A data class to hold a grid of local orientations: for each pixel, a local orientation and magnitude are kept from this representation, it should be fairly easy to implement the Hough transform for different parametric models like straight lines and circles: loop over all possible values of the parameter sets for the lines to check a) based on parameters, generate parametric representation of candidate line (circle, curve) b) for all pixels along this line, check if the orientation map (direction/strength) supports the presence of this candidate if yes, add 1 to accumulator for the given parameters. More...

#include <OrientationGrid.h>

Public Types
enum	ColorExportMode { EXPORT_RGB_MODE, EXPORT_HSV_MODE }
	enum that describes color mapping from angle to color More...

Public Member Functions
	OrientationGrid (int width, int height)
	Constructor. More...

	~OrientationGrid ()
	Destructor.

OrientationGrid *	Clone ()

int	GetWidth ()
	Gets the width of the grid.

int	GetHeight ()
	Gets the height of the grid.

void	SetAngle (int x, int y, double myAngle)
	Sets the local angle at specified position of the grid. More...

double	GetAngle (int x, int y)
	Gets the local angle at specified position of the grid. More...

void	SetOrientation (int x, int y, LocalOrientation *pLocalOrientation)
	Sets the local orientation (angle,magnitude) at specified position of the grid. More...

LocalOrientation *	GetOrientation (int x, int y)
	Gets the local orientation (angle,magnitude) at specified position of the grid. More...

void	SetMagnitude (int x, int y, double myMagnitude)
	Sets the magnitude of the feature response at position (x, y) More...

double	GetMagnitude (int x, int y)
	Gets the magnitude of the feature response at position (x, y) More...

std::vector< LocalOrientation >	GetOrientationVector (int xMin, int yMin, int xMax, int yMax)
	Gets a vector with the orientations within the specified ROI. More...

bool	GetMinMaxMagnitude (double &mmin, double &mmax)
	Gets the minimum and maximum magnitude of the features in the grid. More...

ArrayGrid< double > *	ExtractMagnitudeGrid ()
	extracts a grid with only edge magnitudes Becomes responsability of the caller to free this newly created object

Image *	VisualizeMagnitudeImage ()
	creates current magnitude image ready for exporting to disk or viewing on screen

ArrayGrid< double > *	ExtractOrientationGrid ()
	extracts a grid with only edge orientations Becomes responsability of the caller to free this newly created object

Image *	VisualizeOrientationImage (double threshold, ColorExportMode myMode=EXPORT_HSV_MODE)
	creates current orientation image ready for exporting to disk or viewing on screen More...

Detailed Description

A data class to hold a grid of local orientations: for each pixel, a local orientation and magnitude are kept from this representation, it should be fairly easy to implement the Hough transform for different parametric models like straight lines and circles: loop over all possible values of the parameter sets for the lines to check a) based on parameters, generate parametric representation of candidate line (circle, curve) b) for all pixels along this line, check if the orientation map (direction/strength) supports the presence of this candidate if yes, add 1 to accumulator for the given parameters.

Member Enumeration Documentation

enum stira::imagedata::OrientationGrid::ColorExportMode

enum that describes color mapping from angle to color

Enumerator
EXPORT_RGB_MODE	linear interpolation in R, G and B (see ConvertAngleToRGBMap for details)
EXPORT_HSV_MODE	angle mapped to H, (S and V are both set to 1). HSV is then converted to RGB

Constructor & Destructor Documentation

stira::imagedata::OrientationGrid::OrientationGrid	(	int	width,
		int	height
	)

Constructor.

Parameters

width	width of the grid
height	height of the grid

Member Function Documentation

double stira::imagedata::OrientationGrid::GetAngle	(	int	x,
		int	y
	)

Gets the local angle at specified position of the grid.

Parameters

x	the x coordinate
y	the y coordinate

Referenced by stira::steerable::ComputeSteerableFilteredImages::FilterAlongDominantOrientationG(), stira::steerable::ComputeSteerableFilteredImages::FilterAlongDominantOrientationH(), and stira::imageanalysis::CannyEdgeDetector::~CannyEdgeDetector().

double stira::imagedata::OrientationGrid::GetMagnitude	(	int	x,
		int	y
	)

Gets the magnitude of the feature response at position (x, y)

Parameters

x	the x coordinate
y	the y coordinate

References stira::imagedata::ColorValue::c, stira::common::MathUtils::ToDegrees(), and stira::imagedata::ColorValue::type.

Referenced by stira::imageanalysis::CannyEdgeDetector::~CannyEdgeDetector().

bool stira::imagedata::OrientationGrid::GetMinMaxMagnitude	(	double &	mmin,
		double &	mmax
	)

Gets the minimum and maximum magnitude of the features in the grid.

Parameters

mmin	(output) the minimum magnitude
mmax	(output) the maximum magnitude

LocalOrientation * stira::imagedata::OrientationGrid::GetOrientation	(	int	x,
		int	y
	)

Gets the local orientation (angle,magnitude) at specified position of the grid.

Parameters

x	the x coordinate
y	the y coordinate

std::vector< LocalOrientation > stira::imagedata::OrientationGrid::GetOrientationVector	(	int	xMin,
		int	yMin,
		int	xMax,
		int	yMax
	)

Gets a vector with the orientations within the specified ROI.

Parameters

xMin	minimum x coordinate of the ROI
yMin	minimum y coordinate of the ROI
xMax	maximum x coordinate of the ROI
yMax	maximum y coordinate of the ROI

Referenced by stira::imageanalysis::HOG::HOG().

void stira::imagedata::OrientationGrid::SetAngle	(	int	x,
		int	y,
		double	myAngle
	)

Sets the local angle at specified position of the grid.

Parameters

x	the x coordinate
y	the y coordinate
myAngle	the angle to set for position (x,y)

Referenced by stira::steerable::ComputeOrientation::ComputeDominantOrientationInputComplexPyramidLevel(), stira::steerable::ComputeOrientation::ComputeDominantOrientationInputFreemanAdelson(), stira::filter::NonSeparableFilter::DetermineLocalMagnitudeAndOrientation(), stira::imageanalysis::HOG::HOG(), stira::imageanalysis::StegerLineDetector::Run(), stira::imageanalysis::CannyEdgeDetector::~CannyEdgeDetector(), and ~OrientationGrid().

void stira::imagedata::OrientationGrid::SetMagnitude	(	int	x,
		int	y,
		double	myMagnitude
	)

Sets the magnitude of the feature response at position (x, y)

Parameters

x	the x coordinate
y	the y coordinate
myMagnitude	the magnitude to set for position (x,y)

Referenced by stira::steerable::ComputeOrientation::ComputeDominantOrientationInputComplexPyramidLevel(), stira::steerable::ComputeOrientation::ComputeDominantOrientationInputFreemanAdelson(), stira::filter::NonSeparableFilter::DetermineLocalMagnitudeAndOrientation(), stira::imageanalysis::HOG::HOG(), stira::imageanalysis::StegerLineDetector::Run(), and ~OrientationGrid().

void stira::imagedata::OrientationGrid::SetOrientation	(	int	x,
		int	y,
		LocalOrientation *	pLocalOrientation
	)

Sets the local orientation (angle,magnitude) at specified position of the grid.

Parameters

x	the x coordinate
y	the y coordinate
myLocalOrientation	the local orientation to set for position (x,y)

Image * stira::imagedata::OrientationGrid::VisualizeOrientationImage	(	double	threshold,
		ColorExportMode	myMode = `EXPORT_HSV_MODE`
	)

creates current orientation image ready for exporting to disk or viewing on screen

Parameters

threshold	threshold value for magnitude (below this, we just put black instead of coded orientation)
myMode	color mapping scheme to use

References stira::imagedata::ColorValue::c, stira::imagedata::Image::GetBands(), and stira::imagedata::Image::SetImageName().

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

imagedata/datastructures/OrientationGrid.h
imagedata/datastructures/OrientationGrid.cpp

Public Types

Public Member Functions

Detailed Description

Member Enumeration Documentation

Constructor & Destructor Documentation

Member Function Documentation