Scribe

Scribe is a powerful OCR (Optical Character Recognition) application that extracts text from images using EasyOCR. It provides both a command-line interface (CLI) and a graphical user interface (GUI) for different use cases.

Features

• High Accuracy OCR: Uses EasyOCR with GPU acceleration for fast and accurate text extraction
• Image Processing: Automatically enhances images with resizing, sharpening, and noise reduction
• Dual Interface: Both CLI and GUI options for different workflows
• Clipboard Integration: Automatically copies extracted text to clipboard
• Multiple Image Formats: Supports PNG, JPG, JPEG, BMP, and GIF files
• Confidence Scoring: Shows confidence levels for each detected text segment

Examples

GUI

CLIE

Download

Download latest version.

For Devs

Install Dependencies

pip install easyocr opencv-python pyperclip termcolor click pillow

Run locally

# Run GUI
watchmedo auto-restart --patterns='scribe/gui.py;core.py' --recursive -- python gui_entry.py

# Run CLI
python cli_entry.py path/to/image.png

Build Executables (Optional)

The project includes PyInstaller spec files for building standalone executables:

# Build CLI executable
pyinstaller cli_optimized.spec

# Build GUI executable
pyinstaller gui_entry.spec

# Build both
pyinstaller build.spec

Create installers

Windows

First, make sure Inno is installed.

cd installers ; & "C:\Program Files (x86)\Inno Setup 6\ISCC.exe" scribe.iss

This will generate installers/dist/scribe_installer.exe.

Usage

Command Line Interface (CLI)

The CLI application is perfect for batch processing, automation, and quick text extraction from command line.

Building Executables

# Install PyInstaller
pip install pyinstaller

# Build optimized CLI
pyinstaller --onefile cli_optimized.py

# Build GUI
pyinstaller --onefile gui_entry.py

Performance Tips

• GPU Usage: Ensure CUDA is properly installed for GPU acceleration
• Image Quality: Higher resolution images generally produce better results
• Image Preprocessing: The application automatically enhances images, but clean, well-lit images work best
• Text Orientation: Ensure text is horizontal for best results

Troubleshooting

Common Issues

1. CUDA/GPU Issues: If GPU initialization fails, the application will fall back to CPU processing
2. Image Loading: Ensure the image file exists and is in a supported format
3. Memory Usage: Large images may require significant memory; consider resizing very large images

License

This project is open source. Please check the license file for details.