OCR to Excel Converter

OCR to Excel Converter is a Python-based desktop application that processes images using different OCR engines (PaddleOCR, Tesseract, or EasyOCR) and converts the extracted data into an Excel file. The application features a graphical user interface (GUI) built using Tkinter and ttkbootstrap.

Table of Contents

• Features
• Installation
• Requirements
• Tesseract Installation
• Setup
• Usage
• Building the Executable
• Project Structure
• Troubleshooting
• License

Features

• Support for multiple OCR engines:
  • PaddleOCR (auto-downloads models)
  • EasyOCR (auto-downloads models)
  • Tesseract (requires manual installation)
• Confidence-based Excel highlighting
• Cross-platform support (Windows/macOS)
• GUI with image upload and screenshot capture
• Automatic path detection for Tesseract

Installation

1. Clone the repository:

   git clone https://github.com/a-sajjad72/Medical_OCR_SS_Tool.git
   cd Medical_OCR_SS_Tool

2. Set up a virtual environment:

   python -m venv venv

3. Activate the virtual environment:

   • Windows:

     venv\Scripts\activate

   • macOS/Linux:

     source venv/bin/activate

4. Install dependencies:

   The project contains two requirement files:

   • requirements.txt – List of all dependencies with their versions. (recommended to use)
   • requirements_paddle.txt – list of key dependencies for PaddleOCR.

   pip install -r requirements.txt

Requirements

Python Interpreter

• Requires Python 3.12 (3.12.6 or newer recommended)
• Minimum Supported Version: Python 3.12.0
• Tested Version: Python 3.12.6
• Download: python.org/downloads

Important Notes: ⚠️ PyTorch Compatibility: The application requires PyTorch, which currently only has official support for Python 3.12. Earlier versions (3.11 or below) are not supported due to dependency conflicts.

💡 Installation Tips:

• For Windows users: Check "Add python.exe to PATH" during installation
• For macOS/Linux users: Consider using pyenv for version management
• Verify installation: python --version

🔗 PyTorch Compatibility Reference:
Official PyTorch Python Support Matrix

Tesseract OCR

• Required for Tesseract OCR engine
• Version 5.3.0 or newer
• Installation guides below

Python Dependencies

All Python package requirements are listed in requirements.txt. Key dependencies include:

• PaddleOCR
• EasyOCR
• pytesseract
• OpenCV
• ttkbootstrap
• openpyxl
• pytorch

Tesseract Installation

macOS

# Install using Homebrew
brew install tesseract

# Verify installation
tesseract --version

Windows

1. Download installer from UB-Mannheim Tesseract
2. Run the installer with default settings:
   • Use recommended installation path i.e. (C:\Program Files\Tesseract-OCR)

Custom Installations

Create .env file in project root for custom paths:

TESS_BINARY_PATH=/path/to/tesseract
TESSDATA_PREFIX=/path/to/tessdata

Setup

1. Configuration:

   • For non-standard Tesseract installations, edit the .env file
   • Models for PaddleOCR/EasyOCR will auto-download on first run

2. Verify Paths:

   python -c "from utils import get_tessbin_path, get_tessdata_path; print(f'Tesseract: {get_tessbin_path()}\nTessdata: {get_tessdata_path()}')"

Usage

1. Run the application:

   python main.py

2. Application workflow:

   • Select OCR engine from dropdown
   • Adjust confidence thresholds (High/Medium)
   • Upload image or capture screenshot
   • Processed Excel file saves automatically
   • Results shown with bounding box visualization

Building the Executable

Automated Build Scripts

1. Windows:

   build_windows.bat

2. macOS/Linux:

   source build_mac.sh

Script Features:

• Automatically detects Tesseract installation paths
• Handles both system-wide and custom installations
• Maintains consistent resource paths between dev and prod
• Validates Tesseract presence before building

Manual Build Requirements:

# For reference - use scripts instead
python -c "from utils import get_tessbin_path, get_tessdata_path; print(f'--add-binary {get_tessbin_path()}:models/tesseract --add-data {get_tessdata_path()}:models/tesseract/tessdata')"

Why This Works:

1. Uses your existing path detection logic from utils.py
2. Maintains frozen application structure
3. Eliminates hardcoded paths
4. Automatically adapts to different installations
5. Preserves PyInstaller's resource bundling requirements

Troubleshooting

Tesseract Issues

• Path not found: Verify installation and check .env file
• Missing languages: Install tesseract-lang (macOS) or reinstall with additional languages (Windows)
• Version mismatch: Requires Tesseract 5.3.0+ t

Common Errors

• TESSDATA_PREFIX not set: Verify tessdata directory exists
• No module named...: Reinstall requirements.txt dependencies
• Permission denied: Run as administrator (Windows) or use sudo (macOS)

License

This project is licensed under the Creative Commons License. See the LICENSE file for details.