mds::img::CVolumeGaussFilter< V > Class Template Reference

3D Gaussian smoothing filter. More...

#include <mdsGaussian.h>

Inheritance diagram for mds::img::CVolumeGaussFilter< V >:

List of all members.

Public Types
enum	{ TEMPLATE_PARAMETER_IS_NOT_VOLUME = V::CLASS_VOLUME }
	Check that type V is a volume. More...
typedef CVolumeFilter< V >	base
	Volume filter base.
typedef mds::base::CData < tConvKernelData >	data
	Data base.
typedef data::tConstDataPtr	tConstDataPtr
	Pointer to the constant data.
typedef data::tData	tData
	Data type.
typedef data::tDataPtr	tDataPtr
	Pointer to the data.
typedef mds::base::CFunctor < bool, int, int >	tProgressFunc
	Progress observer (= functor).
typedef CProgressInitializer	tProgressInitializer
	Helper class which can be used to call methods beginProgress() and endProgress() automatically.
typedef base::tVolume	tVolume
	Volume type.
typedef base::tVoxel	tVoxel
	Volume voxel type.
Public Member Functions
tData &	at (tSize x, tSize y, tSize z)
	Returns reference to the kernel coefficient [x][y][z].
const tData &	at (tSize x, tSize y, tSize z) const
void	beginProgress ()
	This function must be called by the class realizing an operation before it starts.
void	create (tSize Size)
	Creates a new 2D convolution kernel.
	CVolumeGaussFilter (double dSigma)
	Constructor that creates Gaussian filter having a specified standard deviation ("sigma" parameter).
	CVolumeGaussFilter (tSize Size)
	Constructor that creates Gaussian filter having the standard deviation corresponding to a given kernel size.
void	deregisterProgressFunc ()
	Sets the default empty observer.
void	endProgress ()
	The function must be called by the class realizing an operation every time the operation finishes.
int	getProgressCount () const
	Returns current value of the internal progress value.
int	getProgressMax () const
	Returns the maximum allowed value of the internal progress counter.
tDataPtr	getPtr (tSize x, tSize y, tSize z)
	Returns pointer to a given kernel data [x][y][z].
tConstDataPtr	getPtr (tSize x, tSize y, tSize z) const
tDataPtr	getPtr ()
	Returns pointer to the kernel data.
tConstDataPtr	getPtr () const
tVoxel	getResponse (const tVolume &SrcVolume, tSize x, tSize y, tSize z)
	Returns filter response at specified volume point.
double	getSigma () const
	Returns filter standard deviation.
tSize	getSize () const
	Returns the kernel size.
bool	operator() (const tVolume &SrcVolume, tVolume &DstVolume)
	Filtering of input/source image.
tData &	operator() (tSize i)
	Returns reference to the subscripted element [i].
const tData &	operator() (tSize i) const
tData &	operator() (tSize x, tSize y, tSize z)
	Returns the subscripted kernel coefficient [x][y][z].
const tData &	operator() (tSize x, tSize y, tSize z) const
bool	progress (int iIncrement=1)
	Calls the registered observer.
void	registerProgressFunc (const tProgressFunc &Func)
	Sets observer/functor called to inform a calling process about the current state of the running operation.
void	resize (tSize Size)
	Sets the filter size.
void	set (tSize x, tSize y, tSize z, const tData &Value)
	Sets the kernel data.
void	setProgressMax (int iValue)
	Sets the maximal counter value.
void	setSigma (double dSigma)
	Sets the filter standard deviation.
	~CVolumeGaussFilter ()
	Destructor.
Static Public Member Functions
static double	getGaussianFuncValue (double dX, double dY, double dZ, double dSigma)
	Returns value of the Gaussian function.
static tSize	sigma2Size (double dSigma)
	Returns kernel size corresponding to a given sigma parameter.
static double	size2Sigma (tSize Size)
	Returns sigma parametr according to a given kernel size.
Protected Member Functions
void	init (tSize Size)
	Allocates data of a given size.
void	initKernel (double dSigma)
	Computes the Gaussian filter kernel.
Protected Attributes
double	m_dSigma
	Standard deviation ("sigma") of the Gaussian filter.
tSize	m_YOffset
	Kernel size.
tSize	m_ZOffset

---

Detailed Description

template<class V>
class mds::img::CVolumeGaussFilter< V >

3D Gaussian smoothing filter.

• Parameter V is a volume type.
• Policy N is used for filter response normalization.
• Gaussian filter kernel is defined by the following function:
  1 (x^2 + y^2 + z^2)
  G_sigma(x,y,z) = ---------------------- * exp(- -----------------)
  (2 * pi)^3/2 * sigma^3 2 * sigma^2
  

---

Member Typedef Documentation

template<class V >

typedef CVolumeFilter<V> mds::img::CVolumeGaussFilter< V >::base

Volume filter base.

Reimplemented from mds::base::CData< tConvKernelData >.

typedef mds::base::CData<tConvKernelData> mds::img::CKernel3D::data [inherited]

Data base.

typedef data::tConstDataPtr mds::img::CKernel3D::tConstDataPtr [inherited]

Pointer to the constant data.

Reimplemented from mds::base::CData< tConvKernelData >.

typedef data::tData mds::img::CKernel3D::tData [inherited]

Data type.

Reimplemented from mds::base::CData< tConvKernelData >.

typedef data::tDataPtr mds::img::CKernel3D::tDataPtr [inherited]

Pointer to the data.

Reimplemented from mds::base::CData< tConvKernelData >.

typedef mds::base::CFunctor<bool, int, int> mds::mod::CProgress::tProgressFunc [inherited]

Progress observer (= functor).

• You may return false from the observer if you want to terminate the running operation.
• Parameters describe current and final state of the pending operation (counter and maximal value).
• If the second parameter is equal to -1, termination time cannot be determined, the current state is unknown.

typedef CProgressInitializer mds::mod::CProgress::tProgressInitializer [inherited]

Helper class which can be used to call methods beginProgress() and endProgress() automatically.

Create static object of this type at the begining of a method that realizes an operation.

template<class V >

typedef base::tVolume mds::img::CVolumeGaussFilter< V >::tVolume

Volume type.

Reimplemented from mds::img::CVolumeFilter< V >.

template<class V >

typedef base::tVoxel mds::img::CVolumeGaussFilter< V >::tVoxel

Volume voxel type.

Reimplemented from mds::img::CVolumeFilter< V >.

---

Member Enumeration Documentation

template<class V>

anonymous enum [inherited]

Check that type V is a volume.

You will see name of this enum somewhere in compiler error message if the type V is not volume.

Enumerator:

TEMPLATE_PARAMETER_IS_NOT_VOLUME

---

Constructor & Destructor Documentation

template<class V >

mds::img::CVolumeGaussFilter< V >::CVolumeGaussFilter

(

double

dSigma

)

Constructor that creates Gaussian filter having a specified standard deviation ("sigma" parameter).

template<class V >

mds::img::CVolumeGaussFilter< V >::CVolumeGaussFilter

(

tSize

Size

)

Constructor that creates Gaussian filter having the standard deviation corresponding to a given kernel size.

template<class V >

mds::img::CVolumeGaussFilter< V >::~CVolumeGaussFilter

(

)

Destructor.

---

Member Function Documentation

tData& mds::img::CKernel3D::at	(	tSize	x,
		tSize	y,
		tSize	z
	)		[inherited]

Returns reference to the kernel coefficient [x][y][z].

• Kernel data are viewed as three dimensional.

const tData& mds::img::CKernel3D::at	(	tSize	x,
		tSize	y,
		tSize	z
	)		const [inherited]

void mds::mod::CProgress::beginProgress

(

)

[inherited]

This function must be called by the class realizing an operation before it starts.

void mds::img::CKernel3D::create

(

tSize

Size

)

[inherited]

Creates a new 2D convolution kernel.

Reimplemented from mds::base::CData< tConvKernelData >.

void mds::mod::CProgress::deregisterProgressFunc

(

)

[inherited]

Sets the default empty observer.

void mds::mod::CProgress::endProgress

(

)

[inherited]

The function must be called by the class realizing an operation every time the operation finishes.

• Sets the default observer.

template<class V >

double CVolumeGaussFilter::getGaussianFuncValue	(	double	dX,
		double	dY,
		double	dZ,
		double	dSigma
	)		[static]

Returns value of the Gaussian function.

int mds::mod::CProgress::getProgressCount

(

)

const [inherited]

Returns current value of the internal progress value.

int mds::mod::CProgress::getProgressMax

(

)

const [inherited]

Returns the maximum allowed value of the internal progress counter.

tDataPtr mds::img::CKernel3D::getPtr	(	tSize	x,
		tSize	y,
		tSize	z
	)		[inherited]

Returns pointer to a given kernel data [x][y][z].

• 3D version.

tConstDataPtr mds::img::CKernel3D::getPtr	(	tSize	x,
		tSize	y,
		tSize	z
	)		const [inherited]

tDataPtr mds::img::CKernel3D::getPtr

(

)

[inherited]

Returns pointer to the kernel data.

Reimplemented from mds::base::CData< tConvKernelData >.

tConstDataPtr mds::img::CKernel3D::getPtr

(

)

const [inherited]

Reimplemented from mds::base::CData< tConvKernelData >.

template<class V >

tVoxel mds::img::CVolumeGaussFilter< V >::getResponse	(	const tVolume &	SrcVolume,
		tSize	x,
		tSize	y,
		tSize	z
	)

Returns filter response at specified volume point.

• Value is not normalized!

template<class V >

double mds::img::CVolumeGaussFilter< V >::getSigma

(

)

const

Returns filter standard deviation.

tSize mds::img::CKernel3D::getSize

(

)

const [inherited]

Returns the kernel size.

Reimplemented from mds::base::CData< tConvKernelData >.

void mds::img::CKernel3D::init

(

tSize

Size

)

[protected, inherited]

Allocates data of a given size.

• This method is primarily designed for kernel data allocation when the default constructor was used.
• Provides no checking of the existing data and its (de,re)allocation.
• Avoid of using it!

template<class V >

void CVolumeGaussFilter::initKernel

(

double

dSigma

)

[protected]

Computes the Gaussian filter kernel.

template<class V >

bool CVolumeGaussFilter::operator()	(	const tVolume &	SrcVolume,
		tVolume &	DstVolume
	)		[virtual]

Filtering of input/source image.

Point filter responses are written to the destination image.

• Returns false on failure.

Implements mds::img::CVolumeFilter< V >.

tData& mds::base::CData< tConvKernelData >::operator()

(

tSize

i

)

[inherited]

Returns reference to the subscripted element [i].

Reimplemented in mds::img::CKernel1D.

const tData& mds::base::CData< tConvKernelData >::operator()

(

tSize

i

)

const [inherited]

Reimplemented in mds::img::CKernel1D.

tData& mds::img::CKernel3D::operator()	(	tSize	x,
		tSize	y,
		tSize	z
	)		[inherited]

Returns the subscripted kernel coefficient [x][y][z].

• Kernel data are viewed as three dimensional.

const tData& mds::img::CKernel3D::operator()	(	tSize	x,
		tSize	y,
		tSize	z
	)		const [inherited]

bool mds::mod::CProgress::progress

(

int

iIncrement = 1

)

[inherited]

Calls the registered observer.

• Returns false if the running operation should be terminated as soon as possible.
• It must be called periodically by the class realizing the operation.

void mds::mod::CProgress::registerProgressFunc

(

const tProgressFunc &

Func

)

[inherited]

Sets observer/functor called to inform a calling process about the current state of the running operation.

• This function should be called before any operation starts.

template<class V >

void CVolumeGaussFilter::resize

(

tSize

Size

)

Sets the filter size.

void mds::img::CKernel3D::set	(	tSize	x,
		tSize	y,
		tSize	z,
		const tData &	Value
	)		[inherited]

Sets the kernel data.

• 3D version.

void mds::mod::CProgress::setProgressMax

(

int

iValue

)

[inherited]

Sets the maximal counter value.

template<class V >

void CVolumeGaussFilter::setSigma

(

double

dSigma

)

Sets the filter standard deviation.

template<class V >

tSize CVolumeGaussFilter::sigma2Size

(

double

dSigma

)

[static]

Returns kernel size corresponding to a given sigma parameter.

template<class V >

double CVolumeGaussFilter::size2Sigma

(

tSize

Size

)

[static]

Returns sigma parametr according to a given kernel size.

---

Member Data Documentation

template<class V >

double mds::img::CVolumeGaussFilter< V >::m_dSigma [protected]

Standard deviation ("sigma") of the Gaussian filter.

tSize mds::img::CKernel3D::m_YOffset [protected, inherited]

Kernel size.

tSize mds::img::CKernel3D::m_ZOffset [protected, inherited]

---

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

• VolumeFilters/mdsGaussian.h
• VolumeFilters/mdsGaussian.hxx