Stemmer.Cvb.Polimago Namespace Reference

Namespace for the Polimago tool. More...

Namespaces
namespace	Testing
	This namespace contains classes and definitions that help testing a given training set versus either a test set or versus itself by means of a leave-out-test.
namespace	Training
	This namespace contains classes and definitions needed for generating classifiers from training data collections.

Classes
class	ClassificationPredictor
	Predictor to classify patterns with. More...
struct	ClassificationResult
	Polimago classification result container. More...
class	PolimagoExtensions
	Extension methods for the Stemmer.Cvb.Polimago library. More...
class	PolimagoFactoryCreatedObject
	Base class for Polimago objects created by one of the factory classes. More...
class	PolimagoRefCountedObject
	Base class for Polimago's reference counted objects. More...
class	PredictorBase
	Base class for Polimago predictors. More...
class	PredictorBaseEx
	Base class for all Polimago predictors. More...
class	RegressionPredictor
	Polimago Regression predictor. More...
class	SearchPredictor
	Predictor that may be used for searching objects. More...
struct	SearchResult
	Search results as provided by a Search Classifier. More...

Enumerations
enum	ClassificationType { None , OneVersusAll , OneVersusOne }
	Determine the classification type to be carried out. More...
enum	InterpolationType { None = 0 , Linear = 1 }
	Interpolation to be used when extracting image data for classifier generation. More...
enum	InvarianceType { None = 0 , Translation = 1 , RotationScaleTranslation = 2 , AffineGroup = 3 }
	Invariance types that can be defined for training. More...

Detailed Description

Namespace for the Polimago tool.

This namespace contains the classes and definitions needed for using the Common Vision Blox Polimago tool.

Online Manual

Search and Classification with Polimago

Enumeration Type Documentation

ClassificationType

enum ClassificationType

Determine the classification type to be carried out.

Enumerator
None	The enum element indicating undefined state.
OneVersusAll	Quick classification that tests versus all classes simultaneously (but potentially at the cost of result quality, especially if many classes are involved).
OneVersusOne	Thorough classification that tests all possible pairs of classes (which yields potentially better results if many classes are involved but takes significantly longer than the OneVersusAll approach).

InterpolationType

enum InterpolationType

Interpolation to be used when extracting image data for classifier generation.

Enumerator
None	Image data is (if necessary) extracted without interpolation.
Linear	Image data is (if necessary) extracted with linear interpolation.

InvarianceType

enum InvarianceType

Invariance types that can be defined for training.

Enumerator
None	No invariance (not valid for training!).
Translation	Translation.
RotationScaleTranslation	Rotation + Scale + Translation.
AffineGroup	Affine group (i.e. 2x2 matrix plus translation)