Tamil OCR Desktop Application

This is a desktop application for performing Optical Character Recognition (OCR) on images and PDF files, with a focus on Tamil and English languages.

Features

• Cross-Platform: Built with PyQt6 and can be compiled into a standalone executable for Linux.
• Image and PDF Support: Open various image formats (PNG, JPG, etc.) and multi-page PDF documents.
• Efficient PDF Processing: Converts PDFs to images in a separate thread to keep the UI responsive, with handling for large files.
• Parallel OCR: Utilizes multiple CPU cores to process pages in parallel, significantly speeding up OCR tasks.
• Tesseract Integration: Powered by the Tesseract OCR engine.
• Custom Models: Comes bundled with a custom Tamil Tesseract model (tam_cus) and the standard English model.
• Interactive Image Viewer:
  • View document pages with zoom and fit-to-screen controls.
  • Highlights recognized words with bounding boxes.
  • Toggle highlights on or off for better readability.
• Advanced OCR Controls:
  • Confidence Threshold: Adjust the minimum confidence level (0-100%) to filter out uncertain results. Changes are reflected in real-time.
  • Language Selection: Easily specify which Tesseract language models to use (e.g., tam_cus+eng).
• Rich Text Editor:
  • View and edit the extracted OCR text for proofreading and corrections.
  • The application tracks edited pages and allows you to reset the text back to the original OCR result.
  • Adjust the editor's font size for comfort.
  • Includes a custom Tamil font (marutham.ttf) for proper rendering.
• Export Functionality: Save the final, proofread text from all pages into a single .txt file.
• Drag and Drop: Quickly open files by dragging them onto the application window.

Setup and Installation

Prerequisites

• Python 3.x
• Tesseract OCR Engine (The application comes bundled with a Tesseract AppImage for Linux).

Steps

1. Clone the repository:

   git clone https://github.com/your-username/Tamil-OCR.git
   cd Tamil-OCR

2. Create a virtual environment (recommended):

   python3 -m venv .venv
   source .venv/bin/activate

3. Install the required Python packages:

   pip install -r requirements.txt

How to Run the Application

Once the setup is complete, you can run the application from the source code:

python main.py

How to Compile (Linux)

This project uses PyInstaller to create a standalone executable.

1. Install PyInstaller:

   pip install pyinstaller

2. Run the PyInstaller command:

The following command will bundle the Python script, assets (fonts, Tesseract data), and the Tesseract AppImage into a single executable file located in the dist directory.

pyinstaller --name "Tamil-OCR" \
            --onefile \
            --windowed \
            --add-data "font/marutham.ttf:font" \
            --add-data "tessdata/eng.traineddata:tessdata" \
            --add-data "tessdata/tam_cus.traineddata:tessdata" \
            --add-binary "binary/linux/tesseract/tesseract.AppImage:binary/linux/tesseract" \
            main.py

3. Make the executable runnable:

After building, you need to give the generated file execute permissions.

chmod +x dist/Tamil-OCR

4. Run the compiled application:

   ./dist/Tamil-OCR