#include <itkCenteredTransformInitializer.h>

Inheritance diagram for itk::CenteredTransformInitializer< TTransform, TFixedImage, TMovingImage >:

Collaboration diagram for itk::CenteredTransformInitializer< TTransform, TFixedImage, TMovingImage >:

Public Types
typedef SmartPointer< const Self >	ConstPointer
typedef FixedImageCalculatorType::Pointer	FixedImageCalculatorPointer
typedef ImageMomentsCalculator < FixedImageType >	FixedImageCalculatorType
typedef FixedImageType::ConstPointer	FixedImagePointer
typedef TFixedImage	FixedImageType
typedef TransformType::InputPointType	InputPointType
typedef MovingImageCalculatorType::Pointer	MovingImageCalculatorPointer
typedef ImageMomentsCalculator < MovingImageType >	MovingImageCalculatorType
typedef MovingImageType::ConstPointer	MovingImagePointer
typedef TMovingImage	MovingImageType
typedef TransformType::OffsetType	OffsetType
typedef TransformType::OutputVectorType	OutputVectorType
typedef SmartPointer< Self >	Pointer
typedef CenteredTransformInitializer	Self
typedef Object	Superclass
typedef TransformType::Pointer	TransformPointer
typedef TTransform	TransformType
Public Member Functions
virtual ::itk::LightObject::Pointer	CreateAnother (void) const
virtual const char *	GetNameOfClass () const
virtual void	InitializeTransform ()
virtual void	SetFixedImage (const FixedImageType *_arg)
virtual void	SetMovingImage (const MovingImageType *_arg)
virtual void	SetTransform (TransformType *_arg)

void	GeometryOn ()
void	MomentsOn ()
Static Public Member Functions
static Pointer	New ()
Static Public Attributes

static const unsigned int	InputSpaceDimension = TransformType::InputSpaceDimension
static const unsigned int	OutputSpaceDimension = TransformType::OutputSpaceDimension
Protected Member Functions
virtual TransformType *	GetTransform ()
void	PrintSelf (std::ostream &os, Indent indent) const
Private Member Functions
	CenteredTransformInitializer (const Self &)
void	operator= (const Self &)
Private Attributes
FixedImageCalculatorPointer	m_FixedCalculator
FixedImagePointer	m_FixedImage
MovingImageCalculatorPointer	m_MovingCalculator
MovingImagePointer	m_MovingImage
TransformPointer	m_Transform
bool	m_UseMoments
virtual const FixedImageCalculatorType *	GetFixedCalculator ()
virtual const MovingImageCalculatorType *	GetMovingCalculator ()
	CenteredTransformInitializer ()
	~CenteredTransformInitializer ()

Detailed Description

template<class TTransform, class TFixedImage, class TMovingImage>
class itk::CenteredTransformInitializer< TTransform, TFixedImage, TMovingImage >

CenteredTransformInitializer is a helper class intended to initialize the center of rotation and the translation of Transforms having the center of rotation among their parameters.

This class is connected to the fixed image, moving image and transform involved in the registration. Two modes of operation are possible:

Geometrical,
Center of mass

In the first mode, the geometrical center of the moving image is passed as initial center of rotation to the transform and the vector from the center of the fixed image to the center of the moving image is passed as the initial translation. This mode basically assumes that the anatomical objects to be registered are centered in their respective images. Hence the best initial guess for the registration is the one that superimposes those two centers.

In the second mode, the moments of gray level values are computed for both images. The center of mass of the moving image is then used as center of rotation. The vector between the two centers of mass is passes as the initial translation to the transform. This second approach assumes that the moments of the anatomical objects are similar for both images and hence the best initial guess for registration is to superimpose both mass centers. Note that this assumption will probably not hold in multi-modality registration.

Definition at line 63 of file itkCenteredTransformInitializer.h.