#include <Wavelet.hh>

Collaboration diagram for Wavelet:

Public Member Functions
int	level (ReferenceVector &dest, ReferenceVector &src, int steps)
	Wavelet (FilterSet &fset)
virtual	~Wavelet (void)
void	analysis (ReferenceVector &dest, ReferenceVector &src, int steps)
void	synthesis (ReferenceVector &dest, ReferenceVector &src, int steps, int prevSteps=0)
virtual void	anastep (ReferenceVector &dest, ReferenceVector &src, int size)
virtual void	synstep (ReferenceVector &dest, ReferenceVector &src, int size)
Protected Member Functions
void	init (FilterSet *pfset=NULL)
Protected Attributes
Filter *	m_analow
Filter *	m_anahigh
Filter *	m_synlow
Filter *	m_synhigh
VectorPosition *	m_apad
VectorPosition *	m_slpad
VectorPosition *	m_shpad
bool	m_symmetric
int	m_npad
int *	m_apositions
int	m_apsize
int *	m_slpositions
int	m_slpsize
int *	m_shpositions
int *	m_shsigns
int	m_shpsize

Detailed Description

A Wavelet. The Wavelet transform is one-dimensional, it thus works on vectors.

Definition at line 24 of file Wavelet.hh.

Constructor & Destructor Documentation

Wavelet::Wavelet ( FilterSet & fset )

Constructor. Sets the the filter.

virtual Wavelet::~Wavelet ( void ) [virtual]

Destructor. Releases allocated objects.

Member Function Documentation

void Wavelet::analysis	(	ReferenceVector &	dest,
		ReferenceVector &	src,
		int	steps
	)

Decompose a vector. The result is written to a second vector.

Exceptions:

invalid_argument low pass subband signal too small (decrease the number of transform steps or increase signal size)

Parameters:

src	A vector of {coeff} containing the original data
dest	A vector of {coeff} to get the transformed data
steps	The last col the decomposition step will be performed to

virtual void Wavelet::anastep	(	ReferenceVector &	dest,
		ReferenceVector &	src,
		int	size
	)		`[virtual]`

A decomposition step on a vector. The result is written to a second vector. Only the result coefficients are being written, the rest may have to be done manually.

Parameters:

dest	A vector of {coeff} to get the transformed data
src	A vector of {coeff} containing the original data
size	The number of values in {src} to perform the decomposition step on

void Wavelet::init ( FilterSet * pfset = NULL ) [protected]

Initialize the filter set to be used. The filters in the set will be duplicated.

Parameters:

pfset a reference to the filter set

int Wavelet::level	(	ReferenceVector &	dest,
		ReferenceVector &	src,
		int	steps
	)

Calculate the level for the transform and execute some consistency checks.

Exceptions:

invalid_argument vector size is not a power of two or the two vectors don't have the same size

Parameters:

dest	the destination vector
src	the source vector
steps	the number of steps for the transform

virtual void Wavelet::synstep	(	ReferenceVector &	dest,
		ReferenceVector &	src,
		int	size
	)		`[virtual]`

A reconstruction step on a vector. The result is written to a second vector. Only the result coefficients are being written, the rest may have to be done manually.

Parameters:

dest	A vector of {coeff} to get the transformed data
src	A vector of {coeff} containing the original data
size	The number of values in {src} to perform the synthesis step on

void Wavelet::synthesis	(	ReferenceVector &	dest,
		ReferenceVector &	src,
		int	steps,
		int	prevSteps = `0`
	)

Reconstruct a vector. The result is written to a second vector.

Parameters:

src	A vector of {coeff} containing the analysisd data
dest	A vector of {coeff} to get the synthesised data
steps	The number of coefficient pairs to synthesis.
prevSteps	if greater than 0, the number of previous decomposition steps to assume. This is necessary to reconstruct images that are not square and have side lengths which are not powers of two.