Accumulator Class Reference

#include <Accumulator.h>

Inheritance diagram for Accumulator:

Collaboration diagram for Accumulator:

Public Types
enum	FieldType {   DATA_SPECIFIC = 32 , QUALITY = 256 , DATA = 'd' , WEIGHT = 'w'|QUALITY ,   COUNT = 'c'|QUALITY , DEVIATION = 's'|QUALITY , WEIGHT_DS = 'W'|QUALITY , COUNT_DS = 'C'|QUALITY ,   DEVIATION_DS = 'S'|QUALITY }
typedef drain::EnumDict< FieldType >::dict_t	dict_t
typedef std::list< FieldType >	FieldList

Public Member Functions
	Accumulator ()
	Accumulator (const Accumulator &acc)
void	setMethod (const std::string &method)
	Set method to some of the predefined methods.
void	setMethod (const std::string &name, const std::string &params)
	Set method to some of the predefined methods.
void	setMethod (const AccumulationMethod &method)
	Copies the method and its parameters.
bool	isMethodSet () const
const AccumulationMethod &	getMethod () const
AccumulationMethod &	getMethod ()
void	add (const size_t i, double value, double weight)
	Adds decoded data that applies natural scaling.
void	add (const size_t i, double value, double weight, unsigned int count)
	Adds decoded data that applies natural scaling.
void	addData (const Image &src, const AccumulationConverter &converter, double weight=1.0, int iOffset=0, int jOffset=0)
	Add (accumulate) data with given prior weight.
void	addData (const Image &src, const Image &srcQuality, const AccumulationConverter &converter, double weight=1.0, int iOffset=0, int jOffset=0)
	Add (accumulate) data weighted with input quality.
void	addData (const Image &src, const Image &srcQuality, const Image &srcCount, const AccumulationConverter &converter)
	Add (accumulate) data weighted with input quality and count; each data(i,j) contributes count(i,j) times.
void	extractField (char field, const AccumulationConverter &converter, Image &dst, const drain::Rectangle< int > &crop) const
	Extracts the accumulated quantity or secondary quantities like weight and standard deviation.
void	extractField (FieldType field, const AccumulationConverter &converter, Image &dst, const drain::Rectangle< int > &crop) const
virtual std::ostream &	toStream (std::ostream &ostr) const

Static Public Member Functions
static char	getFieldChar (FieldType field)
static FieldType	getField (char field)
static bool	isQuality (FieldType field)
static bool	isSpecific (FieldType field)
	Future option to mark scaled/normalized etc.
static void	getFields (const std::string &fieldStr, FieldList &fieldList)
static void	createFieldList (const std::string &fieldChars, FieldList &fieldList)

Public Attributes
AccumulationArray	accArray
	Todo: export.

Protected Member Functions
void	initDst (const AccumulationConverter &coder, Image &dst, drain::Rectangle< int > &cropArea) const

Protected Attributes
AccumulationMethod	undefinedMethod
	A Some predefined methods, that can be set with setMethod(const std::string & key).
AccumulationMethod *	methodPtr

Detailed Description

Accumulator contains five built-in accumulation methods (principles):

1. overwrite
2. maximum
3. average
4. weighted average
5. maximum weight The methods can be extended user defined methods with addMethod() .

Member Enumeration Documentation

FieldType

enum FieldType

FIX: DS = DataSet-specific - forward definition

Constructor & Destructor Documentation

Accumulator()

Accumulator ( )

inline

TODO: a static map?

Member Function Documentation

addData() [1/3]

void addData	(	const Image &	src,
		const AccumulationConverter &	converter,
		double	weight = 1.0,
		int	iOffset = 0,
		int	jOffset = 0
	)

Add (accumulate) data with given prior weight.

Parameters

src	- input data

addData() [2/3]

void addData	(	const Image &	src,
		const Image &	srcQuality,
		const AccumulationConverter &	converter,
		double	weight = 1.0,
		int	iOffset = 0,
		int	jOffset = 0
	)

Add (accumulate) data weighted with input quality.

Parameters

src	- input data
srcQuality	- quality field of input data

addData() [3/3]

void addData	(	const Image &	src,
		const Image &	srcQuality,
		const Image &	srcCount,
		const AccumulationConverter &	converter
	)

Add (accumulate) data weighted with input quality and count; each data(i,j) contributes count(i,j) times.

Parameters

src	- input data
srcQuality	- quality field of input data
srcCount	- counts of input data (makes difference in resulting weights)

Count is bypassed when necessary; for example, when the accumulation method is MAXIMUM.

createFieldList()

void createFieldList ( const std::string &  fieldChars, FieldList &  fieldList  )

static

TODO: add: static bool getField(char field){ looping } to Accumulator

extractField()

void extractField ( char  field, const AccumulationConverter &  converter, Image &  dst, const drain::Rectangle< int > &  crop  ) const

inline

Extracts the accumulated quantity or secondary quantities like weight and standard deviation.

Parameters

field

- data layer to be extracted d = data; w = weight (8bit uchar); W = with same storage type as data c = count (8bit uchar); s = standard deviation, unscaled

setMethod()

void setMethod ( const AccumulationMethod &  method )

inline

Copies the method and its parameters.

That is, does not copy the target (accumulator &). This limits to those already registered in AccMethodBank.

---

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

• src/drain/image/Accumulator.h
• src/drain/image/Accumulator.cpp