mimas::correlation_tracking_tool< T, C > Class Template Reference
[Tracking of Features]

Correlation based tracking of gray-level images and templates The correlation-type C must be an object of class binary_function with back insertion sequences as first and second parameter and double as return value. More...

#include <correlation_tracking_tool.h>

Inheritance diagram for mimas::correlation_tracking_tool< T, C >:

[legend]Collaboration diagram for mimas::correlation_tracking_tool< T, C >:

[legend]List of all members.


Public Member Functions
void	pushPolyCoord (int xcoord, int ycoord)
	push a point of a polygon which specifies the tracking template
const std::list< Cartesian< int > > &	getPolyCoords (void) const
	returns the current list of the polygon template points
void	showPolyCoords (void)
	display the polygon points
void	clearPolyCoords (void)
	clears the polygon points (e.g. if you want to specify a new polygon)
void	sortPolyCoords (void)
	sort polygon points (useful when displaying/debugging)
void	createTemplate (image< T > &trackingImage)
	create the template from the polygon points that have been pushed
void	writeTemplateToFilePGM (char *fn)
	write the template to a PGM file (for debugging)
void	writeMaskedTemplateToFilePGM (char *fn)
	write the template mask to a PGM file (for debugging)
void	showPolygonOnImage (image< T > &someImage)
	display the polygon onto an image
void	showPolygonOnImage (image< T > &someImage, int greylevel)
	display the polyton onto an image using gray colour <greylevel>.
double	getTrackingCorr (void)
	return the last tracking correlation value
int	getTemplateArea (void)
	return the area of a template
void	saveParameters (const char *fileName)
	save parameters to an XML file
void	loadParameters (const char *fileName)
	load parameters from an XML file
Cartesian< int >	trackCorrelation (image< T > &trackingImage, Cartesian< int > &startCoord, Cartesian< int > &range, C &c)
	perform correlation on trackingImage with specified start coordinates and range
Cartesian< int >	trackCorrelation (image< T > &trackingImage, Cartesian< int > &startCoord, C &c)
	perform correlation on trackingImage with specified start coordinates. The tracking range is assumed pre-specified using the setTrackingRange() functions.
Cartesian< int >	trackCorrelation (image< T > &trackingImage, C &c)
	uses the startCoord from the initial image sent to the template, or the end tracking point from the last tracked image.
std::list< Cartesian< int > >	trackCorrelationList (image< T > &trackingImage, Cartesian< int > &startCoord, Cartesian< int > &range, C &c)
	track and return a list of the points of the final polygon after tracking.
void	setAutoUpdateTemplate (bool val)
	automatically update the template when correlation is to low? Default=false;
void	setAutoUpdateCorrVal (double val)
	lowest correlation before template is updated.
void	setTrackingRange (int range)
	tracking range. The single value specifies the ranges in the x and y direction
void	setTrackingRange (int xrange, int yrange)
	overloaded function to set x and y tracking range seperately
Cartesian< int >	getTrackingPoint (void)
	returns the co-ordinates of the main tracking point
Cartesian< int >	getTrackingDisplacement (void)
	returns the last tracking region displacement
	correlation_tracking_tool (void)
	constructor
virtual	~correlation_tracking_tool (void)
	destructor
void	variableDiagnostics (void)
	stick debugging stuff here
Public Attributes
bool	debug
Protected Attributes
int	numPolyPoints
	number of points on polygon
int *	xpoly
int *	ypoly
	point on the polygon
std::list< Cartesian< int > >	polyCoords
std::list< Cartesian< int > >	finalPolyCoords
	the polygon coordinates before and after execution of tracking procedure
image< T >	trackingTemplate
	the template
image< T >	trackingMask
	the rectangular mask used to store shape of template
int	templateRows
int	templateCols
	rows and columns of the bounding box of the polygon template
Cartesian< int >	offsetForFirstCoord
	offset of first coordinate point from top left corner of bounding box of template

Detailed Description

template<typename T, class C = corrcoef< std::vector< double > >>
class mimas::correlation_tracking_tool< T, C >

Author:: Bala Amavasai (bala@amavasai.org)

See also:: corrcoef

Date:

Fri Apr 6 11:54:53 BST 2001

Header: /cvs/mimas2/include/correlation_tracking_tool.h,v 1.1.1.1 2005/08/09 15:37:45 engmb Exp

Definition at line 29 of file correlation_tracking_tool.h.