Topics
	MultiLayer Perceptron (MLP)
	Support-Vector Machines (SVM)

Classes
class	oevislib_net.ComputerVision.OCR.CharacterSample
	A pair composed by the character representation (as a Region) and its corresponding text (as a string). More...
class	oevislib_net.ComputerVision.OCR.Model
	Class containing OCR model data used to recognize text. More...
class	oevislib_net.ComputerVision.OCR.TextSegmentationMethod
	Parameters for text segmentation algorithms. More...
class	oevislib_net.ComputerVision.OCR.TextThresholdingMethod
	Parameters of text extraction algorithm. More...
class	oevislib_net.ComputerVision.OCR.TextSegmentation
	Defines how text should be split. More...

Functions
static Array< CharacterSample >	oevislib_net.ComputerVision.OCR.OCR.MakeCharacterSamples (Array< Region > inRegions, string inCharacters)
	Creates an array of character samples, composed by a region and the corresponding character.
static Array< CharacterSample >	oevislib_net.ComputerVision.OCR.OCR.MakeCharacterSamples (string inFont, int inFontSize, string inCharacters)
	Creates an array of character samples, from a known font and a set of characters.
static Array< Region >	oevislib_net.ComputerVision.OCR.OCR.SplitRegionIntoCharacters (Region inRegion, float inProjectionSmooth=1.0f, int inCharacterWidth=15, Optional< Array< Region > > outClasses=null, Optional< Profile > outProjection=null)
	Splits the input region into an array of character regions.
static Array< Region >	oevislib_net.ComputerVision.OCR.OCR.ExtractText (Image inImage, Rectangle2D inRoi, UCS2D inRoiAlignment=null, TextSegmentation inSegmentationModel=null, Optional< Region > outTextRegion=null, Optional< Array< Region > > outAlignedCharacters=null, Optional< Rectangle2D > outAlignedRoi=null)
	Extracts and splits character in single characters.
static string	oevislib_net.ComputerVision.OCR.OCR.ReadText (Array< Region > inCharacters, Model inModel, float inMinScore, Optional< Array< Conditional< string > > > outCharacters=null, Optional< Conditional< Array< float > > > outScores=null, Optional< bool > outIsValid=null)
	Reads text from images using a pre-trained OCR model.

Detailed Description

Function Documentation

◆ ExtractText()

Array< Region > oevislib_net.ComputerVision.OCR.OCR.ExtractText	(	Image	inImage,
		Rectangle2D	inRoi,
		UCS2D	inRoiAlignment = null,
		TextSegmentation	inSegmentationModel = null,
		Optional< Region >	outTextRegion = null,
		Optional< Array< Region > >	outAlignedCharacters = null,
		Optional< Rectangle2D >	outAlignedRoi = null )

inlinestatic

Extracts and splits character in single characters.

Parameters

inImage	Input image.
inRoi	Where to find text.
inRoiAlignment	Aligns the ROI to the correct position.
inSegmentationModel	Model used for text segmentation.
outTextRegion	Region of text after extraction.
outAlignedCharacters	Extracted character with the original image orientation.
outAlignedRoi	Text extraction zone after alignment.

Returns: Extracted characters (aligned to the ROI).

◆ MakeCharacterSamples() [1/2]

Array< CharacterSample > oevislib_net.ComputerVision.OCR.OCR.MakeCharacterSamples	(	Array< Region >	inRegions,
		string	inCharacters )

inlinestatic

Creates an array of character samples, composed by a region and the corresponding character.

Parameters

inRegions	Array of regions.
inCharacters	Corresponding characters.

Returns: Array of CharacterSamples.

◆ MakeCharacterSamples() [2/2]

Array< CharacterSample > oevislib_net.ComputerVision.OCR.OCR.MakeCharacterSamples	(	string	inFont,
		int	inFontSize,
		string	inCharacters )

inlinestatic

Creates an array of character samples, from a known font and a set of characters.

Parameters

inFont	Font file (.ttf/.ttc/.otf).
inFontSize	Character height.
inCharacters	Set of characters to render.

Returns: Array of CharacterSamples.

◆ ReadText()

string oevislib_net.ComputerVision.OCR.OCR.ReadText	(	Array< Region >	inCharacters,
		Model	inModel,
		float	inMinScore,
		Optional< Array< Conditional< string > > >	outCharacters = null,
		Optional< Conditional< Array< float > > >	outScores = null,
		Optional< bool >	outIsValid = null )

inlinestatic

Reads text from images using a pre-trained OCR model.

Parameters

inCharacters	Character regions.
inModel	Pre-trained OCR model.
inMinScore	Minimal score for the correct reading of a character. Range: [0, 1].
outCharacters	Single read characters. Null represents an invalid read.
outScores	Scores obtained for each character.
outIsValid	false if any character has obtained a score smaller than inMinScore.

Returns: Read text.

◆ SplitRegionIntoCharacters()

Array< Region > oevislib_net.ComputerVision.OCR.OCR.SplitRegionIntoCharacters	(	Region	inRegion,
		float	inProjectionSmooth = 1::0f,
		int	inCharacterWidth = 15,
		Optional< Array< Region > >	outClasses = null,
		Optional< Profile >	outProjection = null )

inlinestatic

Splits the input region into an array of character regions.

Parameters

inRegion	Input region.
inProjectionSmooth	Smooth value applied to the region projection before splitting. Range: [0, +inf).
inCharacterWidth	Single character width. Range: [1, +inf).
outClasses	Regions where characters are connected into a single region (based on inCharacterWidth).
outProjection	Region projection.

Returns: Array single character regions.

Topics

Classes

Functions

Detailed Description

Function Documentation

◆ ExtractText()

◆ MakeCharacterSamples() [1/2]

◆ MakeCharacterSamples() [2/2]

◆ ReadText()

◆ SplitRegionIntoCharacters()