lti::featureSaliencyAx Class Reference

Feature saliency map for AXIOM. More...

#include <ltiFeatureSaliencyAx.h>

Inheritance diagram for lti::featureSaliencyAx:

[legend]

Collaboration diagram for lti::featureSaliencyAx:

[legend]

List of all members.

Classes
class	parameters
	the parameters for the class featureSaliencyAx More...
Public Member Functions
	featureSaliencyAx ()
	featureSaliencyAx (const featureSaliencyAx &other)
virtual	~featureSaliencyAx ()
virtual const char *	getTypeName () const
bool	apply (const image &src, channel &map)
bool	apply (const image &src, channel &conspIntensity, channel &conspColor, channel &conspOrientation, channel &map)
const std::list< int > &	getResolutions () const
featureSaliencyAx &	copy (const featureSaliencyAx &other)
virtual functor *	clone () const
const parameters &	getParameters () const
void	normalize (channel &chnl) const
Protected Member Functions
void	intensityConsp (const channel &intensity, channel &iconsp) const
void	colorConsp (const channel &red, const channel &green, const channel &blue, channel &cconsp) const
void	orientationConsp (const channel &intensity, channel &oconsp) const
void	squarePyramid (const channel &src, pyramid< channel > &pyram) const
void	squarePyramid0 (const channel &src, pyramid< channel > &pyram) const
void	squarePyramid1 (const channel &src, pyramid< channel > &pyram) const
Protected Attributes
std::list< int >	resolutions
int	resWithHighestKullback

---

Detailed Description

Feature saliency map for AXIOM.

The feature saliency map generate a channel, where the pixel values represent "saliency" factors, i.e. how conspicous is the pixel respect its neighbourhood.

This implementation has the same principles of the Itti and Koch saliency map, but the implementation is different, to allow a faster calculation of the conspicous regions, interesting for the object recognition.

The name Ax is related with the Project AXIOM at the Chair of Technical Computer Science (LTI), for which this functor was developed.

---

Constructor & Destructor Documentation

lti::featureSaliencyAx::featureSaliencyAx

(

)

default constructor

lti::featureSaliencyAx::featureSaliencyAx

(

const featureSaliencyAx &

other

)

copy constructor

Parameters:

other

the object to be copied

virtual lti::featureSaliencyAx::~featureSaliencyAx

(

)

[virtual]

destructor

---

Member Function Documentation

bool lti::featureSaliencyAx::apply	(	const image &	src,
		channel &	conspIntensity,
		channel &	conspColor,
		channel &	conspOrientation,
		channel &	map
	)

generate the saliency map for the given image

Parameters:

	src	image with the source data.
	conspIntensity	the conspicuity map based on intensity
	conspColor	conspicuity map based on color
	conspOrientation	conspicuity map based on orientation
	map	the resulting saliency map

Returns:

true if successful, false otherwise.

bool lti::featureSaliencyAx::apply	(	const image &	src,
		channel &	map
	)

generate the saliency map for the given image

Parameters:

	src	image with the source data.
	map	the resulting saliency map

Returns:

true if successful, false otherwise.

virtual functor* lti::featureSaliencyAx::clone

(

)

const [virtual]

returns a pointer to a clone of this functor.

Implements lti::functor.

void lti::featureSaliencyAx::colorConsp	(	const channel &	red,
		const channel &	green,
		const channel &	blue,
		channel &	cconsp
	)			const [protected]

generate color conspicuity map

featureSaliencyAx& lti::featureSaliencyAx::copy

(

const featureSaliencyAx &

other

)

copy data of "other" functor.

Parameters:

other

the functor to be copied

Returns:

a reference to this functor object

Reimplemented from lti::functor.

const parameters& lti::featureSaliencyAx::getParameters

(

)

const

returns used parameters

Reimplemented from lti::modifier.

const std::list<int>& lti::featureSaliencyAx::getResolutions

(

)

const

get the most important resolutions used in the last apply method

virtual const char* lti::featureSaliencyAx::getTypeName

(

)

const [virtual]

returns the name of this type ("featureSaliencyAx")

Reimplemented from lti::modifier.

void lti::featureSaliencyAx::intensityConsp	(	const channel &	intensity,
		channel &	iconsp
	)			const [protected]

generate conspicuity map for intensity contrast This method uses edge detection for the saliency.

void lti::featureSaliencyAx::normalize

(

channel &

chnl

)

const

return a mask with the local maxima of a channel

void lti::featureSaliencyAx::orientationConsp	(	const channel &	intensity,
		channel &	oconsp
	)			const [protected]

generate orientation conspicuity map

void lti::featureSaliencyAx::squarePyramid	(	const channel &	src,
		pyramid< channel > &	pyram
	)			const [protected]

generate a pyramid using square filters with the resolutions i specified in the resolutions-list and the next ones (i and i+1)

void lti::featureSaliencyAx::squarePyramid0	(	const channel &	src,
		pyramid< channel > &	pyram
	)			const [protected]

generate a pyramid using square filters with only the resolutions in the list

void lti::featureSaliencyAx::squarePyramid1	(	const channel &	src,
		pyramid< channel > &	pyram
	)			const [protected]

generate a pyramid using square filters with only the resolutions following the ones in the list.

---

Member Data Documentation

std::list<int> lti::featureSaliencyAx::resolutions [protected]

list with the "most important" resolutions for the given image.

int lti::featureSaliencyAx::resWithHighestKullback [protected]

resolution with the highest kullback

---

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

• ltiFeatureSaliencyAx.h