This function is used in GeodesicActiveContourLevelSetImageFilter to segment structures in an image based on a user supplied edge potential map. More...

#include <itkGeodesicActiveContourLevelSetFunction.h>

Inheritance diagram for itk::GeodesicActiveContourLevelSetFunction< TImageType, TFeatureImageType >:

Collaboration diagram for itk::GeodesicActiveContourLevelSetFunction< TImageType, TFeatureImageType >:

List of all members.

Public Types
typedef SmartPointer< const Self >	ConstPointer
typedef TFeatureImageType	FeatureImageType
typedef Superclass::FeatureScalarType	FeatureScalarType
typedef Superclass::FloatOffsetType	FloatOffsetType
typedef Superclass::GlobalDataStruct	GlobalDataStruct
typedef Superclass::ImageType	ImageType
typedef Superclass::NeighborhoodType	NeighborhoodType
typedef SmartPointer< Self >	Pointer
typedef Superclass::RadiusType	RadiusType
typedef Superclass::ScalarValueType	ScalarValueType
typedef GeodesicActiveContourLevelSetFunction	Self
typedef SegmentationLevelSetFunction < TImageType, TFeatureImageType >	Superclass
typedef Superclass::VectorImageType	VectorImageType
Public Member Functions
virtual void	CalculateAdvectionImage ()
virtual void	CalculateSpeedImage ()
virtual ::itk::LightObject::Pointer	CreateAnother (void) const
virtual ScalarValueType	CurvatureSpeed (const NeighborhoodType &neighborhood, const FloatOffsetType &offset, GlobalDataStruct *gd) const
virtual const char *	GetNameOfClass () const
virtual void	Initialize (const RadiusType &r)

void	SetDerivativeSigma (const double v)
double	GetDerivativeSigma ()
Static Public Member Functions
static Pointer	New ()
Static Public Attributes
static const unsigned int	ImageDimension = Superclass::ImageDimension
Protected Member Functions
	GeodesicActiveContourLevelSetFunction ()
	GeodesicActiveContourLevelSetFunction (const Self &)
void	operator= (const Self &)
void	PrintSelf (std::ostream &os, Indent indent) const
virtual	~GeodesicActiveContourLevelSetFunction ()
Private Attributes
double	m_DerivativeSigma

Detailed Description

template<class TImageType, class TFeatureImageType = TImageType>
class itk::GeodesicActiveContourLevelSetFunction< TImageType, TFeatureImageType >

This function is used in GeodesicActiveContourLevelSetImageFilter to segment structures in an image based on a user supplied edge potential map.

IMPORTANT: The LevelSetFunction class contain additional information necessary to gain full understanding of how to use this function.

GeodesicActiveContourLevelSetFunction is a subclass of the generic LevelSetFunction. It is used to segment structures in an image based on a user supplied edge potential map $ g(I) $ , which has values close to zero in regions near edges (or high image gradient) and values close to one in regions with relatively constant intensity. Typically, the edge potential map is a function of the gradient, for example:

$g(I) = 1 / ( 1 + | (\nabla * G)(I)| )$

$g(I) = \exp^{-|(\nabla * G)(I)|}$

where $ I $ is image intensity and $(\nabla * G)$ is the derivative of Gaussian operator.

The edge potential image is set via the SetFeatureImage() method.

In this function both the propagation term $P(\mathbf{x})$ and the curvature spatial modifier term $Z(\mathbf{x})$ are taken directly from the edge potential image such that:

$P(\mathbf{x}) = g(\mathbf{x})$

$Z(\mathbf{x}) = g(\mathbf{x})$

An advection term $\mathbf{A}(\mathbf{x})$ is constructed from the negative gradient of the edge potential image.

$\mathbf{A}(\mathbf{x}) = -\nabla g(\mathbf{x})$

This term behaves like a doublet attracting the contour to the edges.

This implementation is based on: "Geodesic Active Contours", V. Caselles, R. Kimmel and G. Sapiro. International Journal on Computer Vision, Vol 22, No. 1, pp 61-97, 1997

See also:: LevelSetFunction; SegmentationLevelSetImageFunction; GeodesicActiveContourLevelSetImageFilter

Definition at line 77 of file itkGeodesicActiveContourLevelSetFunction.h.