#include <itkMultiScaleHessianBasedMeasureImageFilter.h>

Inheritance diagram for itk::MultiScaleHessianBasedMeasureImageFilter< TInputImage, THessianImage, TOutputImage >:

Collaboration diagram for itk::MultiScaleHessianBasedMeasureImageFilter< TInputImage, THessianImage, TOutputImage >:

Public Types
typedef UpdateBufferType::ValueType	BufferValueType
typedef SmartPointer< const Self >	ConstPointer
typedef Superclass::DataObjectPointer	DataObjectPointer
typedef ProcessObject::DataObjectPointerArraySizeType	DataObjectPointerArraySizeType
typedef HessianRecursiveGaussianImageFilter < InputImageType, HessianImageType >	HessianFilterType
typedef THessianImage	HessianImageType
typedef ImageToImageFilter < HessianImageType, OutputImageType >	HessianToMeasureFilterType
typedef TInputImage	InputImageType
typedef TInputImage::PixelType	InputPixelType
typedef TOutputImage	OutputImageType
typedef TOutputImage::PixelType	OutputPixelType
typedef TOutputImage::RegionType	OutputRegionType
typedef SmartPointer< Self >	Pointer
typedef Image< ScalesPixelType, itkGetStaticConstMacro(ImageDimension) >	ScalesImageType
typedef float	ScalesPixelType
typedef MultiScaleHessianBasedMeasureImageFilter	Self
enum	SigmaStepMethodType { EquispacedSigmaSteps = 0, LogarithmicSigmaSteps = 1 }
typedef ImageToImageFilter < TInputImage, TOutputImage >	Superclass
typedef Image< double, itkGetStaticConstMacro(ImageDimension) >	UpdateBufferType
Public Member Functions
virtual ::itk::LightObject::Pointer	CreateAnother (void) const
void	EnlargeOutputRequestedRegion (DataObject *)
const HessianImageType *	GetHessianOutput () const
virtual const char *	GetNameOfClass () const
const ScalesImageType *	GetScalesOutput () const
virtual DataObjectPointer	MakeOutput (DataObjectPointerArraySizeType idx)
void	SetSigmaStepMethodToEquispaced ()
void	SetSigmaStepMethodToLogarithmic ()

virtual void	SetSigmaMinimum (double _arg)
virtual double	GetSigmaMinimum () const

virtual void	SetSigmaMaximum (double _arg)
virtual double	GetSigmaMaximum () const

virtual void	SetNumberOfSigmaSteps (unsigned int _arg)
virtual unsigned int	GetNumberOfSigmaSteps () const

virtual void	SetHessianToMeasureFilter (HessianToMeasureFilterType *_arg)
virtual HessianToMeasureFilterType *	GetHessianToMeasureFilter ()

virtual void	SetNonNegativeHessianBasedMeasure (bool _arg)
virtual bool	GetNonNegativeHessianBasedMeasure () const
virtual void	NonNegativeHessianBasedMeasureOn ()
virtual void	NonNegativeHessianBasedMeasureOff ()

virtual void	SetSigmaStepMethod (SigmaStepMethodType _arg)
virtual SigmaStepMethodType	GetSigmaStepMethod () const

virtual void	SetGenerateScalesOutput (bool _arg)
virtual bool	GetGenerateScalesOutput () const
virtual void	GenerateScalesOutputOn ()
virtual void	GenerateScalesOutputOff ()

virtual void	SetGenerateHessianOutput (bool _arg)
virtual bool	GetGenerateHessianOutput () const
virtual void	GenerateHessianOutputOn ()
virtual void	GenerateHessianOutputOff ()
Static Public Member Functions
static Pointer	New ()
Static Public Attributes
static const unsigned int	ImageDimension = ::itk::GetImageDimension< InputImageType >::ImageDimension
Protected Member Functions
void	GenerateData (void)
	MultiScaleHessianBasedMeasureImageFilter ()
void	PrintSelf (std::ostream &os, Indent indent) const
	~MultiScaleHessianBasedMeasureImageFilter ()
Private Member Functions
void	AllocateUpdateBuffer ()
double	ComputeSigmaValue (int scaleLevel)
	MultiScaleHessianBasedMeasureImageFilter (const Self &)
void	operator= (const Self &)
void	UpdateMaximumResponse (double sigma)
Private Attributes
bool	m_GenerateHessianOutput
bool	m_GenerateScalesOutput
HessianFilterType::Pointer	m_HessianFilter
HessianToMeasureFilterType::Pointer	m_HessianToMeasureFilter
bool	m_NonNegativeHessianBasedMeasure
unsigned int	m_NumberOfSigmaSteps
double	m_SigmaMaximum
double	m_SigmaMinimum
SigmaStepMethodType	m_SigmaStepMethod
UpdateBufferType::Pointer	m_UpdateBuffer

Detailed Description

template<typename TInputImage, typename THessianImage, typename TOutputImage = TInputImage>
class itk::MultiScaleHessianBasedMeasureImageFilter< TInputImage, THessianImage, TOutputImage >

A filter to enhance structures using Hessian eigensystem-based measures in a multiscale framework.

The filter evaluates a Hessian-based enhancement measure, such as vesselness or objectness, at different scale levels. The Hessian-based measure is computed from the Hessian image at each scale level and the best response is selected.

Minimum and maximum sigma value can be set using SetMinSigma and SetMaxSigma methods respectively. The number of scale levels is set using SetNumberOfSigmaSteps method. Exponentially distributed scale levels are computed within the bound set by the minimum and maximum sigma values

The filter computes a second output image (accessed by the GetScalesOutput method) containing the scales at which each pixel gave the best reponse.

This code was contributed in the Insight Journal paper: "Generalizing vesselness with respect to dimensionality and shape" by Antiga L. http://hdl.handle.net/1926/576 http://www.insight-journal.org/browse/publication/175

Author:: Luca Antiga Ph.D. Medical Imaging Unit, Bioengineering Deparment, Mario Negri Institute, Italy.

See also:: HessianToObjectnessMeasureImageFilter; Hessian3DToVesselnessMeasureImageFilter; HessianSmoothed3DToVesselnessMeasureImageFilter; HessianRecursiveGaussianImageFilter; SymmetricEigenAnalysisImageFilter; SymmetricSecondRankTensor

Definition at line 67 of file itkMultiScaleHessianBasedMeasureImageFilter.h.