Performs defuzzification on a membership image. More...

#include <itkFuzzyClassifierImageFilter.h>

Inheritance diagram for itk::FuzzyClassifierImageFilter< TInputVectorImage, TLabel >:

Collaboration diagram for itk::FuzzyClassifierImageFilter< TInputVectorImage, TLabel >:

Public Types
typedef SmartPointer< const Self >	ConstPointer

typedef DecisionRuleType::Pointer	DecisionRulePointer

typedef itk::Statistics::MaximumDecisionRule	DecisionRuleType
	Decision rule used in the classification process to compare the membership scores and return a class label.

typedef ImageRegionConstIterator< InputImageType >	InputImageConstIterator
	Other input image typedefs.

typedef Superclass::InputImageConstPointer	InputImageConstPointer

typedef Superclass::InputImagePixelType	InputImagePixelType

typedef Superclass::InputImagePointer	InputImagePointer
	Input image typedef support (types inherited from the superclass).

typedef Superclass::InputImageRegionType	InputImageRegionType

typedef TInputVectorImage	InputImageType
	Convenient typedefs for simplifying declarations.

typedef TLabel	LabelType

typedef ImageRegionIterator< OutputImageType >	OutputImageIterator
	Other output image typedefs.

typedef Superclass::OutputImagePixelType	OutputImagePixelType

typedef Superclass::OutputImagePointer	OutputImagePointer
	Output image typedef support (types inherited from the superclass).

typedef Superclass::OutputImageRegionType	OutputImageRegionType

typedef Image< LabelType, itkGetStaticConstMacro(InputImageDimension) >	OutputImageType
	Convenient typedefs for simplifying declarations.

typedef SmartPointer< Self >	Pointer

typedef FuzzyClassifierImageFilter	Self
	Standard class typedefs.

typedef ImageToImageFilter< InputImageType, OutputImageType >	Superclass
	Standard class typedefs.

Public Member Functions
	itkNewMacro (Self)
	Method for creation through the object factory.

	itkStaticConstMacro (InputImageDimension, unsigned int, InputImageType::ImageDimension)
	Extract dimension from input image.

	itkTypeMacro (FuzzyClassifierImageFilter, ImageToImageFilter)
	Run-time type information (and related methods).

Protected Member Functions
	FuzzyClassifierImageFilter ()
	Constructor.

virtual	~FuzzyClassifierImageFilter ()
	Destructor.

virtual void	Classify ()
	This method computes the labeled map.

void	GenerateData () override
	Standard pipeline method.

void	PrintSelf (std::ostream &os, Indent indent) const override
	Write the name-value pairs of the filter data members to the supplied output stream.

Detailed Description

template<class TInputVectorImage, class TLabel = unsigned char>
class itk::FuzzyClassifierImageFilter< TInputVectorImage, TLabel >

Performs defuzzification on a membership image.

The filter implements the defuzzification (i.e., the assignment of each pixel to a particular class based on their memberships) by applying a Maximum Decision Rule to the membership image.

If a pixel of the membership image has values -1, that pixel is assumed to be a background pixel and should be ignored. The value of the label assigned to those pixels will be equal to the number of classes. Otherwise, the pixel values of the resulting label map will indicate the classes they correspond to (e.g. labels with value 0 belong to the class 0, labels with value 1 belong to the first class, etc.). Thus the range of label values is 0 to the number of classes.

Inputs and Outputs: The input to this filter is an itk::VectorImage that represents pixel memberships to 'n' classes. You may use a subclass of FuzzyClassifierInitializationImageFilter to generate the membership images or specify your own.

: The output of the filter is a label map (an image of unsigned char is the default).

Template parameters: This filter is templated over the input vector image type and the data type of the output label map (defaults to unsigned char).

Version: 0.1

Author: Alberto Rey, Alfonso Castro and Bernardino Arcay. University of A Coruña. Spain

See also: FuzzyClassifierInitializationImageFilter; VectorImage