SceneRec Class Reference

Class to recognize scenes and get landmark positions. More...

#include <RCBot/SceneRec.H>

Inheritance diagram for SceneRec:
Inheritance graph
[legend]
Collaboration diagram for SceneRec:
Collaboration graph
[legend]

List of all members.

Public Member Functions

 SceneRec (OptionManager &mgr, const std::string &descrName="Scene Recognition", const std::string &tagName="SceneRec")
 ~SceneRec ()
 Destructor.
void newInput (const Image< PixRGB< byte > > img)
 Setup up with image to be processed.
bool outputReady ()
 Is the last given input done processing?
short getLandmarkLoc (Point2D< int > &loc)
 Get the feature point and return the current leg.
void computeLocation ()
 Main process for computing locations.
void trainFeature (Image< PixRGB< byte > > &img, Point2D< int > loc, Dims window, short leg)
 Train the system on a given feature location.
void loadVisualDB (short leg)
 Load the current and the next vdb from position 'leg'.

Protected Member Functions

virtual void start1 ()
 This is called from within start() before the subcomponents start.
virtual void stop2 ()
 This is called from within stop() after the subcomponents have stopped.

Detailed Description

Class to recognize scenes and get landmark positions.

Definition at line 52 of file SceneRec.H.


Constructor & Destructor Documentation

SceneRec::~SceneRec (  ) 

Destructor.

Definition at line 94 of file SceneRec.C.


Member Function Documentation

void SceneRec::computeLocation (  ) 

Main process for computing locations.

Definition at line 182 of file SceneRec.C.

References VisualObjectDB::getObjectMatches(), Point2D< T >::i, Image< T >::initialized(), loadVisualDB(), and VisualObjectDB::numObjects().

short SceneRec::getLandmarkLoc ( Point2D< int > &  loc  ) 

Get the feature point and return the current leg.

Definition at line 122 of file SceneRec.C.

void SceneRec::loadVisualDB ( short  leg  ) 

Load the current and the next vdb from position 'leg'.

Definition at line 308 of file SceneRec.C.

References VisualObjectDB::loadFrom(), VisualObjectDB::numObjects(), and VisualObjectDB::saveTo().

Referenced by computeLocation(), start1(), and trainFeature().

void SceneRec::newInput ( const Image< PixRGB< byte > >  img  ) 

Setup up with image to be processed.

Definition at line 98 of file SceneRec.C.

bool SceneRec::outputReady (  ) 

Is the last given input done processing?

Reimplemented in SceneRecServer_i.

Definition at line 110 of file SceneRec.C.

void SceneRec::start1 (  )  [protected, virtual]

This is called from within start() before the subcomponents start.

Reimplemented from ModelComponent.

Definition at line 67 of file SceneRec.C.

References loadVisualDB().

void SceneRec::stop2 (  )  [protected, virtual]

This is called from within stop() after the subcomponents have stopped.

Reimplemented from ModelComponent.

Definition at line 88 of file SceneRec.C.

void SceneRec::trainFeature ( Image< PixRGB< byte > > &  img,
Point2D< int >  loc,
Dims  window,
short  leg 
)

Train the system on a given feature location.

Definition at line 132 of file SceneRec.C.

References VisualObjectDB::addObject(), Dims::h(), height, Point2D< T >::i, loadVisualDB(), VisualObjectDB::saveTo(), and Dims::w().


The documentation for this class was generated from the following files:
Generated on Sun May 8 08:43:47 2011 for iLab Neuromorphic Vision Toolkit by  doxygen 1.6.3