SoftCascadeClassifier< T > Struct Template Reference

#include <SoftCascade.h>

Public Types
typedef T	WeakClassifier
	the inner weak classifier
typedef SoftCascadeStage< T >	Classifier
	A softcascadeboostable classifier is a weak classifier with a weight associated and a rejection threshold.
typedef std::vector< Classifier >	ClassifierList
	List of BoostSoftCascade classifier.
typedef SoftCascadeClassifier < typename T::OptimizedType >	OptimizedType
	the optimized version of the SoftCascadeStage classifier

Public Member Functions
template<class R >
	SoftCascadeClassifier (const SoftCascadeClassifier< R > &src, int scale, int offset, long stride)
	Create and convert a SoftCascade.
template<class R , class Param1Type , class Param2Type >
	SoftCascadeClassifier (const SoftCascadeClassifier< R > &src, Param1Type p1, const Param2Type &p2)
	Create and convert a SoftCascade.
	SoftCascadeClassifier ()
	ctor
	SoftCascadeClassifier (std::istream &in)
	ctor with file with data
	~SoftCascadeClassifier ()
	dtor
unsigned int	length () const
	Return the number of classifiers.
void	insert (const WeakClassifier &weak, float alpha, float reject)
ClassifierList &	list ()
	Return the inner list of classifier.
const ClassifierList &	list () const
template<class FeatureType >
void	export_features (std::vector< FeatureType > &out) const
template<class TParam1 , class TParam2 >
float	raw (const TParam1 data, const TParam2 p1) const
template<class TParam1 >
float	operator() (const TParam1 pIntImage) const
template<class TParam1 , class TParam2 >
float	operator() (const TParam1 pIntImage, const TParam2 stride) const

Static Public Member Functions
static std::string	signature ()
	propagate signature
template<class TParam1 >
static float	compute (const TParam1 pIntImage, const Classifier *classifier, int n)
	the real computer
template<class TParam1 , class TParam2 >
static float	compute (const TParam1 pIntImage, const TParam2 stride, const Classifier *classifier, int n)
	the real computer

Public Attributes
ClassifierList	m_weak_classifiers
	A che collection of SoftCascadeStage weak classifier.

Detailed Description

template<class T>
struct SoftCascadeClassifier< T >

A classifier that classify with majority voting of N (weak) classifier and uses a Soft Cascade rejection mechanism. Result are voted for majority and rejected with threshold.

Parameters

T	a (Boostable) Weak Classifier type

Member Function Documentation

template<class T>

template<class TParam1 >

float SoftCascadeClassifier< T >::operator() ( const TParam1  pIntImage ) const

inline

the operator to perform a classification using the softcascade

Returns

the sum of single classifiers multiplied by weights, or -1 for negative

template<class T>

template<class TParam1 , class TParam2 >

float SoftCascadeClassifier< T >::operator() ( const TParam1  pIntImage, const TParam2  stride  ) const

inline

the operator to perform a classification using the softcascade

Returns

the sum of single classifiers multiplied by weights, or -1 for negative

template<class T>

template<class TParam1 , class TParam2 >

float SoftCascadeClassifier< T >::raw ( const TParam1  data, const TParam2  p1  ) const

inline

avoid rejection mechanism, for training purpose

Todo:

make name homogeneous with others mechanism

---

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

• include/Classifier/SoftCascade.h