lti::kNearestNeighFilter Class Reference

A smoothness filter for label images (channel8,matrix<int>) which uses the idea of the k-Nearest Neighbour concept. More...

#include <ltiKNearestNeighFilter.h>

Inheritance diagram for lti::kNearestNeighFilter:

[legend]

Collaboration diagram for lti::kNearestNeighFilter:

[legend]

List of all members.

Classes
class	parameters
	the parameters for the class kNearestNeighFilter More...
Public Member Functions
	kNearestNeighFilter ()
	kNearestNeighFilter (const kNearestNeighFilter &other)
virtual	~kNearestNeighFilter ()
virtual const char *	getTypeName () const
bool	apply (channel8 &srcdest)
bool	apply (const channel8 &src, channel8 &dest)
bool	apply (imatrix &srcdest)
bool	apply (const imatrix &src, imatrix &dest)
kNearestNeighFilter &	copy (const kNearestNeighFilter &other)
kNearestNeighFilter &	operator= (const kNearestNeighFilter &other)
virtual functor *	clone () const
const parameters &	getParameters () const
Protected Member Functions
int	getMostLabel (const ivector &histogram, const imatrix &src, const int &row, const int &col) const
int	getMedian (const ivector &histogram, const int max, const int numOfMax) const
void	histogramMethodMiddle (const imatrix &src, imatrix &dest, ivector &histogram, const int &row, int &col) const
bool	histogramMethodZero (const imatrix &src, imatrix &dest) const
bool	histogramMethodConstant (const imatrix &src, imatrix &dest) const
bool	histogramMethodPeriodic (const imatrix &src, imatrix &dest) const
bool	histogramMethodMirror (const imatrix &src, imatrix &dest) const
bool	histogramMethodNoBoundary (const imatrix &src, imatrix &dest) const
Protected Attributes
int	sizeOfKernel
int	histoSize

---

Detailed Description

A smoothness filter for label images (channel8,matrix<int>) which uses the idea of the k-Nearest Neighbour concept.

This means, this filter will assign to the origin position of the kernel mask, the label with the most number of instances within the kernel region. The filter removes disturbance pixels and smooths edges.

In case that more than one label has the "maximal" number of instances within the kernel mask, then the assigned label will

1. remain unchanged, if the kernel origin contains already one of the most frequent labels.
2. be the median of all "winner" labels will.

The algorithms uses an optimized histogram actualization method, similar to the one used for the median filter of channel8 (see also Sonka et.al, Image Processing, Analysis, and Machine Vision, 2nd edition, 1998, pp. 74-76

---

Constructor & Destructor Documentation

lti::kNearestNeighFilter::kNearestNeighFilter

(

)

default constructor

lti::kNearestNeighFilter::kNearestNeighFilter

(

const kNearestNeighFilter &

other

)

copy constructor

Parameters:

other

the object to be copied

virtual lti::kNearestNeighFilter::~kNearestNeighFilter

(

)

[virtual]

destructor

---

Member Function Documentation

bool lti::kNearestNeighFilter::apply	(	const imatrix &	src,
		imatrix &	dest
	)

operates on a copy of the given parameters.

Parameters:

	src	channel8 with the source data.
	dest	channel8 where the result will be left.

Returns:

true if apply successful or false otherwise.

bool lti::kNearestNeighFilter::apply

(

imatrix &

srcdest

)

operates on the given parameter.

Parameters:

srcdest

channel8 with the source data. The result will be left here too.

Returns:

true if apply successful or false otherwise.

bool lti::kNearestNeighFilter::apply	(	const channel8 &	src,
		channel8 &	dest
	)

operates on a copy of the given parameters.

Parameters:

	src	channel8 with the source data.
	dest	channel8 where the result will be left.

Returns:

true if apply successful or false otherwise.

bool lti::kNearestNeighFilter::apply

(

channel8 &

srcdest

)

operates on the given parameter.

Parameters:

srcdest

channel8 with the source data. The result will be left here too.

Returns:

true if apply successful or false otherwise.

virtual functor* lti::kNearestNeighFilter::clone

(

)

const [virtual]

returns a pointer to a clone of this functor.

Implements lti::functor.

kNearestNeighFilter& lti::kNearestNeighFilter::copy

(

const kNearestNeighFilter &

other

)

copy data of "other" functor.

Parameters:

other

the functor to be copied

Returns:

a reference to this functor object

Reimplemented from lti::functor.

int lti::kNearestNeighFilter::getMedian	(	const ivector &	histogram,
		const int	max,
		const int	numOfMax
	)			const [protected]

returns the median of all (max)histogram entries

int lti::kNearestNeighFilter::getMostLabel	(	const ivector &	histogram,
		const imatrix &	src,
		const int &	row,
		const int &	col
	)			const [protected]

return the most avaible label in the given histogram

const parameters& lti::kNearestNeighFilter::getParameters

(

)

const

returns used parameters

Reimplemented from lti::modifier.

virtual const char* lti::kNearestNeighFilter::getTypeName

(

)

const [virtual]

returns the name of this type ("kNearestNeighFilter")

Reimplemented from lti::modifier.

bool lti::kNearestNeighFilter::histogramMethodConstant	(	const imatrix &	src,
		imatrix &	dest
	)			const [protected]

applies the histogramMethod for the type boundary Constant

void lti::kNearestNeighFilter::histogramMethodMiddle	(	const imatrix &	src,
		imatrix &	dest,
		ivector &	histogram,
		const int &	row,
		int &	col
	)			const [protected]

runs inside the image src

bool lti::kNearestNeighFilter::histogramMethodMirror	(	const imatrix &	src,
		imatrix &	dest
	)			const [protected]

applies the histogramMethod for the type boundary Mirror

bool lti::kNearestNeighFilter::histogramMethodNoBoundary	(	const imatrix &	src,
		imatrix &	dest
	)			const [protected]

applies the histogramMethod for the type boundary NoBoundary

bool lti::kNearestNeighFilter::histogramMethodPeriodic	(	const imatrix &	src,
		imatrix &	dest
	)			const [protected]

applies the histogramMethod for the type boundary Periodic

bool lti::kNearestNeighFilter::histogramMethodZero	(	const imatrix &	src,
		imatrix &	dest
	)			const [protected]

applies the histogramMethod for the type boundary Zero

kNearestNeighFilter& lti::kNearestNeighFilter::operator=

(

const kNearestNeighFilter &

other

)

alias for copy member

Parameters:

other

the functor to be copied

Returns:

a reference to this functor object

Reimplemented from lti::functor.

---

Member Data Documentation

int lti::kNearestNeighFilter::histoSize [protected]

size of the histogram (= max.

labelnumber from src-image)

int lti::kNearestNeighFilter::sizeOfKernel [protected]

the corrected kernel size (always odd number)

---

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

• ltiKNearestNeighFilter.h