Saving recognition results as a file

To save recognition results to a file or write them to the memory stream, use save() method of RecognitionResult object.

The following file formats are supported:

Format	Description
SaveFormat.TEXT	Plain text with line breaks.
SaveFormat.PDF	Portable Document Format. The original images are placed in the background; the recognized text is placed as an invisible but searchable and selectable overlay on top of the images.
SaveFormat.PDF_NO_IMG	Portable Document Format containing only text. The original images are not saved along with the recognition results.
SaveFormat.DOCX	Microsoft Word document (version 2007 or later).
SaveFormat.XLSX	Microsoft Excel spreadsheet (version 2007 or later).
SaveFormat.JSON	JSON is a popular format widely used in software development and data exchange. The de facto standard for websites and REST APIs.
SaveFormat.XML	Extensible Markup Language (XML), a universal data exchange and storage format for most systems.
SaveFormat.HTML	HTML web page.
SaveFormat.EPUB	ePub, a popular e-book file format.
SaveFormat.RTF	RTF, a universal format for exchanging text documents between different word processing programs.

You can optionally enable automatic spelling corrections for recognition results, provide a custom dictionary, or specify the font to be embedded into a PDF document. The latter is only applicable when saving recognition results into text-only PDF (SaveFormat.PDF_NO_IMG).

# Instantiate Aspose.OCR API
api = AsposeOcr()
# Add image to the recognition batch
input = OcrInput(InputType.SINGLE_IMAGE)
input.add("source.png")
# Recognize the image
results = api.recognize(input)
# Save recognition result
results[0].save("result.txt", SaveFormat.TEXT)

Saving recognition results as a multi-page document

You can merge several recognition results into one document using save_multipage_document() method of AsposeOcr class. This can be useful for recognizing books, contracts, articles, and other printouts consisting of multiple pages, as well as for batch recognition.

You can optionally enable automatic spelling corrections for recognition results and provide a custom dictionary.

# Instantiate Aspose.OCR API
api = AsposeOcr()
# Add image to the recognition batch
input = OcrInput(InputType.SINGLE_IMAGE)
input.add("page1.png")
input.add("page2.png")
input.add("page3.png")
# Recognize the image
results = api.recognize(input)
# Save recognition result
save_multipage_document("result.docx", SaveFormat.DOCX, results)