Aspose.OCR for Python via Java 25.2.0 - Release Notes

What was changed

Public API changes and backwards compatibility

This section lists all public API changes introduced in Aspose.OCR for Python via Java 25.2.0 that may affect the code of existing applications.

Added public APIs:

The following public APIs have been introduced in 25.2.0 release:

Format.HOCR

Instructs Aspose.OCR library to save the recognition results in hOCR format - an open standard of data representation for formatted text obtained from OCR. It includes extracted text, style, layout, and other information.

OcrOutput class

A container class that stores recognition results, which is returned from all recognition methods instead of a collection of recognition results.

It is fully backward compatible with the previously used collection of RecognitionResult objects, which means you do not have to update your existing code.

Resources.release_memory()

Unload all OCR modules to free up memory. The downloaded module files will remain stored on your system.

If you need to use the OCR module again later, it will automatically reload into memory. Please note that the first recognition attempt may be slightly slower due to the reloading process.

save_multipage_pdf_optimized()

This method allows to balance between file size and image quality of saved PDFs, using the optional optimizePdf parameter, which accepts the value of PdfOptimizationMode enumeration.

Updated public APIs:

The following public APIs have been changed in 25.2.0 release:

Language enumeration

Aspose.OCR for Python via Java can now recognize a number of new languages, including texts in mixed scripts:

Removed public APIs:

The following APIs deprecated during 2024 have been removed in this release:

recognize_street_photo()

Use the universal recognize() method which allows to control recognition settings, multiple languages, image regions, spellcheck, and other advanced features.

DetectAreasMode.NONE

Use DetectAreasMode.LEAN, which provides the same functionality.

DetectAreasMode.DOCUMENT

Use DetectAreasMode.MULTICOLUMN, which provides the same functionality.

DetectAreasMode.TEXT_IN_WILD

Use DetectAreasMode.UNIVERSAL, which detects all blocks of text in the image, including sparse and irregular text on street photos.

DetectAreasMode.COMBINE

Use DetectAreasMode.UNIVERSAL, which works best with sparse irregular text.

DetectAreasMode.PHOTO

Use DetectAreasMode.UNIVERSAL, which is optimal for all types of text except for multi-column layouts and tables.

Language.CZE

Use Language.CES to recognize texts in Czech.

Language.DUM

Use Language.NLD to recognize texts in Dutch.

Language.RUM

Use Language.RON to recognize texts in Romanian.

Language.SRP_HRV

Use Language.HBS to recognize texts in Serbo-Croatian (Latin alphabet).

Language.CHI

Use Language.CHINESE to recognize all Chinese languages, including mixed-language Chinese/English texts.

Language.NONE

Specify the recognition language directly. If the language is not specified, the OCR engine defaults to a lightweight Latin character model, which does not support diacritics.

SpellCheckLanguage.CZE

Use SpellCheckLanguage.CES to check the spelling in Czech texts.

SpellCheckLanguage.DUM

Use SpellCheckLanguage.NLD to check the spelling in Dutch texts.

SpellCheckLanguage.RUM

Use SpellCheckLanguage.RON to check the spelling in Romanian texts.

Examples

The examples below illustrates the changes introduced in this release:

Japanese characters recognition

import aspose as ocr

ocr_resources = []
ocr_resources.append('aspose-ocr-japanese-v1')
ocr.Resources.fetch_resources(ocr_resources)

api = ocr.AsposeOcr()
images = ocr.OcrInput(ocr.InputType.SINGLE_IMAGE)
images.add('source.png')

recognitionSettings = RecognitionSettings()
recognitionSettings.setLanguage(Language.JPN)

result = api.recognize(images, recognitionSettings)
print(result[0].recognition_text)

Unloading OCR modules from memory

import aspose as ocr

ocr_resources = []
ocr_resources.append('aspose-ocr-japanese-v1')
ocr.Resources.fetch_resources(ocr_resources)

api = ocr.AsposeOcr()
images = ocr.OcrInput(ocr.InputType.SINGLE_IMAGE)
images.add('source.png')

recognitionSettings = RecognitionSettings()
recognitionSettings.setLanguage(Language.JPN)

result = api.recognize(images, recognitionSettings)
print(result[0].recognition_text)
ocr.Resources.release_memory()

Save recognition results to optimized PDF

import aspose as ocr

api = ocr.AsposeOcr()
images = ocr.OcrInput(ocr.InputType.SINGLE_IMAGE)
images.add('page1.png')
images.add('page2.png')

results = api.recognize(images, recognitionSettings)
save_multipage_pdf_optimized('target.pdf', results, "", PdfOptimizationMode.BALANCED)