#include <ObjectEdgeDetector.h>

Public Types
enum	Mode { BEST, GPU, CPU }

Public Member Functions
	ObjectEdgeDetector (Mode mode=BEST)
	Create new ObjectEdgeDetector. More...

virtual	~ObjectEdgeDetector ()
	Destructor. More...

const core::Img8u &	calculate (const core::Img32f &depthImage, bool filter, bool average, bool gauss)
	One call function for calculation of the complete processingpipeline. More...

void	setDepthImage (const core::Img32f &depthImg)
	Sets the input depth image (input for median filter). More...

void	applyMedianFilter ()
	Calculates a filtered image using a median filter. More...

const core::Img32f &	getFilteredDepthImage ()
	Returns the filtered depth image. More...

void	setFilteredDepthImage (const core::Img32f &filteredImg)
	Sets the (filtered) depth image (input for normal calculation) More...

void	applyNormalCalculation ()
	Calculates the point normals. More...

void	applyLinearNormalAveraging ()
	Recalculates the normals by linear averaging. More...

void	applyGaussianNormalSmoothing ()
	Recalculates the normals by gaussian smoothing in a given range. More...

const core::DataSegment< float, 4 >	getNormals ()
	Returns the Pointer to the normals. More...

void	applyWorldNormalCalculation (const Camera &cam)
	Transforms the normals to the world space and calculates normal image. More...

const core::DataSegment< float, 4 >	getWorldNormals ()
	Returns the point normals in world space. More...

const core::Img8u &	getRGBNormalImage ()
	Returns the RGB normal image. More...

void	setNormals (core::DataSegment< float, 4 > pNormals)
	Sets the point normals (input for angle image calculation). More...

void	applyAngleImageCalculation ()
	Calculates the angle image. More...

const core::Img32f &	getAngleImage ()
	Returns the angle image. More...

void	setAngleImage (const core::Img32f &angleImg)
	Sets the angle image (input for image binarization). More...

void	applyImageBinarization ()
	Binarizes the angle image to detect edges. More...

const core::Img8u &	getBinarizedAngleImage ()
	Returns the binarized angle image (final output). More...

void	setMedianFilterSize (int size)
	Sets the mask size for applyMedianFilter() More...

void	setNormalCalculationRange (int range)
	Sets the range for applyNormalCalculation(). More...

void	setNormalAveragingRange (int range)
	Sets the averaging range for applyNormalAveraging() More...

void	setAngleNeighborhoodMode (int mode)
	Sets the neighborhood mode for applyAngleImageCalculation() More...

void	setAngleNeighborhoodRange (int range)
	Sets the neighborhood range for applyAngleImageCalculation() More...

void	setBinarizationThreshold (float threshold)
	Sets the binarization threshold for applyImageBinarization(). More...

void	setUseNormalAveraging (bool use)
	Sets normal averaging enabled/disabled. More...

void	setUseGaussSmoothing (bool use)
	Sets normal averaging by gauss smoothing enabled/disabled. More...

Private Member Functions
void	initialize (utils::Size size)

Private Attributes
Data *	m_data
	internal data type More...

ObjectEdgeDetectorPlugin *	objectEdgeDetector

Detailed Description

This class calculates an edge image based on angles between normals from an input depth image (e.g. Kinect). The common way to use this class is the calculate() method, getting a depth image and returning an edge image. This method computes the whole pipeline (image filtering, normal calculation and smoothing, angle image calculation and binarization). The performance of this method is optimized with minimal read/write for the underlying OpenCL implementation. The interim results can be accessed with getNormals() and getAngleImage() afterwards. It is also possible to use subparts of the pipeline using the setter methods to set the interim data.

Member Enumeration Documentation

◆ Mode

enum icl::geom::ObjectEdgeDetector::Mode

Enumerator
BEST
GPU
CPU

Constructor & Destructor Documentation

◆ ObjectEdgeDetector()

icl::geom::ObjectEdgeDetector::ObjectEdgeDetector ( Mode mode = BEST )

Create new ObjectEdgeDetector.

Constructs an object of this class. All default parameters are set. Use setters for desired values.

Parameters

mode	selects the implementation (GPU, CPU or BEST)

◆ ~ObjectEdgeDetector()

virtual icl::geom::ObjectEdgeDetector::~ObjectEdgeDetector ( )

virtual

Destructor.

Member Function Documentation

◆ applyAngleImageCalculation()

void icl::geom::ObjectEdgeDetector::applyAngleImageCalculation ( )

Calculates the angle image.

The mode is set by setAngleNeighborhoodMode(int mode). The range is set by setAngleNeighborhoodRange(int range)

◆ applyGaussianNormalSmoothing()

void icl::geom::ObjectEdgeDetector::applyGaussianNormalSmoothing ( )

Recalculates the normals by gaussian smoothing in a given range.

Alternative to applyTemporalNormalAveraging that uses a simpler gaussian filter for smoothing

if it is enabled width setUseNormalAveraging(bool use). The range is set by setNormalAveragingRange(int range). Alternative: normalAveraging() with setUseGaussSmoothing(false).

◆ applyImageBinarization()

void icl::geom::ObjectEdgeDetector::applyImageBinarization ( )

Binarizes the angle image to detect edges.

The threshold is set by setBinarizationThreshold(float threshold).

◆ applyLinearNormalAveraging()

void icl::geom::ObjectEdgeDetector::applyLinearNormalAveraging ( )

Recalculates the normals by linear averaging.

This reduces the noise of the normal image. The function is called from calculateNormals() if it is enabled width setUseNormalAveraging(bool use). The range is set by setNormalAveragingRange(int range).
Alternative: normalGaussSmoothing() with setUseGaussSmoothing(true).

◆ applyMedianFilter()

void icl::geom::ObjectEdgeDetector::applyMedianFilter ( )

Calculates a filtered image using a median filter.

The mask size is set by setMedianFilterSize(int size)

◆ applyNormalCalculation()

void icl::geom::ObjectEdgeDetector::applyNormalCalculation ( )

Calculates the point normals.

The range for calculation is set by setNormalCalculationRange(int range).

◆ applyWorldNormalCalculation()

void icl::geom::ObjectEdgeDetector::applyWorldNormalCalculation ( const Camera & cam )

Transforms the normals to the world space and calculates normal image.

Parameters

cam	the camera of the depth image

◆ calculate()

const core::Img8u& icl::geom::ObjectEdgeDetector::calculate	(	const core::Img32f &	depthImage,
		bool	filter,
		bool	average,
		bool	gauss
	)

One call function for calculation of the complete processingpipeline.

Order: ((filter)->normals->(normalAvg)->angles->binarization)

Parameters

depthImage	the input depth image
filter	enable/disable filtering
average	enable/disable normal averaging
gauss	true=gauss smoothing, false=linear smoothing

Returns: the binarized angle image

◆ getAngleImage()

const core::Img32f& icl::geom::ObjectEdgeDetector::getAngleImage ( )

Returns the angle image.

Returns: the angle image

◆ getBinarizedAngleImage()

const core::Img8u& icl::geom::ObjectEdgeDetector::getBinarizedAngleImage ( )

Returns the binarized angle image (final output).

Returns: the (final) binarized angle image

◆ getFilteredDepthImage()

const core::Img32f& icl::geom::ObjectEdgeDetector::getFilteredDepthImage ( )

Returns the filtered depth image.

Returns: the filtered depth image

◆ getNormals()

const core::DataSegment<float,4> icl::geom::ObjectEdgeDetector::getNormals ( )

Returns the Pointer to the normals.

Returns: the point normals

◆ getRGBNormalImage()

const core::Img8u& icl::geom::ObjectEdgeDetector::getRGBNormalImage ( )

Returns the RGB normal image.

Returns: the RGB normal image

◆ getWorldNormals()

const core::DataSegment<float,4> icl::geom::ObjectEdgeDetector::getWorldNormals ( )

Returns the point normals in world space.

Returns: the point normals in world space

◆ initialize()

void icl::geom::ObjectEdgeDetector::initialize ( utils::Size size )

private

◆ setAngleImage()

void icl::geom::ObjectEdgeDetector::setAngleImage ( const core::Img32f & angleImg )

Sets the angle image (input for image binarization).

This call is not necessary if angleImageCalculation() is executed before.

Parameters

angleImg the angle image

◆ setAngleNeighborhoodMode()

void icl::geom::ObjectEdgeDetector::setAngleNeighborhoodMode ( int mode )

Sets the neighborhood mode for applyAngleImageCalculation()

0=max, 1=mean. (default 0)

Parameters

mode	the neighborhood mode

◆ setAngleNeighborhoodRange()

void icl::geom::ObjectEdgeDetector::setAngleNeighborhoodRange ( int range )

Sets the neighborhood range for applyAngleImageCalculation()

(default 3, min 1)

Parameters

range the neighborhood range

◆ setBinarizationThreshold()

void icl::geom::ObjectEdgeDetector::setBinarizationThreshold ( float threshold )

Sets the binarization threshold for applyImageBinarization().

Value n for acos(n). A value of 0 maps to 90 degree, a value of 1 maps to o degree (default 0.89)

Parameters

threshold binarization threshold

◆ setDepthImage()

void icl::geom::ObjectEdgeDetector::setDepthImage ( const core::Img32f & depthImg )

Sets the input depth image (input for median filter).

Parameters

depthImg the input depth image

◆ setFilteredDepthImage()

void icl::geom::ObjectEdgeDetector::setFilteredDepthImage ( const core::Img32f & filteredImg )

Sets the (filtered) depth image (input for normal calculation)

This call is not necessary if medianFilter() is executed before

Parameters

filteredImg the (filtered) depth image

◆ setMedianFilterSize()

void icl::geom::ObjectEdgeDetector::setMedianFilterSize ( int size )

Sets the mask size for applyMedianFilter()

size n corresponds to mask size n x n. (default 3, min 3, max 9, odd only)

Parameters

size	the mask size

◆ setNormalAveragingRange()

void icl::geom::ObjectEdgeDetector::setNormalAveragingRange ( int range )

Sets the averaging range for applyNormalAveraging()

(default 1)

Parameters

range the normal averaging range

◆ setNormalCalculationRange()

void icl::geom::ObjectEdgeDetector::setNormalCalculationRange ( int range )

Sets the range for applyNormalCalculation().

(default 2)

Parameters

range the normal calculation range

◆ setNormals()

void icl::geom::ObjectEdgeDetector::setNormals ( core::DataSegment< float, 4 > pNormals )

Sets the point normals (input for angle image calculation).

This call is not necessary if normalCalculation() is executed before.

Parameters

pNormals the point normals

◆ setUseGaussSmoothing()

void icl::geom::ObjectEdgeDetector::setUseGaussSmoothing ( bool use )

Sets normal averaging by gauss smoothing enabled/disabled.

(default false=linear smoothing if normal averaging enabled)

Parameters

use	enable/disable gauss smoothing

◆ setUseNormalAveraging()

void icl::geom::ObjectEdgeDetector::setUseNormalAveraging ( bool use )

Sets normal averaging enabled/disabled.

(default true=enabled)

Parameters

use	enable/disable normal averaging

Member Data Documentation

◆ m_data

Data* icl::geom::ObjectEdgeDetector::m_data

private

internal data type

internal data pointer

◆ objectEdgeDetector

ObjectEdgeDetectorPlugin* icl::geom::ObjectEdgeDetector::objectEdgeDetector

private

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

/Users/alneuman/vm/icl/ICLGeom/src/ICLGeom/ObjectEdgeDetector.h

Public Types

Public Member Functions

Private Member Functions

Private Attributes

Detailed Description

Member Enumeration Documentation

◆ Mode

Constructor & Destructor Documentation

◆ ObjectEdgeDetector()

◆ ~ObjectEdgeDetector()

Member Function Documentation

◆ applyAngleImageCalculation()

◆ applyGaussianNormalSmoothing()

◆ applyImageBinarization()

◆ applyLinearNormalAveraging()

◆ applyMedianFilter()

◆ applyNormalCalculation()

◆ applyWorldNormalCalculation()

◆ calculate()

◆ getAngleImage()

◆ getBinarizedAngleImage()

◆ getFilteredDepthImage()

◆ getNormals()

◆ getRGBNormalImage()

◆ getWorldNormals()

◆ initialize()

◆ setAngleImage()

◆ setAngleNeighborhoodMode()

◆ setAngleNeighborhoodRange()

◆ setBinarizationThreshold()

◆ setDepthImage()

◆ setFilteredDepthImage()

◆ setMedianFilterSize()

◆ setNormalAveragingRange()

◆ setNormalCalculationRange()

◆ setNormals()

◆ setUseGaussSmoothing()

◆ setUseNormalAveraging()

Member Data Documentation

◆ m_data

◆ objectEdgeDetector