itk::ShapePriorMAPCostFunction< TFeatureImage, TOutputPixel > Class Template Reference
[Numerics, Optimizers]

Represents the maximum aprior (MAP) cost function used ShapePriorSegmentationLevelSetImageFilter to estimate the shape paramaeters. More...

#include <itkShapePriorMAPCostFunction.h>

Inheritance diagram for itk::ShapePriorMAPCostFunction< TFeatureImage, TOutputPixel >:

[legend]Collaboration diagram for itk::ShapePriorMAPCostFunction< TFeatureImage, TOutputPixel >:

[legend]List of all members.

[NOHEADER]
typedef FixedArray< double, 4 >	WeightsType
virtual void	SetWeights (WeightsType _arg)
virtual const WeightsType &	GetWeights ()
Public Types
typedef ShapePriorMAPCostFunction	Self
typedef ShapePriorMAPCostFunctionBase< TFeatureImage, TOutputPixel >	Superclass
typedef SmartPointer< Self >	Pointer
typedef SmartPointer< const Self >	ConstPointer
typedef Superclass::ParametersType	ParametersType
typedef Superclass::FeatureImageType	FeatureImageType
typedef Superclass::FeatureImagePointer	FeatureImagePointer
typedef Superclass::MeasureType	MeasureType
typedef Superclass::PixelType	PixelType
typedef Superclass::NodeType	NodeType
typedef Superclass::NodeContainerType	NodeContainerType
typedef Superclass::ShapeFunctionType	ShapeFunctionType
typedef Array< double >	ArrayType
Public Member Functions
virtual const char *	GetNameOfClass () const
	itkStaticConstMacro (ImageDimension, unsigned int, TFeatureImage::ImageDimension)
virtual MeasureType	ComputeLogInsideTerm (const ParametersType &parameters) const
virtual MeasureType	ComputeLogGradientTerm (const ParametersType &parameters) const
virtual MeasureType	ComputeLogShapePriorTerm (const ParametersType &parameters) const
virtual MeasureType	ComputeLogPosePriorTerm (const ParametersType &parameters) const
virtual void	Initialize (void) throw ( ExceptionObject )
virtual void	SetShapeParameterMeans (ArrayType _arg)
virtual ArrayType	GetShapeParameterMeans ()
virtual void	SetShapeParameterStandardDeviations (ArrayType _arg)
virtual ArrayType	GetShapeParameterStandardDeviations ()
Static Public Member Functions
Pointer	New ()
Protected Member Functions
	ShapePriorMAPCostFunction ()
virtual	~ShapePriorMAPCostFunction ()
void	PrintSelf (std::ostream &os, Indent indent) const

---

Detailed Description

template<class TFeatureImage, class TOutputPixel>
class itk::ShapePriorMAPCostFunction< TFeatureImage, TOutputPixel >

Represents the maximum aprior (MAP) cost function used ShapePriorSegmentationLevelSetImageFilter to estimate the shape paramaeters.

This class follows the shape and pose parameters estimation developed in [1]. Note that this class returns the negative log of the MAP function. Using the negative function make this cost function compatible with generic optimizers which seeks the minimum of a cost function.

This class has two template parameters, the feature image type representing the edge potential map and the pixel type used to represent the output level set in the ShapePriorSegmentationLevelSetImageFilter.

See also:

ShapePriorSegmentationLevelSetImageFilter

REFERENCES

[1] Leventon, M.E. et al. "Statistical Shape Influence in Geodesic Active Contours", CVPR 2000.

Definition at line 49 of file itkShapePriorMAPCostFunction.h.

---

Member Typedef Documentation

template<class TFeatureImage, class TOutputPixel> typedef Array<double> itk::ShapePriorMAPCostFunction< TFeatureImage, TOutputPixel >::ArrayType

Type of the array for storing shape parameter mean and standard deivation. Definition at line 92 of file itkShapePriorMAPCostFunction.h. Referenced by itk::ShapePriorMAPCostFunction< TFeatureImage, TOutputPixel >::~ShapePriorMAPCostFunction().

template<class TFeatureImage, class TOutputPixel> typedef SmartPointer<const Self> itk::ShapePriorMAPCostFunction< TFeatureImage, TOutputPixel >::ConstPointer

Reimplemented from itk::ShapePriorMAPCostFunctionBase< TFeatureImage, TOutputPixel >. Definition at line 57 of file itkShapePriorMAPCostFunction.h.

template<class TFeatureImage, class TOutputPixel> typedef Superclass::FeatureImagePointer itk::ShapePriorMAPCostFunction< TFeatureImage, TOutputPixel >::FeatureImagePointer

Reimplemented from itk::ShapePriorMAPCostFunctionBase< TFeatureImage, TOutputPixel >. Definition at line 71 of file itkShapePriorMAPCostFunction.h.

template<class TFeatureImage, class TOutputPixel> typedef Superclass::FeatureImageType itk::ShapePriorMAPCostFunction< TFeatureImage, TOutputPixel >::FeatureImageType

Type of the feature image representing the edge potential map. Reimplemented from itk::ShapePriorMAPCostFunctionBase< TFeatureImage, TOutputPixel >. Definition at line 70 of file itkShapePriorMAPCostFunction.h.

template<class TFeatureImage, class TOutputPixel> typedef Superclass::MeasureType itk::ShapePriorMAPCostFunction< TFeatureImage, TOutputPixel >::MeasureType

Type of the return measure value. Reimplemented from itk::ShapePriorMAPCostFunctionBase< TFeatureImage, TOutputPixel >. Definition at line 74 of file itkShapePriorMAPCostFunction.h.

template<class TFeatureImage, class TOutputPixel> typedef Superclass::NodeContainerType itk::ShapePriorMAPCostFunction< TFeatureImage, TOutputPixel >::NodeContainerType

Type of container used to store the level set nodes. Reimplemented from itk::ShapePriorMAPCostFunctionBase< TFeatureImage, TOutputPixel >. Definition at line 86 of file itkShapePriorMAPCostFunction.h.

template<class TFeatureImage, class TOutputPixel> typedef Superclass::NodeType itk::ShapePriorMAPCostFunction< TFeatureImage, TOutputPixel >::NodeType

Type of node used to represent the active region around the zero set. Reimplemented from itk::ShapePriorMAPCostFunctionBase< TFeatureImage, TOutputPixel >. Definition at line 83 of file itkShapePriorMAPCostFunction.h.

template<class TFeatureImage, class TOutputPixel> typedef Superclass::ParametersType itk::ShapePriorMAPCostFunction< TFeatureImage, TOutputPixel >::ParametersType

ParametersType typedef. It defines a position in the optimization search space. Reimplemented from itk::ShapePriorMAPCostFunctionBase< TFeatureImage, TOutputPixel >. Definition at line 67 of file itkShapePriorMAPCostFunction.h.

template<class TFeatureImage, class TOutputPixel> typedef Superclass::PixelType itk::ShapePriorMAPCostFunction< TFeatureImage, TOutputPixel >::PixelType

Type of pixel used to represent the level set. Reimplemented from itk::ShapePriorMAPCostFunctionBase< TFeatureImage, TOutputPixel >. Definition at line 80 of file itkShapePriorMAPCostFunction.h.

template<class TFeatureImage, class TOutputPixel> typedef SmartPointer<Self> itk::ShapePriorMAPCostFunction< TFeatureImage, TOutputPixel >::Pointer

Reimplemented from itk::ShapePriorMAPCostFunctionBase< TFeatureImage, TOutputPixel >. Definition at line 56 of file itkShapePriorMAPCostFunction.h.

template<class TFeatureImage, class TOutputPixel> typedef ShapePriorMAPCostFunction itk::ShapePriorMAPCostFunction< TFeatureImage, TOutputPixel >::Self

Standard class typedefs. Reimplemented from itk::ShapePriorMAPCostFunctionBase< TFeatureImage, TOutputPixel >. Definition at line 54 of file itkShapePriorMAPCostFunction.h. Referenced by itk::ShapePriorMAPCostFunction< TFeatureImage, TOutputPixel >::~ShapePriorMAPCostFunction().

template<class TFeatureImage, class TOutputPixel> typedef Superclass::ShapeFunctionType itk::ShapePriorMAPCostFunction< TFeatureImage, TOutputPixel >::ShapeFunctionType

Type of the shape signed distance function. Reimplemented from itk::ShapePriorMAPCostFunctionBase< TFeatureImage, TOutputPixel >. Definition at line 89 of file itkShapePriorMAPCostFunction.h.

template<class TFeatureImage, class TOutputPixel> typedef ShapePriorMAPCostFunctionBase<TFeatureImage,TOutputPixel> itk::ShapePriorMAPCostFunction< TFeatureImage, TOutputPixel >::Superclass

Reimplemented from itk::ShapePriorMAPCostFunctionBase< TFeatureImage, TOutputPixel >. Definition at line 55 of file itkShapePriorMAPCostFunction.h.

template<class TFeatureImage, class TOutputPixel> typedef FixedArray<double,4> itk::ShapePriorMAPCostFunction< TFeatureImage, TOutputPixel >::WeightsType

Set/Get the weights for each term. Default is a vector of all ones. The weights are applied to terms in the following order: LogInsideTerm, LogGradientTerm, LogShapePriorTerm and LogPosePriorTerm. Definition at line 107 of file itkShapePriorMAPCostFunction.h. Referenced by itk::ShapePriorMAPCostFunction< TFeatureImage, TOutputPixel >::~ShapePriorMAPCostFunction().

---

Constructor & Destructor Documentation

template<class TFeatureImage, class TOutputPixel> itk::ShapePriorMAPCostFunction< TFeatureImage, TOutputPixel >::ShapePriorMAPCostFunction (   )  [protected]

template<class TFeatureImage, class TOutputPixel> virtual itk::ShapePriorMAPCostFunction< TFeatureImage, TOutputPixel >::~ShapePriorMAPCostFunction (   )  [inline, protected, virtual]

Definition at line 143 of file itkShapePriorMAPCostFunction.h. References itk::ShapePriorMAPCostFunction< TFeatureImage, TOutputPixel >::ArrayType, itk::GaussianKernelFunction::Pointer, itk::ShapePriorMAPCostFunction< TFeatureImage, TOutputPixel >::Self, and itk::ShapePriorMAPCostFunction< TFeatureImage, TOutputPixel >::WeightsType.

---

Member Function Documentation

template<class TFeatureImage, class TOutputPixel> virtual MeasureType itk::ShapePriorMAPCostFunction< TFeatureImage, TOutputPixel >::ComputeLogGradientTerm (  const ParametersType &  parameters  )  const [virtual]

Compute the gradient term component of the MAP cost function. In particular, this method assume that ( 1 - FeatureImage ) approximates a Gaussian (zero mean, unit variance) algon the normal of the evolving contour. The gradient term is then given by a Laplacian of the goodness of fit of the Gaussian. Implements itk::ShapePriorMAPCostFunctionBase< TFeatureImage, TOutputPixel >.

template<class TFeatureImage, class TOutputPixel> virtual MeasureType itk::ShapePriorMAPCostFunction< TFeatureImage, TOutputPixel >::ComputeLogInsideTerm (  const ParametersType &  parameters  )  const [virtual]

Compute the inside term component of the MAP cost function. In particular, the method sums the number of pixels inside the current contour (defined by nodes of the active region that are less than zero) which are outside the shape specified by the input parameters. Implements itk::ShapePriorMAPCostFunctionBase< TFeatureImage, TOutputPixel >.

template<class TFeatureImage, class TOutputPixel> virtual MeasureType itk::ShapePriorMAPCostFunction< TFeatureImage, TOutputPixel >::ComputeLogPosePriorTerm (  const ParametersType &  parameters  )  const [virtual]

Compute the pose prior component of the MAP cost function. In particular, this method assumes that the pose parameters are uniformly distributed and returns a constant of zero. Implements itk::ShapePriorMAPCostFunctionBase< TFeatureImage, TOutputPixel >.

template<class TFeatureImage, class TOutputPixel> virtual MeasureType itk::ShapePriorMAPCostFunction< TFeatureImage, TOutputPixel >::ComputeLogShapePriorTerm (  const ParametersType &  parameters  )  const [virtual]

Compute the shape prior component of the MAP cost function. In particular, the method assumes that the shape parameters comes from independent Gaussian distributions defined by the ShapeParameterMeans and ShapeParameterVariances array. Implements itk::ShapePriorMAPCostFunctionBase< TFeatureImage, TOutputPixel >.

template<class TFeatureImage, class TOutputPixel> virtual const char* itk::ShapePriorMAPCostFunction< TFeatureImage, TOutputPixel >::GetNameOfClass (   )  const [virtual]

Run-time type information (and related methods). Reimplemented from itk::ShapePriorMAPCostFunctionBase< TFeatureImage, TOutputPixel >.

template<class TFeatureImage, class TOutputPixel> virtual ArrayType itk::ShapePriorMAPCostFunction< TFeatureImage, TOutputPixel >::GetShapeParameterMeans (   )  [virtual]

Set/Get the array of shape parameters mean.

template<class TFeatureImage, class TOutputPixel> virtual ArrayType itk::ShapePriorMAPCostFunction< TFeatureImage, TOutputPixel >::GetShapeParameterStandardDeviations (   )  [virtual]

Set/Get the array of shape parameters standard deviation.

template<class TFeatureImage, class TOutputPixel> virtual const WeightsType& itk::ShapePriorMAPCostFunction< TFeatureImage, TOutputPixel >::GetWeights (   )  [virtual]

Set/Get the weights for each term. Default is a vector of all ones. The weights are applied to terms in the following order: LogInsideTerm, LogGradientTerm, LogShapePriorTerm and LogPosePriorTerm.

template<class TFeatureImage, class TOutputPixel> virtual void itk::ShapePriorMAPCostFunction< TFeatureImage, TOutputPixel >::Initialize (  void   )  throw ( ExceptionObject ) [virtual]

Initialize the cost function by making sure that all the components are present. Reimplemented from itk::ShapePriorMAPCostFunctionBase< TFeatureImage, TOutputPixel >.

template<class TFeatureImage, class TOutputPixel> itk::ShapePriorMAPCostFunction< TFeatureImage, TOutputPixel >::itkStaticConstMacro (  ImageDimension  , unsigned  int, TFeatureImage::ImageDimension  )

Dimension constant. Reimplemented from itk::ShapePriorMAPCostFunctionBase< TFeatureImage, TOutputPixel >.

template<class TFeatureImage, class TOutputPixel> Pointer itk::ShapePriorMAPCostFunction< TFeatureImage, TOutputPixel >::New (   )  [static]

Method for creation through the object factory. Reimplemented from itk::Object.

template<class TFeatureImage, class TOutputPixel> void itk::ShapePriorMAPCostFunction< TFeatureImage, TOutputPixel >::PrintSelf (  std::ostream &  os, Indent  indent )  const [protected, virtual]

Methods invoked by Print() to print information about the object including superclasses. Typically not called by the user (use Print() instead) but used in the hierarchical print process to combine the output of several classes. Reimplemented from itk::ShapePriorMAPCostFunctionBase< TFeatureImage, TOutputPixel >.

template<class TFeatureImage, class TOutputPixel> virtual void itk::ShapePriorMAPCostFunction< TFeatureImage, TOutputPixel >::SetShapeParameterMeans (  ArrayType  _arg  )  [virtual]

Set/Get the array of shape parameters mean.

template<class TFeatureImage, class TOutputPixel> virtual void itk::ShapePriorMAPCostFunction< TFeatureImage, TOutputPixel >::SetShapeParameterStandardDeviations (  ArrayType  _arg  )  [virtual]

Set/Get the array of shape parameters standard deviation.

template<class TFeatureImage, class TOutputPixel> virtual void itk::ShapePriorMAPCostFunction< TFeatureImage, TOutputPixel >::SetWeights (  WeightsType  _arg  )  [virtual]

Set/Get the weights for each term. Default is a vector of all ones. The weights are applied to terms in the following order: LogInsideTerm, LogGradientTerm, LogShapePriorTerm and LogPosePriorTerm.

---

The documentation for this class was generated from the following file:

• itkShapePriorMAPCostFunction.h

---