📄 PDF OCR Text Extractor

High-quality text extraction from scanned PDFs with Persian/Arabic (RTL) support.

This project provides a powerful Python script that extracts text from PDFs — especially scanned or image-based documents — using Tesseract OCR, advanced image preprocessing, and robust text-cleaning tailored for Persian/Arabic languages.
It outputs both clean UTF-8 text and HTML with proper RTL support.

---

✨ Features

🔍 OCR Extraction

• Works with scanned PDFs
• Supports Farsi, Arabic, and English
• Runs multiple OCR configurations and selects the best output

🖼 Advanced Preprocessing

• Uses OpenCV (if available):
  • Adaptive thresholding
  • Noise reduction
  • Morphological cleanup
• Uses Pillow fallback when OpenCV is not installed

🇮🇷 RTL Persian/Arabic Handling

• Auto-detects Persian text
• Normalizes common OCR mistakes
• Fixes diacritics, punctuation, and spacing
• Converts digits contextually
• Uses arabic-reshaper + bidi for correct display

📝 Output Formats

• extracted_text_clean.txt
• extracted_text_clean.html
• Page-separated output

---

📦 Requirements

Install Python Dependencies

pip install pytesseract PyMuPDF Pillow opencv-python arabic-reshaper python-bidi numpy

Install Tesseract OCR

Windows: https://github.com/UB-Mannheim/tesseract/wiki

Ubuntu :

sudo apt-get install tesseract-ocr tesseract-ocr-fas

MacOs:

brew install tesseract
brew install tesseract-lang

Usage

Place the PDF into the same directory and name it: Sample.pdf

Run:

python pdf_ocr.py