📰 Newspaper OCR Enhancement & Text-Layering Tool

This project enhances the readability and text extraction capability of scanned newspaper PDFs. It first applies color-preserving image preprocessing and then adds an invisible OCR-based text layer so that users can search and copy text from the resulting PDF without altering the original visual appearance.

---

📌 Features

✅ High-resolution image conversion (300 DPI)
✅ Color-preserving contrast enhancement & denoising
✅ Image reassembly into a PDF while maintaining full color
✅ Text-layer embedding using ocrmypdf
✅ Searchable and selectable final PDF output
⚠️ Note: You should download output pdf files to try text extraction

---

📅 Requirements

Python 3.x

Python packages:

• ocrmypdf
• pdf2image
• opencv-python
• numpy
• Pillow

Poppler utilities:

• macOS: brew install poppler
• Ubuntu: sudo apt-get install poppler-utils

---

🔍 Main Function Explanations

• convert_from_path() : Converts PDF pages to high-resolution color images
• cv2.convertScaleAbs() : Enhances contrast and brightness
• cv2.fastNlMeansDenoisingColored() : Denoises while preserving color
• Image.fromarray() : Converts NumPy array back to PIL image for saving
• ocrmypdf.ocr() : Adds an invisible, selectable OCR text layer to the PDF

---

🔄 How to Run

1. Place the original scanned newspaper PDF in the ./data folder.

2. Update the file name in the script (input_pdf variable).

3. Run the Python script:

   python main.py