Image Text Reader

The image-text-reader library allows you to extract hungarian text from images using Optical Character Recognition (OCR) with the help of the pytesseract library and Pillow for image processing.

note: the current code only works if tessaract is set up for hungarian language pack

Table of Contents

• Prerequisites
• Installation
• Usage
• Code Explanation
• Contributing
• License

Prerequisites

• Python 3.x
• Tesseract-OCR

Installation

1. Install the required Python libraries:

   pip install image-text-reader

2. Install Tesseract-OCR:

   • Windows: Download and install from here.

   • macOS: Use Homebrew to install:

     brew install tesseract

   • Linux: Use your package manager, for example:

     sudo apt-get install tesseract-ocr

Usage

1. Create a Python script (e.g., test_script.py) and import the ocr_image function from the image_text_reader library:

   from image_text_reader import ocr_image

2. Set the path to your image and Tesseract-OCR executable:

   # Update these paths for your system
   image_path = 'C:/path_to_your_image.jpg'  # Replace with the path to your test image
   tesseract_cmd = 'C:/Program Files/Tesseract-OCR/tesseract.exe'  # Path to Tesseract executable
   
   extracted_text = ocr_image(image_path, tesseract_cmd=tesseract_cmd)
   print("Extracted Text:")
   print(extracted_text)

3. Run your script:

   python test_script.py

Code Explanation

• Preprocessing Function:

  The preprocess_image function prepares the image for OCR by converting it to grayscale, sharpening it, and enhancing its contrast:

  def preprocess_image(image_path):
      image = Image.open(image_path).convert('L')
      image = image.filter(ImageFilter.SHARPEN)
      enhancer = ImageEnhance.Contrast(image)
      image = enhancer.enhance(2)
      return image

• OCR Function:

  The ocr_image function processes the image and then extracts the text using pytesseract:

  def ocr_image(image_path, tesseract_cmd=None):
      if tesseract_cmd:
          pytesseract.pytesseract.tesseract_cmd = tesseract_cmd
      image = preprocess_image(image_path)
      text = pytesseract.image_to_string(image, lang='eng')
      return text

Contributing

Contributions are welcome! Please open an issue or submit a pull request for any changes.

License

This project is licensed under the MIT License. See the LICENSE file for more details.

For more information, visit the image-text-reader library page on PyPI.