Arabic-PDF-OCR-Text-Extraction

This project provides a Python-based solution to extract Arabic text from PDF documents using Google Document AI. It processes PDFs to generate formatted .txt files containing the extracted text.

Features

• High-Accuracy OCR: Employs Google Cloud Document AI for robust and scalable Optical Character Recognition of Arabic text in PDFs.
• Comprehensive Text Processing: Includes functions for:
• Normalizing Arabic text (removing tashkeel, etc.).
• Correcting common spelling errors.
• Removing unwanted characters and whitespace.
• Improving overall text readability.
• Optional Diacritization: Integrates Farasa for adding diacritics (tashkeel) to the extracted text, enhancing linguistic accuracy.
• Asynchronous Processing: Utilizes asyncio and concurrent.futures to process multiple files concurrently, significantly improving performance for large datasets.

Requirements

Ensure the following dependencies are installed:

• google-cloud-documentai
• PyPDF2
• python-dotenv
• arabic_reshaper
• python-bidi
• tqdm

Install these dependencies using the provided requirements.txt file:

pip install -r requirements.txt

Note: Ensure that you have access to Google Document AI and have set up the necessary authentication credentials.

Usage

1. Set Up Google Document AI Credentials: Follow the Google Cloud documentation to set up authentication and obtain your credentials.

2. Create .env File:

   • Create .env file and add
   • project_id=
   • location=
   • processor_id=

3. Configure the Scripts:

   • Specify the path to your input PDF file in main.py.

4. Run the Scripts:

   • Use main.py to extract text from PDF files. The extracted and formatted text will be saved as .txt files in the output directory.

Output

• Text Files: .txt files containing the extracted Arabic text, formatted for readability and ease of use.

Example

Here's how to set the input PDF path in the scripts:

# Set the path to the input PDF file in main.py
pdf_file_path = '/path/to/your/input.pdf'

# Run main.py 

After running the scripts, the extracted and processed text files will be saved in the output directory with the same name as the pdf file.

License

• This project is licensed under the MIT License.

Notes

• Ensure that your Google Cloud credentials are correctly set up and that you have the necessary permissions to use Document AI.
• The script is designed to handle PDFs containing Arabic text. For other languages, adjust the Document AI settings accordingly.

For more details and updates, visit the GitHub repository.