mirtk::BSplineFreeFormTransformationSV, mirtk::FluidFreeFormTransformation, mirtk::MultiLevelFreeFormTransformation, mirtk::BSplineFreeFormTransformationTD, and mirtk::FreeFormTransformation3D.

RequiresCachingOfDisplacements()

bool mirtk::Transformation::RequiresCachingOfDisplacements ( ) const

inlinevirtual

Whether the caching of the transformation displacements is required (or preferred) by this transformation. For some transformations such as those parameterized by velocities, caching of the displacements for each target voxel results in better performance or is needed for example for the scaling and squaring method.

Reimplemented in mirtk::BSplineFreeFormTransformationSV, mirtk::MultiLevelTransformation, mirtk::PartialMultiLevelStationaryVelocityTransformation, mirtk::MultiLevelStationaryVelocityTransformation, and mirtk::PartialBSplineFreeFormTransformationSV.

Definition at line 793 of file Transformation.h.

Verify()

virtual void mirtk::Transformation::Verify ( )

virtual

Verifies that the transformation is well constructed according to class-specific rules

Reimplemented in mirtk::BSplineFreeFormTransformationStatistical.

---

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

• Transformation.h