Isis::Equalization Class Reference

This class can be used to calculate, read in, and/or apply equalization statistics for a list of files. More...

#include <Equalization.h>

Inheritance diagram for Isis::Equalization:

Collaboration diagram for Isis::Equalization:

Classes
class	ApplyFunctor
class	CalculateFunctor
class	ImageAdjustment

Public Member Functions
	Equalization (OverlapNormalization::SolutionType sType, QString fromListName)
void	addHolds (QString holdListName)
void	calculateStatistics (double sampPercent, int mincnt, bool wtopt, LeastSquares::SolveMethod methodType)
void	importStatistics (QString instatsFileName)
void	applyCorrection (QString toListName)
PvlGroup	getResults ()
void	write (QString outstatsFileName)
double	evaluate (double dn, int imageIndex, int bandIndex) const

Protected Member Functions
void	loadInputs (QString fromListName)
void	setInput (int index, QString value)
const FileList &	getInputs () const
virtual void	fillOutList (FileList &outList, QString toListName)
virtual void	errorCheck (QString fromListName)
void	generateOutputs (FileList &outList)
void	loadOutputs (FileList &outList, QString toListName)
void	loadHolds (OverlapNormalization *oNorm)
void	setResults (std::vector< OverlapStatistics > &overlapStats)
void	setResults ()
void	clearAdjustments ()
void	addAdjustment (ImageAdjustment *adjustment)
void	addValid (int count)
void	addInvalid (int count)

Private Member Functions
void	init ()
std::vector< int >	validateInputStatistics (QString instatsFileName)

Private Attributes
FileList	m_imageList
std::vector< ImageAdjustment * >	m_adjustments
std::vector< int >	m_holdIndices
int	m_validCnt
int	m_invalidCnt
int	m_mincnt
bool	m_wtopt
int	m_maxCube
int	m_maxBand
OverlapNormalization::SolutionType	m_sType
Pvl *	m_results

Detailed Description

This class can be used to calculate, read in, and/or apply equalization statistics for a list of files.

• Calculating equalization statistics
  • An optional list of images to hold may be given before calculations.
  • In order to calculate statistics, the class will need to know the following:
    • percentage of lines to be used for calculations,
    • the minimum number of points in overlapping area to be used,
    • whether overlapping areas should be weighted based on number of valid pixels,
    • whether to calculate gain, offset, or both
    • which LeastSquares solve method.
  • Once calculated, these statistics can be returned as a PvlGroup, written to a text file, and/or applied to the images in the input file list.
• Importing equalization statistics
  • Statistics may be imported from a given file name and then applied to the images in the input file list.
• Applying equalization statistics
  • Statistics must be calculated or imported before they can be applied to the images in the input file list.

Code example for calculating statistics, writing results to a Pvl, writing results to a file, and applying results: Equalization eq(inputCubeListFileName); // calculate eq.addHolds(holdListFileName); eq.calculateStatistics(sampPercent, mincnt, wtopt, sType, methodType); // write to PvlGroup PvlGroup resultsGroup = eq.getResults(); // write to file eq.write(outputStatisticsFileName); // apply corrections to cubes in input list eq.applyCorrection();

Code example for importing statistics and applying them. Equalization eq(listFileName); eq.importStatistics(inputStatisticsFileName); // create new output cubes from equalized cube list and apply corrections eq.applyCorrection(equalizedCubeListFileName);

This class contains the classes ImageAdjustment, CalculateFunctor, and ApplyFunctor.

Author

????-??-?? Unknown

History:

2012-04-26 Jeannie Backer - Added includes to Filelist, Pvl, PvlGroup, and Statistics classes to the implementation file.

2013-01-29 Jeannie Backer - Added input parameter to calculateStatistics() method. Added error catch to improve thrown message. Fixes #962.

2013-01-29 Jeannie Backer - Fixed bugs from previous checkin. Improved test coverage by more than 23% for each coverage type.

2013-02-06 Steven Lambright - Made the OverlapNormalization::SolutionType into a member required upon instantiation in order to support gain adjustments without normalization, which uses a separate formula in evaluate() instead of just modifying how statistics are computed. Implemented the GainsWithoutNormalization solution type. References #911.

Definition at line 124 of file Equalization.h.

Constructor & Destructor Documentation

Isis::Equalization::Equalization	(	OverlapNormalization::SolutionType	sType,
		QString	fromListName
	)

Parameters

sType

An integer value corresponding to the enumerated value of the OverlapNormalization::SolutionType to be used.

Definition at line 34 of file Equalization.cpp.

Member Function Documentation

void Isis::Equalization::calculateStatistics	(	double	percent,
		int	mincnt,
		bool	wtopt,
		LeastSquares::SolveMethod	methodType
	)

Parameters

percent	Percentage of the lines to consider when gathering overall cube statistics and overlap statistics
mincnt	Minimum number of points in overlapping area required to be used in the solution
wtopt	Indicates whether overlaps should be weighted
methodType	An integer value corresponding to the enumerated value of the desired LeastSquares::SolveMethod to be used.

Definition at line 91 of file Equalization.cpp.

References _FILEINFO_, Isis::ProcessByBrick::EndProcess(), Isis::OverlapStatistics::GetMStats(), Isis::OverlapStatistics::HasOverlap(), Isis::Cube::open(), Isis::ProcessByLine::ProcessCubeInPlace(), Isis::Process::Progress(), Isis::ProcessByLine::SetInputCube(), Isis::OverlapStatistics::SetMincount(), Isis::Progress::SetText(), Isis::toString(), Isis::MultivariateStatistics::ValidPixels(), Isis::MultivariateStatistics::X(), and Isis::MultivariateStatistics::Y().

---

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

• src/base/objs/Equalization/Equalization.h
• src/base/objs/Equalization/Equalization.cpp