lti::crossCorrelationCoefficient Class Reference

This class computes the normalized cross correlation coefficient between two matrices or vectors. More...

#include <ltiCrossCorrelationCoefficient.h>

Inheritance diagram for lti::crossCorrelationCoefficient:

[legend]

Collaboration diagram for lti::crossCorrelationCoefficient:

[legend]

List of all members.

Classes
class	parameters
	The parameters for the class crossCorrelationCoefficient. More...
Public Member Functions
	crossCorrelationCoefficient ()
	crossCorrelationCoefficient (const parameters &par)
	crossCorrelationCoefficient (const crossCorrelationCoefficient &other)
virtual	~crossCorrelationCoefficient ()
virtual const char *	getTypeName () const
crossCorrelationCoefficient &	copy (const crossCorrelationCoefficient &other)
crossCorrelationCoefficient &	operator= (const crossCorrelationCoefficient &other)
virtual functor *	clone () const
const parameters &	getParameters () const
Apply methods with one set of control points
bool	apply (const matrix< ubyte > &src1, const matrix< ubyte > &src2, const pointList &controlPoints, const fvector &transf, float &dest) const
bool	apply (const matrix< float > &src1, const matrix< float > &src2, const pointList &controlPoints, const fvector &transf, float &dest) const
bool	apply (const matrix< ubyte > &src1, const matrix< ubyte > &src2, const pointList &controlPoints, const dvector &transf, double &dest) const
bool	apply (const matrix< float > &src1, const matrix< float > &src2, const pointList &controlPoints, const dvector &transf, double &dest) const
Apply methods with two sets of control points
bool	apply (const matrix< ubyte > &src1, const matrix< ubyte > &src2, const pointList &pts1, const pointList &pts2, float &dest) const
bool	apply (const matrix< float > &src1, const matrix< float > &src2, const pointList &pts1, const pointList &pts2, float &dest) const
bool	apply (const matrix< ubyte > &src1, const matrix< ubyte > &src2, const pointList &pts1, const pointList &pts2, double &dest) const
bool	apply (const matrix< float > &src1, const matrix< float > &src2, const pointList &pts1, const pointList &pts2, double &dest) const
Apply methods with two points
bool	apply (const matrix< ubyte > &src1, const matrix< ubyte > &src2, const ipoint &pt1, const ipoint &pt2, float &dest) const
bool	apply (const matrix< float > &src1, const matrix< float > &src2, const ipoint &pt1, const ipoint &pt2, float &dest) const
bool	apply (const matrix< ubyte > &src1, const matrix< ubyte > &src2, const ipoint &pt1, const ipoint &pt2, double &dest) const
bool	apply (const matrix< float > &src1, const matrix< float > &src2, const ipoint &pt1, const ipoint &pt2, double &dest) const
Apply methods for entire matrices
bool	apply (const matrix< ubyte > &src1, const matrix< ubyte > &src2, float &dest) const
bool	apply (const matrix< float > &src1, const matrix< float > &src2, float &dest) const
bool	apply (const matrix< ubyte > &src1, const matrix< ubyte > &src2, double &dest) const
bool	apply (const matrix< float > &src1, const matrix< float > &src2, double &dest) const
Apply methods for vectors
bool	apply (const vector< ubyte > &src1, const vector< ubyte > &src2, float &dest) const
bool	apply (const vector< float > &src1, const vector< float > &src2, float &dest) const
bool	apply (const vector< ubyte > &src1, const vector< ubyte > &src2, double &dest) const
bool	apply (const vector< float > &src1, const vector< float > &src2, double &dest) const

---

Detailed Description

This class computes the normalized cross correlation coefficient between two matrices or vectors.

Some applies allow to transform the points in one source using nearest neighbor interpolation prior to computing the cross correlation coefficient. This is only done if you specify a homographyEstimatorBase in the parameters, though. The transformation is a short cut bypassing classes like geometricTransform, which is useful when estimating and verifying transformations (coming soon).

CAVEAT: This class could use more iterator arithmetic inside and thus is not implemented efficiently everywhere. Some applies also are less efficient than they could be because code duplication is avoided.

---

Constructor & Destructor Documentation

lti::crossCorrelationCoefficient::crossCorrelationCoefficient

(

)

Default constructor.

lti::crossCorrelationCoefficient::crossCorrelationCoefficient

(

const parameters &

par

)

Construct a functor using the given parameters.

lti::crossCorrelationCoefficient::crossCorrelationCoefficient

(

const crossCorrelationCoefficient &

other

)

Copy constructor.

Parameters:

other

the object to be copied

virtual lti::crossCorrelationCoefficient::~crossCorrelationCoefficient

(

)

[virtual]

Destructor.

---

Member Function Documentation

bool lti::crossCorrelationCoefficient::apply	(	const vector< float > &	src1,
		const vector< float > &	src2,
		double &	dest
	)			const

Computes the normalized cross correlation coefficient between two vectors only considering all positions inside both vectors.

No window is used. Both vectors most have the same size.

Parameters:

	src1	vector with the first intensities.
	src2	vector with the second intensities.
	dest	the computed normalized cross correlation coefficient.

Returns:

true if apply successful or false otherwise.

bool lti::crossCorrelationCoefficient::apply	(	const vector< ubyte > &	src1,
		const vector< ubyte > &	src2,
		double &	dest
	)			const

Computes the normalized cross correlation coefficient between two vectors only considering all positions inside both vectors.

No window is used. Both vectors most have the same size.

Parameters:

	src1	vector with the first intensities.
	src2	vector with the second intensities.
	dest	the computed normalized cross correlation coefficient.

Returns:

true if apply successful or false otherwise.

bool lti::crossCorrelationCoefficient::apply	(	const vector< float > &	src1,
		const vector< float > &	src2,
		float &	dest
	)			const

Computes the normalized cross correlation coefficient between two vectors only considering all positions inside both vectors.

No window is used. Both vectors most have the same size.

Parameters:

	src1	vector with the first intensities.
	src2	vector with the second intensities.
	dest	the computed normalized cross correlation coefficient.

Returns:

true if apply successful or false otherwise.

bool lti::crossCorrelationCoefficient::apply	(	const vector< ubyte > &	src1,
		const vector< ubyte > &	src2,
		float &	dest
	)			const

Computes the normalized cross correlation coefficient between two vectors only considering all positions inside both vectors.

No window is used. Both vectors most have the same size.

Parameters:

	src1	vector with the first intensities.
	src2	vector with the second intensities.
	dest	the computed normalized cross correlation coefficient.

Returns:

true if apply successful or false otherwise.

bool lti::crossCorrelationCoefficient::apply	(	const matrix< float > &	src1,
		const matrix< float > &	src2,
		double &	dest
	)			const

Computes the normalized cross correlation coefficient between two channels only considering all positions inside both channels.

No window is used. Both channels most have the same size.

Parameters:

	src1	channel with the first intensities.
	src2	channel with the second intensities.
	dest	the computed normalized cross correlation coefficient.

Returns:

true if apply successful or false otherwise.

bool lti::crossCorrelationCoefficient::apply	(	const matrix< ubyte > &	src1,
		const matrix< ubyte > &	src2,
		double &	dest
	)			const

Computes the normalized cross correlation coefficient between two channels only considering all positions inside both channels.

No window is used. Both channels most have the same size.

Parameters:

	src1	channel8 with the first intensities.
	src2	channel8 with the second intensities.
	dest	the computed normalized cross correlation coefficient.

Returns:

true if apply successful or false otherwise.

bool lti::crossCorrelationCoefficient::apply	(	const matrix< float > &	src1,
		const matrix< float > &	src2,
		float &	dest
	)			const

Computes the normalized cross correlation coefficient between two channels only considering all positions inside both channels.

No window is used. Both channels most have the same size.

Parameters:

	src1	channel with the first intensities.
	src2	channel with the second intensities.
	dest	the computed normalized cross correlation coefficient.

Returns:

true if apply successful or false otherwise.

bool lti::crossCorrelationCoefficient::apply	(	const matrix< ubyte > &	src1,
		const matrix< ubyte > &	src2,
		float &	dest
	)			const

Computes the normalized cross correlation coefficient between two channels only considering all positions inside both channels.

No window is used. Both channels most have the same size.

Parameters:

	src1	channel8 with the first intensities.
	src2	channel8 with the second intensities.
	dest	the computed normalized cross correlation coefficient.

Returns:

true if apply successful or false otherwise.

bool lti::crossCorrelationCoefficient::apply	(	const matrix< float > &	src1,
		const matrix< float > &	src2,
		const ipoint &	pt1,
		const ipoint &	pt2,
		double &	dest
	)			const

Computes the normalized cross correlation coefficient between two channels only considering a window located at a single control point.

Only positions inside both channels are considered. In both channels the same window is used.

Parameters:

	src1	channel with the first intensities.
	src2	channel with the second intensities.
	pt1	the first control points.
	pt2	the second control points.
	dest	the computed normalized cross correlation coefficient.

Returns:

true if apply successful or false otherwise.

bool lti::crossCorrelationCoefficient::apply	(	const matrix< ubyte > &	src1,
		const matrix< ubyte > &	src2,
		const ipoint &	pt1,
		const ipoint &	pt2,
		double &	dest
	)			const

Computes the normalized cross correlation coefficient between two channels only considering a window located at a single control point.

Only positions inside both channels are considered. In both channels the same window is used.

Parameters:

	src1	channel8 with the first intensities.
	src2	channel8 with the second intensities.
	pt1	the first control points.
	pt2	the second control points.
	dest	the computed normalized cross correlation coefficient.

Returns:

true if apply successful or false otherwise.

bool lti::crossCorrelationCoefficient::apply	(	const matrix< float > &	src1,
		const matrix< float > &	src2,
		const ipoint &	pt1,
		const ipoint &	pt2,
		float &	dest
	)			const

Computes the normalized cross correlation coefficient between two channels only considering a window located at a single control point.

Only positions inside both channels are considered. In both channels the same window is used.

Parameters:

	src1	channel with the first intensities.
	src2	channel with the second intensities.
	pt1	the first control points.
	pt2	the second control points.
	dest	the computed normalized cross correlation coefficient.

Returns:

true if apply successful or false otherwise.

bool lti::crossCorrelationCoefficient::apply	(	const matrix< ubyte > &	src1,
		const matrix< ubyte > &	src2,
		const ipoint &	pt1,
		const ipoint &	pt2,
		float &	dest
	)			const

Computes the normalized cross correlation coefficient between two channels only considering a window located at a single control point.

Only positions inside both channels are considered. In both channels the same window is used.

Parameters:

	src1	channel8 with the first intensities.
	src2	channel8 with the second intensities.
	pt1	the first control points.
	pt2	the second control points.
	dest	the computed normalized cross correlation coefficient.

Returns:

true if apply successful or false otherwise.

bool lti::crossCorrelationCoefficient::apply	(	const matrix< float > &	src1,
		const matrix< float > &	src2,
		const pointList &	pts1,
		const pointList &	pts2,
		double &	dest
	)			const

Computes the normalized cross correlation coefficient between two channels only considering a window located at each control point.

This is useful to exlude e.g. homogenous areas. Only positions inside both channels are considered. Please specify the same number of points in both point lists. In both channels the same window is used.

Parameters:

	src1	channel with the first intensities.
	src2	channel with the second intensities.
	pts1	pointList with the first control points.
	pts2	pointList with the second control points.
	dest	the computed normalized cross correlation coefficient.

Returns:

true if apply successful or false otherwise.

bool lti::crossCorrelationCoefficient::apply	(	const matrix< ubyte > &	src1,
		const matrix< ubyte > &	src2,
		const pointList &	pts1,
		const pointList &	pts2,
		double &	dest
	)			const

Computes the normalized cross correlation coefficient between two channels only considering a window located at each control point.

This is useful to exlude e.g. homogenous areas. Only positions inside both channels are considered. Please specify the same number of points in both point lists. In both channels the same window is used.

Parameters:

	src1	channel8 with the first intensities.
	src2	channel8 with the second intensities.
	pts1	pointList with the first control points.
	pts2	pointList with the second control points.
	dest	the computed normalized cross correlation coefficient.

Returns:

true if apply successful or false otherwise.

bool lti::crossCorrelationCoefficient::apply	(	const matrix< float > &	src1,
		const matrix< float > &	src2,
		const pointList &	pts1,
		const pointList &	pts2,
		float &	dest
	)			const

Computes the normalized cross correlation coefficient between two channels only considering a window located at each control point.

This is useful to exlude e.g. homogenous areas. Only positions inside both channels are considered. Please specify the same number of points in both point lists. In both channels the same window is used.

Parameters:

	src1	channel with the first intensities.
	src2	channel with the second intensities.
	pts1	pointList with the first control points.
	pts2	pointList with the second control points.
	dest	the computed normalized cross correlation coefficient.

Returns:

true if apply successful or false otherwise.

bool lti::crossCorrelationCoefficient::apply	(	const matrix< ubyte > &	src1,
		const matrix< ubyte > &	src2,
		const pointList &	pts1,
		const pointList &	pts2,
		float &	dest
	)			const

Computes the normalized cross correlation coefficient between two channels only considering a window located at each control point.

This is useful to exlude e.g. homogenous areas. Only positions inside both channels are considered. Please specify the same number of points in both point lists. In both channels the same window is used.

Parameters:

	src1	channel8 with the first intensities.
	src2	channel8 with the second intensities.
	pts1	pointList with the first control points.
	pts2	pointList with the second control points.
	dest	the computed normalized cross correlation coefficient.

Returns:

true if apply successful or false otherwise.

bool lti::crossCorrelationCoefficient::apply	(	const matrix< float > &	src1,
		const matrix< float > &	src2,
		const pointList &	controlPoints,
		const dvector &	transf,
		double &	dest
	)			const

Computes the normalized cross correlation coefficient between two channels only considering a window located at each control point.

This is useful to exlude e.g. homogenous areas. Only positions inside both channels are considered. If the channels are recorded from different positions it is advantageous to supply a transformation vector. Please do no forget to define a transformation in the parameters. On default no transformation is executed. E.g. the control points and the points inside the window have the same coordinates in both channels.

Parameters:

	src1	channel with the first intensities.
	src2	channel with the second intensities.
	controlPoints	pointList with the control points.
	transf	dvector with the transform by which the control points are fit inside the the second image. The window in the second image around the control points is distorted, too.
	dest	the computed normalized cross correlation coefficient.

Returns:

true if apply successful or false otherwise.

bool lti::crossCorrelationCoefficient::apply	(	const matrix< ubyte > &	src1,
		const matrix< ubyte > &	src2,
		const pointList &	controlPoints,
		const dvector &	transf,
		double &	dest
	)			const

Computes the normalized cross correlation coefficient between two channels only considering a window located at each control point.

This is useful to exlude e.g. homogenous areas. Only positions inside both channels are considered. If the channels are recorded from different positions it is advantageous to supply a transformation vector. Please do no forget to define a transformation in the parameters. On default no transformation is executed. E.g. the control points and the points inside the window have the same coordinates in both channels.

Parameters:

	src1	channel8 with the first intensities.
	src2	channel8 with the second intensities.
	controlPoints	pointList with the control points.
	transf	dvector with the transform by which the control points are fit inside the the second image. The window in the second image around the control points is distorted, too.
	dest	the computed normalized cross correlation coefficient.

Returns:

true if apply successful or false otherwise.

bool lti::crossCorrelationCoefficient::apply	(	const matrix< float > &	src1,
		const matrix< float > &	src2,
		const pointList &	controlPoints,
		const fvector &	transf,
		float &	dest
	)			const

Computes the normalized cross correlation coefficient between two channels only considering a window located at each control point.

This is useful to exlude e.g. homogenous areas. Only positions inside both channels are considered. If the channels are recorded from different positions it is advantageous to supply a transformation vector. Please do no forget to define a transformation in the parameters. On default no transformation is executed. E.g. the control points and the points inside the window have the same coordinates in both channels.

Parameters:

	src1	channel with the first intensities.
	src2	channel with the second intensities.
	controlPoints	pointList with the control points.
	transf	fvector with the transform by which the control points are fit inside the the second image. The window in the second image around the control points is distorted, too.
	dest	the computed normalized cross correlation coefficient.

Returns:

true if apply successful or false otherwise.

bool lti::crossCorrelationCoefficient::apply	(	const matrix< ubyte > &	src1,
		const matrix< ubyte > &	src2,
		const pointList &	controlPoints,
		const fvector &	transf,
		float &	dest
	)			const

Computes the normalized cross correlation coefficient between two channels only considering a window located at each control point.

This is useful to exlude e.g. homogenous areas. Only positions inside both channels are considered. If the channels are recorded from different positions it is advantageous to supply a transformation vector. Please do no forget to define a transformation in the parameters. On default no transformation is executed. E.g. the control points and the points inside the window have the same coordinates in both channels.

Parameters:

	src1	channel8 with the first intensities.
	src2	channel8 with the second intensities.
	controlPoints	pointList with the control points.
	transf	fvector with the transform by which the control points are fit inside the the second image. The window in the second image around the control points is distorted, too.
	dest	the computed normalized cross correlation coefficient.

Returns:

true if apply successful or false otherwise.

virtual functor* lti::crossCorrelationCoefficient::clone

(

)

const [virtual]

Returns a pointer to a clone of this functor.

Reimplemented from lti::featureExtractor.

crossCorrelationCoefficient& lti::crossCorrelationCoefficient::copy

(

const crossCorrelationCoefficient &

other

)

Copy data of "other" functor.

Parameters:

other

the functor to be copied

Returns:

a reference to this functor object

Reimplemented from lti::featureExtractor.

const parameters& lti::crossCorrelationCoefficient::getParameters

(

)

const

Returns used parameters.

Reimplemented from lti::featureExtractor.

virtual const char* lti::crossCorrelationCoefficient::getTypeName

(

)

const [virtual]

Returns the name of this type ("crossCorrelationCoefficient").

Reimplemented from lti::featureExtractor.

crossCorrelationCoefficient& lti::crossCorrelationCoefficient::operator=

(

const crossCorrelationCoefficient &

other

)

Alias for copy member.

Parameters:

other

the functor to be copied

Returns:

a reference to this functor object

Reimplemented from lti::functor.

---

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

• ltiCrossCorrelationCoefficient.h