CVB.Net 15.0
Stemmer.Cvb.Manto Namespace Reference

Classes

class  Classifier
 Manto classifier object. More...
 
class  ClassifierFactory
 Learner object that creates a classifier from an image list. More...
 
class  NamespaceDoc
 The namespace and assembly Stemmer.Cvb.Manto contains the classes and definitions needed for using the Common Vision Blox Manto tool. More...
 
class  SampleImageList
 Class that wraps a Manto SIL More...
 
struct  SearchResult
 Results provided by Manto classifier searches. More...
 

Enumerations

enum  DecisionBoundary { Soft = 0 , Medium = 1 , Hard = 2 }
 Type of decision boundary to be used during classifier generation. More...
 
enum  Symmetries {
  None = 0 , Identity = 1 , MirrorX = 2 , MirrorY = 4 ,
  MirrorXY = 8 , All = 15
}
 Symmetry transformation to be applied temporarily to an SampleImageList during the learning phase. More...
 
enum  InstanceGeneration { Random , Deterministic }
 Approach to be taken if additional instances are to be generated during the learning phase of Manto. More...
 
enum  LearnStage { Unknown , Preprocessing , Learning }
 The states the learner can be in during a learn operation (used in the LearnProgress event). More...
 

Enumeration Type Documentation

◆ DecisionBoundary

Type of decision boundary to be used during classifier generation.

Enumerator
Soft 

Soft decision boundary uses a narrow hinge loss function to allow for a fair amount of closeness to the decision boundary or even boundary violations.

Medium 

Intermediate value between

Soft and Hard.

Hard 

Considerable rigidity to misclassified samples due to broad hinge loss function.

◆ InstanceGeneration

Approach to be taken if additional instances are to be generated during the learning phase of Manto.

Enumerator
Random 

additional instances will be generated randomly during the learning phase (i.e. each learning phase will use a different set of additionally generated images)

Deterministic 

additional instances will be generated deterministically during the learning phase (i.e. each learning phase will always use the same set of additionally generated images)

◆ LearnStage

enum LearnStage

The states the learner can be in during a learn operation (used in the LearnProgress event).

Enumerator
Unknown 

indeterminate learn stage

Preprocessing 

preprocessing the images in the sample image list

Learning 

learning a classifier from the preprocessed images

◆ Symmetries

enum Symmetries

Symmetry transformation to be applied temporarily to an SampleImageList during the learning phase.

Enumerator
None 

no symmetries are to be used during learning; this setting is point- less for an application, but serves as the "null" element for assembling bit fields

Identity 

use the images as they are in the SampleImageList

MirrorX 

mirror the images along the X axis

MirrorY 

mirror the images along the y axis

MirrorXY 

mirror the images along the x and y axis (effectively a 180° rotation)

All 

Combine and use all Symmetries