Computes similarity between a moving spatial obejct and an Image to be registered. More...

#include <itkImageToSpatialObjectMetric.h>

Inheritance diagram for itk::ImageToSpatialObjectMetric< TFixedImage, TMovingSpatialObject >:

Collaboration diagram for itk::ImageToSpatialObjectMetric< TFixedImage, TMovingSpatialObject >:

Public Types
typedef SmartPointer< const Self >	ConstPointer
typedef Superclass::ParametersValueType	CoordinateRepresentationType
typedef Superclass::DerivativeType	DerivativeType
typedef FixedImageType::ConstPointer	FixedImageConstPointer
typedef FixedImageType::Pointer	FixedImagePointer
typedef TFixedImage	FixedImageType
typedef TransformType::InputPointType	InputPointType
typedef InterpolatorType::Pointer	InterpolatorPointer
typedef LinearInterpolateImageFunction < TFixedImage, CoordinateRepresentationType >	InterpolatorType
typedef Superclass::MeasureType	MeasureType
typedef MovingSpatialObjectType::ConstPointer	MovingSpatialObjectConstPointer
typedef MovingSpatialObjectType::Pointer	MovingSpatialObjectPointer
typedef TMovingSpatialObject	MovingSpatialObjectType
typedef TransformType::OutputPointType	OutputPointType
typedef Superclass::ParametersType	ParametersType
typedef SmartPointer< Self >	Pointer
typedef ImageToSpatialObjectMetric	Self
typedef SingleValuedCostFunction	Superclass
typedef TransformType::JacobianType	TransformJacobianType
typedef TransformType::ParametersType	TransformParametersType
typedef TransformType::Pointer	TransformPointer
typedef Transform < CoordinateRepresentationType, itkGetStaticConstMacro(ObjectDimension), itkGetStaticConstMacro(ImageDimension) >	TransformType
typedef vnl_vector_fixed < double, itkGetStaticConstMacro(ObjectDimension) >	VectorType
Public Member Functions
virtual const FixedImageType *	GetFixedImage ()
virtual InterpolatorType *	GetInterpolator ()
virtual const ParametersType &	GetLastTransformParameters ()
virtual const MovingSpatialObjectType *	GetMovingSpatialObject ()
virtual const char *	GetNameOfClass () const
virtual void	GetValueAndDerivative (const ParametersType &parameters, MeasureType &Value, DerivativeType &Derivative) const =0
virtual void	Initialize (void) throw ( ExceptionObject )
virtual void	SetFixedImage (const FixedImageType *_arg)
virtual void	SetInterpolator (InterpolatorType *_arg)
virtual void	SetMovingSpatialObject (const MovingSpatialObjectType *_arg)
virtual void	SetTransform (TransformType *_arg)
Static Public Attributes
static const unsigned int	ImageDimension = FixedImageType::ImageDimension
static const unsigned int	ObjectDimension = MovingSpatialObjectType::ObjectDimension
Protected Member Functions
	ImageToSpatialObjectMetric ()
	ImageToSpatialObjectMetric (const Self &)
void	operator= (const Self &)
void	PrintSelf (std::ostream &os, Indent indent) const
virtual	~ImageToSpatialObjectMetric ()
Protected Attributes
FixedImageConstPointer	m_FixedImage
InterpolatorPointer	m_Interpolator
ParametersType	m_LastTransformParameters
MeasureType	m_MatchMeasure
DerivativeType	m_MatchMeasureDerivatives
MovingSpatialObjectConstPointer	m_MovingSpatialObject
TransformPointer	m_Transform

Detailed Description

template<class TFixedImage, class TMovingSpatialObject>
class itk::ImageToSpatialObjectMetric< TFixedImage, TMovingSpatialObject >

Computes similarity between a moving spatial obejct and an Image to be registered.

The ImageToSpatialObjectMetric is different from the rest of the registration framework in ITK regarding the interpretation of the Transform with respect to the Fixed and Moving objects. In most of the ITK registration framework, the Transform computed by the optimizer is the one that maps points from the space of the Fixed object into the space of the Moving object. This direction of the transform is the one that makes easier to resample the Moving object into the space of the Fixed object.

In the particular case of the ImageToSpatialObject registration, the Transform to be computed is the one mapping points from the SpatialObject into the Image, despite the fact that the SpatialObject is called the "Moving" object and the image is called the "Fixed" object. This change of reference system is the consequence of using this type of registration in applications that are based on Visualization. In the context of such visualizations it is simpler to think in terms of the Transform that can be used for displaying the SpatialObject in the appropriate position with respect to the image. Since this process does not involve resampling, but providing a Transform to a visualization routine, it is usually more natural to use the Transform that maps points from the SpatialObject space the image space.

A full discussion of the Transform directions in the ITK registration framework can be found in the ITK Software Guide.

Definition at line 61 of file itkImageToSpatialObjectMetric.h.