Rank filter of a greyscale image. More...

#include <itkRankImageFilter.h>

Inheritance diagram for itk::RankImageFilter< TInputImage, TOutputImage, TKernel >:

Collaboration diagram for itk::RankImageFilter< TInputImage, TOutputImage, TKernel >:

Public Types
typedef SmartPointer< const Self >	ConstPointer
typedef Superclass::HistogramType	HistogramType
typedef TInputImage::IndexType	IndexType
typedef TInputImage	InputImageType
typedef TInputImage::PixelType	InputPixelType
typedef KernelType::ConstIterator	KernelIteratorType
typedef TKernel	KernelType
typedef TInputImage::OffsetType	OffsetType
typedef Superclass::OutputImageRegionType	OutputImageRegionType
typedef TOutputImage	OutputImageType
typedef TOutputImage::PixelType	OutputPixelType
typedef TInputImage::PixelType	PixelType
typedef SmartPointer< Self >	Pointer
typedef KernelType::SizeType	RadiusType
typedef TInputImage::RegionType	RegionType
typedef RankImageFilter	Self
typedef TInputImage::SizeType	SizeType
typedef MovingHistogramImageFilter < TInputImage, TOutputImage, TKernel, Function::RankHistogram < typename TInputImage::PixelType > >	Superclass
Public Member Functions
virtual ::itk::LightObject::Pointer	CreateAnother (void) const
virtual const char *	GetNameOfClass () const
virtual float	GetRank () const
bool	GetUseVectorBasedAlgorithm ()
virtual void	SetRank (float _arg)
Static Public Member Functions
static Pointer	New ()
Static Public Attributes
static const unsigned int	ImageDimension = TInputImage::ImageDimension
Protected Member Functions
virtual void	ConfigureHistogram (HistogramType &histogram)
void	PrintSelf (std::ostream &os, Indent indent) const
	RankImageFilter ()
	~RankImageFilter ()
Private Member Functions
void	operator= (const Self &)
	RankImageFilter (const Self &)
Private Attributes
float	m_Rank

Detailed Description

template<class TInputImage, class TOutputImage, class TKernel = FlatStructuringElement< ::itk::GetImageDimension< TInputImage >::ImageDimension >>
class itk::RankImageFilter< TInputImage, TOutputImage, TKernel >

Rank filter of a greyscale image.

Nonlinear filter in which each output pixel is a user defined rank of input pixels in a user defined neighborhood. The default rank is 0.5 (median). The boundary conditions are different to the standard itkMedianImageFilter. In this filter the neighborhood is cropped at the boundary, and is therefore smaller.

This filter uses a recursive implementation - essentially the one by Huang 1979, I believe, to compute the rank, and is therefore usually a lot faster than the direct implementation. The extensions to Huang are support for arbitary pixel types (using c++ maps) and arbitary neighborhoods. I presume that these are not new ideas.

This filter is based on the sliding window code from the consolidatedMorphology package on InsightJournal.

The structuring element is assumed to be composed of binary values (zero or one). Only elements of the structuring element having values > 0 are candidates for affecting the center pixel.

This code was contributed in the Insight Journal paper: "Efficient implementation of kernel filtering" by Beare R., Lehmann G http://hdl.handle.net/1926/555 http://www.insight-journal.org/browse/publication/160

Author:: Richard Beare

Definition at line 68 of file itkRankImageFilter.h.