This class implements the filter for computing the normal vectors from a scalar implicit function (i.e. a levelset image) and processing them. More...

#include <itkImplicitManifoldNormalVectorFilter.h>

Inheritance diagram for itk::ImplicitManifoldNormalVectorFilter< TInputImage, TSparseOutputImage >:

Collaboration diagram for itk::ImplicitManifoldNormalVectorFilter< TInputImage, TSparseOutputImage >:

List of all members.

Public Types
typedef SmartPointer< const Self >	ConstPointer
typedef Superclass::FiniteDifferenceFunctionType	FiniteDifferenceFunctionType
typedef Superclass::IndexType	IndexType
typedef ConstNeighborhoodIterator < InputImageType >	InputImageIteratorType
typedef Superclass::InputImageType	InputImageType
typedef Superclass::NodeListType	NodeListType
typedef Superclass::NodeValueType	NodeValueType
typedef Superclass::OutputNodeType	NormalBandNodeType
typedef NormalVectorFunctionBase < SparseOutputImageType >	NormalFunctionType
typedef Superclass::NodeDataType	NormalVectorType
typedef SmartPointer< Self >	Pointer
typedef FiniteDifferenceFunctionType::RadiusType	RadiusType
typedef ImplicitManifoldNormalVectorFilter	Self
typedef Superclass::SparseOutputImageType	SparseOutputImageType
typedef FiniteDifferenceSparseImageFilter < TInputImage, TSparseOutputImage >	Superclass
Public Member Functions
virtual ::itk::LightObject::Pointer	CreateAnother (void) const
virtual NodeValueType	GetIsoLevelHigh () const
virtual NodeValueType	GetIsoLevelLow () const
virtual unsigned int	GetMaxIteration () const
virtual NodeValueType	GetMinVectorNorm () const
virtual const char *	GetNameOfClass () const
virtual bool	GetUnsharpMaskingFlag () const
virtual NodeValueType	GetUnsharpMaskingWeight () const
virtual void	SetIsoLevelHigh (NodeValueType _arg)
virtual void	SetIsoLevelLow (NodeValueType _arg)
virtual void	SetMaxIteration (unsigned int _arg)
virtual void	SetMinVectorNorm (NodeValueType _arg)
void	SetNormalFunction (NormalFunctionType *nf)
virtual void	SetUnsharpMaskingFlag (bool _arg)
virtual void	SetUnsharpMaskingWeight (NodeValueType _arg)
Static Public Member Functions
static Pointer	New ()
Static Public Attributes
static const unsigned int	ImageDimension = Superclass::ImageDimension
Protected Member Functions
virtual void	CopyInputToOutput ()
virtual NormalVectorType	DataConstraint (const NormalVectorType &data) const
	ImplicitManifoldNormalVectorFilter ()
virtual void	Initialize ()
void	InitializeNormalBandNode (NormalBandNodeType *node, const InputImageIteratorType &it)
virtual void	PostProcessOutput ()
void	PrintSelf (std::ostream &os, Indent indent) const
void	SetNormalBand ()
	~ImplicitManifoldNormalVectorFilter ()

virtual bool	Halt ()
Private Member Functions
	ImplicitManifoldNormalVectorFilter (const Self &)
void	operator= (const Self &)
Private Attributes
NodeValueType	m_DimConst
NodeValueType	m_DimConst2
unsigned long	m_Indicator [itkGetStaticConstMacro(ImageDimension)]
NodeValueType	m_IsoLevelHigh
NodeValueType	m_IsoLevelLow
RadiusType	m_ManifoldRadius
unsigned int	m_MaxIteration
NodeValueType	m_MinVectorNorm
NormalFunctionType *	m_NormalFunction
unsigned int	m_NumVertex
bool	m_UnsharpMaskingFlag
NodeValueType	m_UnsharpMaskingWeight

Detailed Description

template<class TInputImage, class TSparseOutputImage>
class itk::ImplicitManifoldNormalVectorFilter< TInputImage, TSparseOutputImage >

This class implements the filter for computing the normal vectors from a scalar implicit function (i.e. a levelset image) and processing them.

: This is a ready-to-use class for filtering normal vectors of an implicit manifold image. The normal vectors of the input image are computed and processed in a band where the values of the input image fall in the region [IsoLevelLow, IsoLevelHigh]. The processing is done by m_NormalFunction. This class also defines a Halt method which uses the iteration counter and the parameter m_MaxIteration to determine when to stop the processing.

INPUTS: The input is a scalar image. Even though this can be any scalar image, this filter class is intended to work with the image of an implicit function. One such example is the output of the SparseFieldLevelSetImageFilter class.

OUTPUTS: The output is a sparse image. The m_Data member variables of the nodes of the sparse image will contain the filtered output normal vectors. The sparse image has valid node pointers only in the band [IsoLevelLow, IsoLevelHigh]. Pixels of the sparse image outside this band will be null pointers.

PARAMETERS: IsoLevelLow and IsoLevelHigh define the working band for this filter. Pixels of the input scalar image whose values fall between these low and high limits will be operated on. MaxIteration determines the number of iterations this filter will perform. (Default is 25) The MinVectorNorm parameter determines the minumum vector norm allowed (to avoid divide by 0). The default for MinVectorNorm is 10^-6 which is designed to work with an input image of floats. This value can be lowered to 10^-12 if the input image is doubles.

IMPORTANT: The TSparseOutputImage template parameter must be a sparse image templated over a NodeType which at least has the following members: m_Data, m_InputData, m_Update and m_MAnifoldNormal. Depending on the Function object being used it might need other members. For instance, NormalVectorDiffusionFunction will also require that the NodeType has the following additional members: m_Flux.

Definition at line 73 of file itkImplicitManifoldNormalVectorFilter.h.