lti::grayWorldNormalization Class Reference
[Color Analysis]

Performs a color normalization on an lti::image using a gray world approach, in order to eliminate effects of illumination colour. More...

#include <ltiGrayWorldNormalization.h>

Inheritance diagram for lti::grayWorldNormalization:

[legend]

Collaboration diagram for lti::grayWorldNormalization:

[legend]

List of all members.

Classes
class	parameters
	The parameters for the class grayWorldNormalization. More...
Public Member Functions
	grayWorldNormalization ()
	grayWorldNormalization (const parameters &par)
	grayWorldNormalization (const grayWorldNormalization &other)
virtual	~grayWorldNormalization ()
virtual const char *	getTypeName () const
bool	apply (image &srcdest) const
bool	apply (const image &src, image &dest) const
bool	apply (const channel &srcR, const channel &srcG, const channel &srcB, image &dest) const
bool	apply (const image &src, channel &chR, channel &chG, channel &chB) const
bool	apply (const channel &srcR, const channel &srcG, const channel &srcB, channel &chR, channel &chG, channel &chB) const
bool	apply (channel &srcdestR, channel &srcdestG, channel &srcdestB) const
grayWorldNormalization &	copy (const grayWorldNormalization &other)
virtual functor *	clone () const
const parameters &	getParameters () const

Detailed Description

Performs a color normalization on an lti::image using a gray world approach, in order to eliminate effects of illumination colour.

Gray world methods and white world methods assume that some low order statistics on the colors of a real world scene lie on the gray line of a color space. The most simple approach uses only the mean color and maps it linearly into one specific point on the gray line, assuming that a diagonal transform (a simple channel scaling) is valid.

Normalization is done under the assumptiomn of mondrian worlds, constant point light sources and delta function camera sensors.

For each color channel $c\in{r,g,b}$ , the normalization will be

$c_n \rightarrow \frac{c} {(1-\gamma) + \gamma(\nu \cdot c_\mu + \kappa \cdot c_\sigma)} *$

With:

$\gamma$ : normalization strength between 0 (no normalization) and 1 (complete normalization).
$c_\mu$ : mean value of the c channel.
$c_\sigma$ : standard deviation of the c channel.
$\nu$ : mean value scaling factor.
$\kappa$ : standard deviation scaling factor.

Initial theory: Diploma Thesis, Dipl.-Ing. Jose Pablo Alvarado Moya, 1998 Dept. of Technical Computer Science, RWTH Aachen, Germany.

Constructor & Destructor Documentation

lti::grayWorldNormalization::grayWorldNormalization ( )

default constructor

lti::grayWorldNormalization::grayWorldNormalization ( const parameters & par )

default constructor

lti::grayWorldNormalization::grayWorldNormalization ( const grayWorldNormalization & other )

copy constructor

Parameters:

other

the object to be copied

virtual lti::grayWorldNormalization::~grayWorldNormalization ( ) [virtual]

destructor

Member Function Documentation

bool lti::grayWorldNormalization::apply	(	channel &	srcdestR,
		channel &	srcdestG,
		channel &	srcdestB
	)			const

operates on the given parameter.

Parameters:

	srcdestR	red channel of the source and destination.
	srcdestG	green channel of the source and destination.
	srcdestB	blue channel of the source and destination.

Returns:: true if successful, false otherwise.

bool lti::grayWorldNormalization::apply	(	const channel &	srcR,
		const channel &	srcG,
		const channel &	srcB,
		channel &	chR,
		channel &	chG,
		channel &	chB
	)			const

operates on a copy of the given parameters.

Parameters:

	srcR	red channel of the source data
	srcG	green channel of the source data
	srcB	blue channel of the source data
	chR	red component of the normalized input image.
	chG	green component of the normalized input image.
	chB	blue component of the normalized input image.

Returns:: true if successful, false otherwise.

bool lti::grayWorldNormalization::apply	(	const image &	src,
		channel &	chR,
		channel &	chG,
		channel &	chB
	)			const

operates on a copy of the given parameters.

Parameters:

	src	image with the source data.
	chR	red component of the normalized input image.
	chG	green component of the normalized input image.
	chB	blue component of the normalized input image.

Returns:: true if successful, false otherwise.

bool lti::grayWorldNormalization::apply	(	const channel &	srcR,
		const channel &	srcG,
		const channel &	srcB,
		image &	dest
	)			const

operates on a copy of the given parameters.

Parameters:

	srcR	red channel of the source data
	srcG	green channel of the source data
	srcB	blue channel of the source data
	dest	image where the result will be left.