VisionProcessor Class Reference

The VisionProcessor class performs image processing to the sensor data in order to detect the hands and the fingertips. More...

#include <visionprocessor.h>

Collaboration diagram for VisionProcessor:

Public Member Functions
	VisionProcessor ()
virtual void	init (int wRes=1024, int hRes=768)
	init initializes the paramameters for the image processing algorithms More...
virtual void	getScreenTransformation ()
	getScreenTransformation segments and creates the transformation for going from sensor space to screen space More...
virtual void	createMaxDepthWithHistogram (int numFrames=200)
	createMaxDepthWithHistogram generates a noise robust background map More...
virtual void	processFrame (FrameData &frameData, double &processingDuration)
	processFrame waits for a frame and does the processing More...
cv::Size2i	getKinectResolution () const
ProjectiveMapping	getKinect2screen () const
void	setKinect2screen (const ProjectiveMapping &value)

Protected Member Functions
virtual int	segmentScreen (const Mat &grayFrame, int Ith, float epsilon, vector< Point2i > &screenContours, Mat &maskShow)

Protected Attributes
KinectHandler	kinect
Mat	mask
Mat	maxDepth
ProjectiveMapping	kinect2screen
float	errorTolerance
float	thicknessHand
float	thicknesFinger
int	sizes [2]
float	tolerance
int	epsilon
Mat	minDepthHand
Mat	minDepthFinger
float	minAreaContour
float	maxAreaContour
Mat	erodeK
bool	useScreenSeg
FourierDescriptor	fdDictionary
float	maxAreaFinger
float	minAreaFinger
float	maxExcentricity
float	minExcentricity
float	maxCirc
float	minHandExc
float	maxDist2Class
int	descriptorNum
int	wResolution
int	hResolution
vector< Point2f >	screenContoursCorners

Static Protected Attributes
static const int	HAND_CLASS_ID =1
static const double	PI = 3.14159265358979323846

Detailed Description

The VisionProcessor class performs image processing to the sensor data in order to detect the hands and the fingertips.

Constructor & Destructor Documentation

VisionProcessor::VisionProcessor

(

)

Member Function Documentation

void VisionProcessor::createMaxDepthWithHistogram ( int  numFrames = 200 )

virtual

createMaxDepthWithHistogram generates a noise robust background map

Parameters

numFrames

number of frames to generate the histogram, 200 for default

ProjectiveMapping VisionProcessor::getKinect2screen

(

)

const

Size2i VisionProcessor::getKinectResolution

(

)

const

void VisionProcessor::getScreenTransformation ( )

virtual

getScreenTransformation segments and creates the transformation for going from sensor space to screen space

void VisionProcessor::init ( int  wRes = 1024, int  hRes = 768  )

virtual

init initializes the paramameters for the image processing algorithms

Parameters

wRes	is the width of the resolution the screen is working on, 1024 for default
hRes	is the height of the resolution the screen is working on, 768 for default

void VisionProcessor::processFrame ( FrameData &  frameData, double &  processingDuration  )

virtual

processFrame waits for a frame and does the processing

Parameters

frameData	input/output variable with the frame data processed
processingDuration	output variable with processing time needed in seconds

int VisionProcessor::segmentScreen ( const Mat &  grayFrame, int  Ith, float  epsilon, vector< Point2i > &  screenContours, Mat &  maskShow  )

protectedvirtual

void VisionProcessor::setKinect2screen

(

const ProjectiveMapping &

value

)

Member Data Documentation

int VisionProcessor::descriptorNum

protected

int VisionProcessor::epsilon

protected

Mat VisionProcessor::erodeK

protected

float VisionProcessor::errorTolerance

protected

FourierDescriptor VisionProcessor::fdDictionary

protected

const int VisionProcessor::HAND_CLASS_ID =1

staticprotected

int VisionProcessor::hResolution

protected

KinectHandler VisionProcessor::kinect

protected

ProjectiveMapping VisionProcessor::kinect2screen

protected

Mat VisionProcessor::mask

protected

float VisionProcessor::maxAreaContour

protected

float VisionProcessor::maxAreaFinger

protected

float VisionProcessor::maxCirc

protected

Mat VisionProcessor::maxDepth

protected

float VisionProcessor::maxDist2Class

protected

float VisionProcessor::maxExcentricity

protected

float VisionProcessor::minAreaContour

protected

float VisionProcessor::minAreaFinger

protected

Mat VisionProcessor::minDepthFinger

protected

Mat VisionProcessor::minDepthHand

protected

float VisionProcessor::minExcentricity

protected

float VisionProcessor::minHandExc

protected

const double VisionProcessor::PI = 3.14159265358979323846

staticprotected

vector<Point2f> VisionProcessor::screenContoursCorners

protected

int VisionProcessor::sizes[2]

protected

float VisionProcessor::thicknesFinger

protected

float VisionProcessor::thicknessHand

protected

float VisionProcessor::tolerance

protected

bool VisionProcessor::useScreenSeg

protected

int VisionProcessor::wResolution

protected

---

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

• visionprocessor.h
• visionprocessor.cpp