OCR Image to Text Conversion

This guide provides instructions on how to perform Optical Character Recognition (OCR) to convert images to text using Tesseract-OCR and Python.

Tesseract-OCR Method

Installation

1. Install Tesseract-OCR: sudo apt install tesseract-ocr
2. Download the Bulgarian language dictionary: bul.traineddata
3. Move the downloaded dictionary to the Tesseract-OCR data directory: mv bul.traineddata /usr/share/tesseract-ocr/4.00/tessdata/
4. Install jq sudo apt install jq

Install Python Libraries

1. pip3 install pytesseract
2. pip3 install pillow
3. pip3 install pandas
4. pip3 install argparse
5. pip3 install pdf2image

Convert PDF into Images

1. Install ImageMagic: sudo apt install imagemagick
2. Add this policy in /etc/ImageMagick-6/policy.xml:
3. Convert it: convert -density 300 input.pdf output.png

Instructions

1. Run the python file to OCR the pdf file

• python your_script.py your_json_file.json --delete