new OCROptions()
Methods
-
addDPI(dpi)
-
Sets the value for DPI in the options object.
Knowing proper image resolution is important, as it enables the OCR engine to translate pixel heights of characters to their respective font sizes.
We do our best to retrieve resolution information from the input's metadata, however it occasionally can be corrupt or missing. Hence we allow manual override of source's resolution, which supersedes any metadata found (both explicit as in image metadata and implicit as in PDF).
If input is a PDF file, the SDK will render PDF page for OCR as a raster image with 300 DPI resolution. Otherwise, if input is an image, SDK will pass to OCR raster image in original resolution. Exceptionally, if the user, by using this method, suggests the resolution (say 300 DPI), the SDK will try do as suggested.
Here, some restrictions apply:
- New image size in pixels (Width X Height) cannot exceed 75 MP (Mega Pixels)
- Each image side (Width or Height), in pixels, cannot exceed 16 bits (65,535)
To achieve that, SDK will iteratively scale down the DPI till both conditions are met.
Parameters:
Name Type Description dpinumber Image resolution Returns:
This object, for call chaining. -
addIgnoreZonesForPage(regions, page_num)
-
Adds a collection of ignorable regions for the given page.
It is an optional list of areas that will be excluded from analysis.
The meaning and handling of the passed region coordinates will vary depending on:
- Whether input is PDF (rather than an image), and
- Whether SetUsePDFPageCoords() is called with value true.
If either of above is true, then the region coordinates have to be given in PDF Points and coodinate space origin at bottom-left.
Otherwise:
- The input is a raster image,
- The coordinates has to be given in pixel units and with coordinate space origin at top-left
See also: ProcessPDF() ImageToPDF() SetUsePDFPageCoords() AddTextZonesForPage().
Parameters:
Name Type Description regionsArray.<PDFNet.Rect> The new zones to be added to ignore list. page_numnumber The page number the added regions belong to. Returns:
This object, for call chaining. -
addLang(lang)
-
Adds a language to the list of to be considered when processing this document.
OCR engine is prepared for recognition of a predefined set of language scripts. Various scripts have differences in glyphs and it is recommended that OCR engine is instructed to use a correct glyph set to match the appropriate Unicode. At this moment, there are two avaliable engines for use. Let we refer them by their executable names: OCRModule and OCRModuleIRIS.
Here is a list of supported languages by their codes that may be passed as a parameter. One language per call to AddLang(). Use multiple calls to AddLang() to add additional languages.
Warning: Be aware that adding too many language codes at once may lead to wrong altertnatives regarding Latin characters with accents.
Language Code Language Code English eng Chinese Simpl. chi_sim French fra Chinese Trad. chi_tra German deu Japanese jpn Italian ita Korean kor Russian rus Spanish spa Here are the languages supported out of the box per engine:
Executable Name Supported Codes OCRModule eng fra deu ita spa rus OCRModuleIRIS eng fra deu ita spa rus chi_sim chi_tra jpn kor Note: OCRModuleIRIS allows mix of a single Asian language and just English.
Parameters:
Name Type Description langstring The new language to be added to language list. Returns:
This object, for call chaining. -
addTextZonesForPage(regions, page_num)
-
Adds a collection of text regions of interest for the given page.
It is as an optional list of known text zones that will be used to improve OCR quality.
See also: AddIgnoreZonesForPage().
Parameters:
Name Type Description regionsArray.<PDFNet.Rect> the zones to be added to the text region list. page_numnumber The page number the added regions belong to Returns:
This object, for call chaining. -
getAutoRotate()
-
Gets the value AutoRotate from the options object.
Default value is false.
Note: This function doesn't apply to IRIS OCR module.
Returns:
The current value for AutoRotate.- Type
- boolean
-
getDPI()
-
Gets the value DPI from the options object.
Knowing proper image resolution is important, as it enables the OCR engine to translate pixel heights of characters to their respective font sizes.
We do our best to retrieve resolution information from the input's metadata, however it occasionally can be corrupt or missing. Hence we allow manual override of source's resolution, which supersedes any metadata found (both explicit as in image metadata and implicit as in PDF).
If input is a PDF file, the SDK will render PDF page for OCR as a raster image with 300 DPI resolution. Otherwise, if input is an image, SDK will pass to OCR raster image in original resolution. Exceptionally, if the user, by using this method, suggests the resolution (say 300 DPI), the SDK will try do as suggested.
Here, some restrictions apply:
- New image size in pixels (Width X Height) cannot exceed 75 MP (Mega Pixels)
- Each image side (Width or Height), in pixels, cannot exceed 16 bits (65,535)
To achieve that, SDK will iteratively scale down the DPI till both conditions are met.
Returns:
The current value of DPI.- Type
- number
-
getIgnoreExistingText()
-
Gets the value IgnoreExistingText from the options object.
Default value is false, so that areas with existing text will be automatically skipped during OCR. Setting to true will cause a pre-existing text to be duplicated with the OCR-ed ones in the PDF document or in GetOCRJsonFromPDF() and GetOCRXmlFromPDF() results.
Returns:
The current value for IgnoreExistingText.- Type
- boolean
-
getOCREngine()
-
Gets the value OCREngine from the options object.
Options include 'default' or 'iris'. Chosen module must be present and correctly licensed.
Returns:
The current value for OCREngine.- Type
- string
-
getUsePDFPageCoords()
-
Gets the value UsePDFPageCoords from the options object.
Defines the coordinate system, scaling and units. SDK and OCRModule will refer to this setting while dealing with potential input zone rectangle(s) (ignorable or text) and with the output result as well. The default value is false which corresponds to raster image input.
Here are the meanings for value:
Value Origin +Y axis dir. Unit Size Expected input false top-left Downwards Pixel 1/DPI inches Image true bottom-left Upwards Point 1/72 inches PDF Note that, OCRModule backend has no notion about the SDK's real input. So, if not explicitelly instructed via this call and value, it will work by default and will report the coordinates as for false. This is important to know for cases when the call for OCR service comes from GetOCRJsonFromPDF() or GetOCRXmlFromPDF() where the results should be correct in JSON or XML format - and removes the need for user to do additional adjustments.
See also: AddIgnoreZonesForPage() AddTextZonesForPage() AddDPI() GetOCRJsonFromImage() GetOCRXmlFromImage().
Returns:
The current value for UsePDFPageCoords.- Type
- boolean
-
setAutoRotate(value)
-
Sets the value for AutoRotate in the options object.
Default value is false. Setting to true will deskew the image before conducting OCR.
Note: This function doesn't apply to IRIS OCR module.
Parameters:
Name Type Description valueboolean The new value for AutoRotate Returns:
This object, for call chaining. -
setIgnoreExistingText(value)
-
Sets the value for IgnoreExistingText in the options object.
Default value is false, so that areas with existing text will be automatically skipped during OCR. Setting to true will cause a pre-existing text to be duplicated with the OCR-ed ones in the PDF document or in GetOCRJsonFromPDF() and GetOCRXmlFromPDF() results.
Parameters:
Name Type Description valueboolean The new value for IgnoreExistingText Returns:
This object, for call chaining. -
setOCREngine(value)
-
Sets the value for OCREngine in the options object.
Options include 'default' or 'iris'. Chosen module must be present and correctly licensed.
Parameters:
Name Type Description valuestring The new value for OCREngine Returns:
This object, for call chaining. -
setUsePDFPageCoords(value)
-
Sets the value for UsePDFPageCoords in the options object.
Defines the coordinate system, scaling and units. SDK and OCRModule will refer to this setting while dealing with potential input zone rectangle(s) (ignorable or text) and with the output result as well. The default value is false which corresponds to raster image input.
Here are the meanings for value:
Value Origin +Y axis dir. Unit Size Expected input false top-left Downwards Pixel 1/DPI inches Image true bottom-left Upwards Point 1/72 inches PDF Note that, OCRModule backend has no notion about the SDK's real input. So, if not explicitelly instructed via this call and value, it will work by default and will report the coordinates as for false. This is important to know for cases when the call for OCR service comes from GetOCRJsonFromPDF() or GetOCRXmlFromPDF() where the results should be correct in JSON or XML format - and removes the need for user to do additional adjustments.
See also: AddIgnoreZonesForPage() AddTextZonesForPage() AddDPI() GetOCRJsonFromImage() GetOCRXmlFromImage().
Parameters:
Name Type Description valueboolean The new value for UsePDFPageCoords Returns:
This object, for call chaining.