Extracting handwritten text

Aspose.OCR offers a special recognition algorithm for extracting handwritten text from images. It supports a number of European languages based on Extended Lain alphabet. Handwritten recognition has various business, government and personal applications:

  • Extract text from notes, memos and letters.
  • Convert historical records and documents into digital formats for archival purposes.
  • Parse resumes and employee records.
  • Analyze information from handwritten inventory lists to maintain accurate and up-to-date databases.
  • Analyze customer feedback and reviews to gain insights into customer preferences.
  • And many more.

To extract handwritten text from images, use recognize_handwritten_text() method of AsposeOcr class.

The method takes OcrInput object and returns OcrOutput object containing the text from images.

import aspose.ocr as ocr
# Initialize an instance of Aspose.OCR API
api = ocr.AsposeOcr()
# Add image to the recognition batch
input = ocr.OcrInput(ocr.InputType.SINGLE_IMAGE)
input.add("memo.jpg")
# Extract and show text
results = api.recognize_handwritten_text(input)
print(results[0].recognition_text)

Live demo

Handwritten note 
WE STARF WITH GOOD
BECAUSE ALL BUSINESSES SHOULD
BE DOING SOMETHING GOOD

Limitations

  • The method only supports a limited subset of Extended Latin letters and numbers.
  • Both uppercase and lowercase letters are recognized. However, the resulting text will be in uppercase.
  • This method does not support recognition settings. The recognition language is detected automatically.
Extracting text from street photos