ShapeEstimator Class Reference

everything related to the shape estimation procedure More...

#include <Neuro/ShapeEstimator.H>

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

List of all members.

Public Member Functions

 ShapeEstimator (OptionManager &mgr, const std::string &descrName="Shape Estimator", const std::string &tagName="shapeestimator", const nub::soft_ref< VisualCortex > vcx=nub::soft_ref< VisualCortex >())
 Constructor.

Protected Member Functions

 SIMCALLBACK_DECLARE (ShapeEstimator, SimEventWTAwinner)
 Callback for when a new WTA winner is available.
 SIMCALLBACK_DECLARE (ShapeEstimator, SimEventSaveOutput)
 Callback for every time we should save our outputs.
virtual void reset1 ()
 Reset ShapeEstimator.

Protected Attributes

nub::ref< SpatialMetricsitsMetrics
OModelParam< std::stringitsLogFile
 Text log file name.
OModelParam< ShapeEstimatorModeitsMode
 Determines what information is used as a source for the shape estimation.
OModelParam
< ShapeEstimatorSmoothMethod
itsSmMethod
 Determines Smoothing Method used to extract a map in image coordinates.
OModelParam< bool > itsSaveObjMask
 Save our internals?
OModelParam< bool > itsUseLargeNeigh
 use a larger neighborhood when tracking down a local max?

Detailed Description

everything related to the shape estimation procedure

class to take care of all things related to the shape estimation procedure published in Walther et al, BMCV 02

Definition at line 61 of file ShapeEstimator.H.


Constructor & Destructor Documentation

ShapeEstimator::ShapeEstimator ( OptionManager mgr,
const std::string descrName = "Shape Estimator",
const std::string tagName = "shapeestimator",
const nub::soft_ref< VisualCortex vcx = nub::soft_ref<VisualCortex>() 
)

Constructor.

Parameters:
vcx this is the hook for the class to obtain its information on channels, submaps etc.

Definition at line 78 of file ShapeEstimator.C.

References ModelComponent::addSubComponent().


Member Function Documentation

void ShapeEstimator::reset1 (  )  [protected, virtual]

Reset ShapeEstimator.

See the base function in ModelComponent.H for info.

Reimplemented from ModelComponent.

Definition at line 99 of file ShapeEstimator.C.

References Image< T >::freeMem().

ShapeEstimator::SIMCALLBACK_DECLARE ( ShapeEstimator  ,
SimEventSaveOutput   
) [protected]

Callback for every time we should save our outputs.

metrics that depend on the input size:

ShapeEstimator::SIMCALLBACK_DECLARE ( ShapeEstimator  ,
SimEventWTAwinner   
) [protected]

Callback for when a new WTA winner is available.


Member Data Documentation

Text log file name.

Definition at line 85 of file ShapeEstimator.H.

Determines what information is used as a source for the shape estimation.

possible values:

  • "None" - shape estimator is not used;
  • "FeatureMap" - the shape is extracted from the winning Feature Map;
  • "ConspicuityMap" - shape is extracted from the winning Conspicuity Map;
  • "SaliencyMap" - the shape is extracted from the Saliency Map

Definition at line 92 of file ShapeEstimator.H.

Save our internals?

Definition at line 103 of file ShapeEstimator.H.

Determines Smoothing Method used to extract a map in image coordinates.

possible values:

  • "None" - no smoothing, the result is just scaled up in blocks;
  • "Gaussian" - smoothing by convolving with a large 2D Gaussian kernel;
  • "Chamfer" - smoothing with Opening and Chamfering (takes about half the time compared to "Gaussian"

Definition at line 100 of file ShapeEstimator.H.

use a larger neighborhood when tracking down a local max?

Definition at line 106 of file ShapeEstimator.H.


The documentation for this class was generated from the following files:
Generated on Sun May 8 08:24:55 2011 for iLab Neuromorphic Vision Toolkit by  doxygen 1.6.3