com.itextpdf.pdfocr.tesseract4

Class Tesseract4OcrEngineProperties

java.lang.Object

com.itextpdf.pdfocr.OcrEngineProperties

com.itextpdf.pdfocr.tesseract4.Tesseract4OcrEngineProperties

public class Tesseract4OcrEngineProperties
extends OcrEngineProperties

Properties that will be used by the IOcrEngine.

Constructor Summary

Constructors

Constructor and Description
Tesseract4OcrEngineProperties() Creates a new Tesseract4OcrEngineProperties instance.
Tesseract4OcrEngineProperties(Tesseract4OcrEngineProperties other) Creates a new Tesseract4OcrEngineProperties instance based on another Tesseract4OcrEngineProperties instance (copy constructor).

Method Summary

Modifier and Type	Method and Description
String	getDefaultLanguage() Gets default language for ocr.
String	getDefaultUserWordsSuffix() Gets default user words suffix.
ImagePreprocessingOptions	getImagePreprocessingOptions() Gets imagePreprocessingOptions.
int	getMinimalConfidenceLevel() Gets minimal confidence level for HOCR line to be considered as properly recognized.
Integer	getPageSegMode() Gets Page Segmentation Mode.
File	getPathToTessData() Gets path to directory with tess data.
TextPositioning	getTextPositioning() Defines the way text is retrieved from tesseract output using TextPositioning.
boolean	isPreprocessingImages() Checks whether image preprocessing is needed.
boolean	isUseTxtToImproveHocrParsing() Gets useTxtToImproveHocrParsing.
Tesseract4OcrEngineProperties	setImagePreprocessingOptions(ImagePreprocessingOptions imagePreprocessingOptions) Sets imagePreprocessingOptions.
Tesseract4OcrEngineProperties	setMinimalConfidenceLevel(int minimalConfidenceLevel) Sets minimal confidence level for HOCR line to be considered as properly recognized.
Tesseract4OcrEngineProperties	setPageSegMode(Integer mode) Sets Page Segmentation Mode.
Tesseract4OcrEngineProperties	setPathToTessData(File tessData) Sets path to directory with tess data.
Tesseract4OcrEngineProperties	setPreprocessingImages(boolean preprocess) Sets true if image preprocessing is needed.
Tesseract4OcrEngineProperties	setTextPositioning(TextPositioning positioning) Defines the way text is retrieved from tesseract output using TextPositioning.
Tesseract4OcrEngineProperties	setUseTxtToImproveHocrParsing(boolean useTxtToImproveHocrParsing) Sets useTxtToImproveHocrParsing.

Methods inherited from class com.itextpdf.pdfocr.OcrEngineProperties

getLanguages, setLanguages

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

Tesseract4OcrEngineProperties

public Tesseract4OcrEngineProperties()

Creates a new Tesseract4OcrEngineProperties instance.

Tesseract4OcrEngineProperties

public Tesseract4OcrEngineProperties(Tesseract4OcrEngineProperties other)

Creates a new Tesseract4OcrEngineProperties instance based on another Tesseract4OcrEngineProperties instance (copy constructor).

Parameters:

other - the other Tesseract4OcrEngineProperties instance

Method Detail

getDefaultLanguage

public final String getDefaultLanguage()

Gets default language for ocr.

Returns:

default language - "eng"

getDefaultUserWordsSuffix

public final String getDefaultUserWordsSuffix()

Gets default user words suffix.

Returns:

default suffix for user words files

getPathToTessData

public final File getPathToTessData()

Gets path to directory with tess data.

Returns:

path to directory with tess data

setPathToTessData

public final Tesseract4OcrEngineProperties setPathToTessData(File tessData)

Sets path to directory with tess data.

Parameters:

tessData - path to train directory as File

Returns:

the Tesseract4OcrEngineProperties instance

Throws:

PdfOcrTesseract4Exception - if path to tess data directory is null or empty, or provided directory does not exist? or it is not a directory

getPageSegMode

public final Integer getPageSegMode()

Gets Page Segmentation Mode.

Returns:

psm mode as Integer

setPageSegMode

public final Tesseract4OcrEngineProperties setPageSegMode(Integer mode)

Sets Page Segmentation Mode. More detailed explanation about psm modes could be found here https://github.com/tesseract-ocr/tesseract/blob/master/doc/tesseract.1.asc#options Note that in documentation it is stated that default value of PSM is 3. This is true for tesseract executable, but for tesseract lib it is -1 which has negative impact on some documents. That's why in the code we set it explicitly to 3.

Parameters:

mode - psm mode as Integer

Returns:

the Tesseract4OcrEngineProperties instance

isPreprocessingImages

public final boolean isPreprocessingImages()

Checks whether image preprocessing is needed.

Returns:

true if images need to be preprocessed, otherwise - false

setPreprocessingImages

public final Tesseract4OcrEngineProperties setPreprocessingImages(boolean preprocess)

Sets true if image preprocessing is needed.

Parameters:

preprocess - true if images need to be preprocessed, otherwise - false

Returns:

the Tesseract4OcrEngineProperties instance

getTextPositioning

public final TextPositioning getTextPositioning()

Defines the way text is retrieved from tesseract output using TextPositioning.

Returns:

the way text is retrieved

setTextPositioning

public final Tesseract4OcrEngineProperties setTextPositioning(TextPositioning positioning)

Defines the way text is retrieved from tesseract output using TextPositioning.

Parameters:

positioning - the way text is retrieved

Returns:

the Tesseract4OcrEngineProperties instance

isUseTxtToImproveHocrParsing

public final boolean isUseTxtToImproveHocrParsing()

Gets useTxtToImproveHocrParsing. Used to make HOCR recognition result more precise. This is needed for cases of Thai language or some Chinese dialects where every character is interpreted as a single word. For more information see https://github.com/tesseract-ocr/tesseract/issues/2702

Returns:

useTxtToImproveHocrParsing

setUseTxtToImproveHocrParsing

public final Tesseract4OcrEngineProperties setUseTxtToImproveHocrParsing(boolean useTxtToImproveHocrParsing)

Sets useTxtToImproveHocrParsing. Used to make HOCR recognition result more precise. This is needed for cases of Thai language or some Chinese dialects where every character is interpreted as a single word. For more information see https://github.com/tesseract-ocr/tesseract/issues/2702

Parameters:

useTxtToImproveHocrParsing - useTxtToImproveHocrParsing

Returns:

this Tesseract4OcrEngineProperties instance.

getImagePreprocessingOptions

public final ImagePreprocessingOptions getImagePreprocessingOptions()

Gets imagePreprocessingOptions.

Returns:

ImagePreprocessingOptions

setImagePreprocessingOptions

public final Tesseract4OcrEngineProperties setImagePreprocessingOptions(ImagePreprocessingOptions imagePreprocessingOptions)

Sets imagePreprocessingOptions.

Parameters:

imagePreprocessingOptions - ImagePreprocessingOptions

Returns:

the Tesseract4OcrEngineProperties instance

getMinimalConfidenceLevel

public final int getMinimalConfidenceLevel()

Gets minimal confidence level for HOCR line to be considered as properly recognized. If real confidence level is lower then line is ignored Default value is 0 which means that everything is considered as properly recognized Value may vary in range of 0-100

Returns:

minimal confidence level

setMinimalConfidenceLevel

public final Tesseract4OcrEngineProperties setMinimalConfidenceLevel(int minimalConfidenceLevel)

Sets minimal confidence level for HOCR line to be considered as properly recognized. If real confidence level is lower then line is ignored Default value is 0 which means that everything is considered as properly recognized Value may vary in range of 0-100

Parameters:

minimalConfidenceLevel - minimal confidence level value

Returns:

this Tesseract4OcrEngineProperties instance