cvr::usePalette Class Reference
[Region analysis]

Use color or gray-valued palette. More...

#include <cvrUsePalette.h>

Inheritance diagram for cvr::usePalette:

[legend]

Collaboration diagram for cvr::usePalette:

[legend]

List of all members.

Classes
class	parameters
	the parameters for the class usePalette More...
Public Member Functions
	usePalette ()
	usePalette (const usePalette &other)
virtual	~usePalette ()
virtual const std::string &	name () const
bool	apply (const matrix< ubyte > &chnl, image &other) const
bool	apply (const matrix< int > &chnl, image &other) const
bool	apply (const matrix< ubyte > &chnl, const palette &thePalette, image &other) const
bool	apply (const matrix< int > &chnl, const palette &thePalette, image &other) const
bool	apply (const matrix< ubyte > &regions, const vector< float > &values, fmatrix &dest) const
bool	apply (const matrix< int > &regions, const vector< float > &values, fmatrix &dest) const
bool	apply (const image &img, matrix< ubyte > &chnl)
bool	apply (const image &img, matrix< int > &chnl)
bool	apply (const image &img, const palette &thePalette, matrix< ubyte > &chnl)
bool	apply (const image &img, const palette &thePalette, matrix< int > &chnl)
bool	apply (const image &img, const kdTree< rgbaPixel, int > &tree, matrix< ubyte > &chnl) const
bool	apply (const image &img, const kdTree< rgbaPixel, int > &tree, matrix< int > &chnl) const
const kdTree< rgbaPixel, int > &	getKdTree () const
usePalette &	copy (const usePalette &other)
virtual usePalette *	clone () const
virtual usePalette *	newInstance () const
const parameters &	getParameters () const
virtual bool	updateParameters ()
Protected Member Functions
bool	buildKdTree (const palette &pal)
Protected Attributes
kdTree< rgbaPixel, int >	tree_
bool	treeBuilded_

---

Detailed Description

Use color or gray-valued palette.

This class has two uses:

• For a given labeled mask (which is usually an cvr::imatrix or a cvr::channel8) and a palette, replace in a destination image each region label by the corresponding palette entry. This is a simple but frequent used operation when dealing with segmentation results, color quantization results, and other image partitioning operators.

• For a given image and a palette, find which palette entry in the image is the most similar to each corresponding pixel, and generate an index image (cvr::imatrix or cvr::channel8) containing those indices.

For the second operation mode you can choose in the parameters to use a cvr::kdTree in order to avoid a "brute-force" search.

See also:

cvr::computePalette

---

Constructor & Destructor Documentation

cvr::usePalette::usePalette

(

)

default constructor

cvr::usePalette::usePalette

(

const usePalette &

other

)

copy constructor

Parameters:

other

the object to be copied

virtual cvr::usePalette::~usePalette

(

)

[virtual]

destructor

---

Member Function Documentation

bool cvr::usePalette::apply	(	const image &	img,
		const kdTree< rgbaPixel, int > &	tree,
		matrix< int > &	chnl
	)			const

Find for each pixel in the given image the nearest entry in the given k-d Tree and leave the data component of that entry in the correspondig pixel of the matrix<ubyte>.

Note that with this method the parameters::colors attribute or previously computed k-d trees will be ignored.

Parameters:

	img	the color image (true color)
	tree	k-d Tree coding a color palette. The index in the integer data component of the tree elements will be assumed to be the index of a color palette.
	chnl	the data of the nearest component for each pixel in img will be stored in the correponding pixel of this channel.

bool cvr::usePalette::apply	(	const image &	img,
		const kdTree< rgbaPixel, int > &	tree,
		matrix< ubyte > &	chnl
	)			const

Find for each pixel in the given image the nearest entry in the given k-d Tree and leave the data component of that entry in the correspondig pixel of the matrix<ubyte>.

Note that with this method the parameters::colors attribute or previously computed k-d trees will be ignored.

Parameters:

	img	the color image (true color)
	tree	k-d Tree coding a color palette. The index in the integer data component of the tree elements will be assumed to be the index of a color palette.
	chnl	the data of the nearest component for each pixel in img will be stored in the correponding pixel of this channel.

bool cvr::usePalette::apply	(	const image &	img,
		const palette &	thePalette,
		matrix< int > &	chnl
	)

Find for each pixel in the given image the best entry of the given palette and leave the index of that entry in the correspondig pixel of the matrix<int>.

Note that with this method the parameters::colors attribute will be ignored.

Parameters:

	img	the color image (true color)
	thePalette	a color palette. The index of the most similar color entry within this palette to each pixel of the img will be left in the chnl
	chnl	the indices for each pixel in img of the correponding palette entry in thePalette

bool cvr::usePalette::apply	(	const image &	img,
		const palette &	thePalette,
		matrix< ubyte > &	chnl
	)

Find for each pixel in the given image the best entry of the given palette and leave the index of that entry in the correspondig pixel of the matrix<ubyte>.

Note that with this method the parameters::colors attribute will be ignored. The internal k-d Tree will be computed using thePalette given here.

Parameters:

	img	the color image (true color)
	thePalette	a color palette. The index of the most similar color entry within this palette to each pixel of the img will be left in the chnl
	chnl	the indices for each pixel in img of the correponding palette entry in thePalette

bool cvr::usePalette::apply	(	const image &	img,
		matrix< int > &	chnl
	)

Find for each pixel in the given image the best entry in the palette found in the parameters and leave the index of the entry in the correspondig pixel of the matrix<int>.

Parameters:

	img	the color image (true color)
	chnl	the indices for each pixel in img of the correponding palette entry in the parameters object

bool cvr::usePalette::apply	(	const image &	img,
		matrix< ubyte > &	chnl
	)

For each pixel in the given image, find the "best" entry in the palette found in the parameters and leave the index of the entry in the correspondig pixel of the matrix<ubyte>.

Parameters:

	img	the color image (true color)
	chnl	the indices for each pixel in img of the correponding palette entry in the parameters object

bool cvr::usePalette::apply	(	const matrix< int > &	regions,
		const vector< float > &	values,
		fmatrix &	dest
	)			const

Use gray valued palette.

This method is provided for completeness. It is not for color images, but for gray valued ones, but since the principle used is the same than with colored palettes, the method seems to be in a proper place here. It complements the corresponding apply method of cvr::computePalette to compute gray-valued palettes from channels.

Parameters:

	regions	regions map, containing for each region a palette index entry.
	values	gray-valued palette entries. For each region with index i, the output channel will contain values.at(i).
	dest	output channel, the result of using the gray-valued palette for the given regions map will be left here.

Returns:

true if successful, false otherwise.

bool cvr::usePalette::apply	(	const matrix< ubyte > &	regions,
		const vector< float > &	values,
		fmatrix &	dest
	)			const

Use gray valued palette.

This method is provided for completeness. It is not for color images, but for gray valued ones, but since the principle used is the same than with colored palettes, the method seems to be in a proper place here. It complements the corresponding apply method of cvr::computePalette to compute gray-valued palettes from channels.

Parameters:

	regions	regions map, containing for each region a palette index entry.
	values	gray-valued palette entries. For each region with index i, the output channel will contain values.at(i).
	dest	output channel, the result of using the gray-valued palette for the given regions map will be left here.

Returns:

true if successful, false otherwise.

bool cvr::usePalette::apply	(	const matrix< int > &	chnl,
		const palette &	thePalette,
		image &	other
	)			const

Takes the matrix<int> and uses its elements as index for the given palette.

The result will be left in a new image.

Parameters:

	chnl	the matrix<int> with the indices for the palette
	thePalette	palette to be used
	other	the resulting image

Returns:

true if successful, false otherwise

bool cvr::usePalette::apply	(	const matrix< ubyte > &	chnl,
		const palette &	thePalette,
		image &	other
	)			const

Takes the matrix<ubyte> and uses its elements as index for the given palette.

The result will be left in a new image.

Parameters:

	chnl	the matrix<ubyte> with the indices for the palette
	thePalette	palette to be used
	other	the resulting image

Returns:

true if successful, false otherwise

bool cvr::usePalette::apply	(	const matrix< int > &	chnl,
		image &	other
	)			const

Takes the matrix<int> and uses its elements as index for the palette in the parameters.

The result will be left in a new image.

Parameters:

	chnl	the matrix<int> with the indices for the palette
	other	the resulting image

Returns:

true if successful, false otherwise

bool cvr::usePalette::apply	(	const matrix< ubyte > &	chnl,
		image &	other
	)			const

Takes the matrix<ubyte> and uses its elements as index for the palette in the parameters.

The result will be left in a new image.

Parameters:

	chnl	the matrix<ubyte> with the indices for the palette
	other	the resulting image

Returns:

true if successful, false otherwise

bool cvr::usePalette::buildKdTree

(

const palette &

pal

)

[protected]

Build the k-d Tree from the given color palette.

virtual usePalette* cvr::usePalette::clone

(

)

const [virtual]

Returns a pointer to a clone of this functor.

Implements cvr::functor.

usePalette& cvr::usePalette::copy

(

const usePalette &

other

)

Copy data of "other" functor.

Parameters:

other

the functor to be copied

Returns:

a reference to this functor object

const kdTree<rgbaPixel,int>& cvr::usePalette::getKdTree

(

)

const

Get a constant reference to the internal k-d tree.

const parameters& cvr::usePalette::getParameters

(

)

const

Returns used parameters.

Reimplemented from cvr::parametersManager.

virtual const std::string& cvr::usePalette::name

(

)

const [virtual]

returns the name of this type

Implements cvr::functor.

virtual usePalette* cvr::usePalette::newInstance

(

)

const [virtual]

Returns a pointer to a new instance of this functor.

Implements cvr::functor.

virtual bool cvr::usePalette::updateParameters

(

)

[virtual]

Set functor's parameters.

This member makes a copy of theParam: the functor will keep its own copy of the parameters!

Returns:

true if successful, false otherwise

Reimplemented from cvr::parametersManager.

---

Member Data Documentation

kdTree<rgbaPixel,int> cvr::usePalette::tree_ [protected]

The parameters palette is converted (if desired) into this k-d Tree.

The "data" type int states for the entry index.

bool cvr::usePalette::treeBuilded_ [protected]

Flag to indicate if the k-d Tree has already been build or not.

---

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

• cvrUsePalette.h