autoocr

A Python wrapper for cross platform tesseract OCR engine with multiple languages (e.g. Bangla)

Installations

pip3 install autoocr

Usage

Mac OS

• Import the library

from autoocr import AutoOCR # import the AutoOCR class

• Specify the language

oa = AutoOCR(lang='bangla') # specify the language code

• Set the tessdata folder, on mac you can do brew list tesseract to get the path. This is only needed once.

oa.set_datapath('/usr/local/Cellar/tesseract/4.0.0_1/share/tessdata')

• Get the text from image by passing the path to image

out_text = oa.get_text('image_ocr.jpg')

Windows

• Install tesseract engine

• Import the library

from autoocr import AutoOCR # import the AutoOCR class

• Specify the language

oa = AutoOCR(lang='bangla') # specify the language code

• Set the tessdata folder. This is only needed once.

oa.set_datapath('/path/to/tessdata')

• Get the text from image by passing the path to image

out_text = oa.get_text('image_ocr.jpg')

Linux

• Install tesseract engine. Follow this page tesseract-ocr

• Import the library

from autoocr import AutoOCR # import the AutoOCR class

• Specify the language

oa = AutoOCR(lang='bangla') # specify the language code

• Set the tessdata folder. This is only needed once. Run, rpm -ql tesseract for yum to get the location.

oa.set_datapath('/path/to/tessdata')

• Get the text from image by passing the path to image

out_text = oa.get_text('image_ocr.jpg')

License

This project is licensed under the MIT License - see the LICENSE file for details.