LabelCraft 1.8.14

LabelCraft - A modern graphical image annotation tool based on labelImg

LabelCraft - Modern Image Annotation Tool

LabelCraft is a powerful and user-friendly graphical image annotation tool, enhanced from the popular labelImg project. It provides an intuitive interface for creating bounding box annotations for object detection and machine learning tasks.

✨ Key Features

• Multiple Format Support: Export annotations in PASCAL VOC (XML), YOLO (TXT), CreateML (JSON), COCO (JSON), and CSV formats
• Project Management: Create and manage annotation projects with organized directory structures
• Smart Workflow: Pending queue system for efficient batch annotation
• Multi-language Support: English, Simplified Chinese, Traditional Chinese, Japanese, German, French
• Image Enhancement: Real-time brightness adjustment for better visibility
• Flexible Navigation: Zoom, pan, and keyboard shortcuts for fast annotation
• Predefined Classes: Load and manage custom class labels
• Annotation Verification: Mark verified annotations for quality control
• Dark/Light Mode: Automatic theme adaptation for comfortable viewing
• Auto-save: Optional automatic saving when switching images
• Cross-platform: Works on Windows, macOS, and Linux

🚀 Quick Start

Installation

Install via pip::

    pip install labelcraft

Or run from source::

    git clone https://github.com/syd168/LabelCraft.git
    cd LabelCraft
    pip install -r requirements.txt
    python main.py

Usage
~~~~~

Basic usage::

    labelcraft                              # Start without parameters
    labelcraft /path/to/images              # Open images from directory
    labelcraft --classes classes.txt        # Use custom class file
    labelcraft --save-dir ./output          # Set default save directory
    labelcraft --version                    # Show version

Command-line Options

• image_dir: Directory containing images to annotate (optional)
• --classes, -c: Path to predefined classes file
• --save-dir, -s: Default directory to save annotations
• --version, -v: Display program version

📖 Documentation

For detailed tutorials and documentation, visit:

• GitHub Repository <https://github.com/syd168/LabelCraft>_
• Installation Guide <https://github.com/syd168/LabelCraft/blob/master/doc/安装标注工具.md>_
• Tutorial (English) <https://github.com/syd168/LabelCraft/blob/master/doc/tutorial.md>_
• Tutorial (中文) <https://github.com/syd168/LabelCraft/blob/master/doc/tutorial_zh-CN.md>_

🛠️ Development

Requirements

- Python 3.8 or higher
- PySide6 >= 6.5.0
- lxml >= 4.9.0

Install dependencies::

    pip install pyside6>=6.5.0 lxml>=4.9.0

💡 Tips
------

- Press ``W`` to start drawing bounding boxes
- Use ``D`` and ``A`` to navigate between images
- Press ``Space`` to verify current annotation
- Use mouse wheel to zoom, drag to pan
- Adjust brightness with Ctrl+Shift+[/]

📄 License
---------

MIT License - See LICENSE file for details

🙏 Acknowledgments
-----------------

This project is based on the original `labelImg <https://github.com/tzutalin/labelImg>`_ by TzuTa Lin.

Note: This project is actively maintained and enhanced with modern features.


History

1.8.8 (2025-04-12)
------------------

* Fixed i18n: Added get_str() instance method for all dialogs
* Fixed export: Enhanced image path detection for YOLO and CSV formats
* Fixed export: Added SameFileError check when copying images
* Fixed project close: Auto-save current image annotation before closing
* Updated start scripts to English output for international users
* Improved README with pip installation guide
* Removed deprecated language files (fa-FA, de-DE)
* Enhanced annotation converter with multi-location image search

1.8.7 (2024-XX-XX)
------------------

* Based on labelImg project
* Added modern UI improvements
* Enhanced multi-language support
* Added brightness adjustment feature
* Improved annotation workflow
* Added GitHub Actions CI/CD

Note: This project is a fork and enhancement of the original labelImg project.
=======

1.8.6 (2024-XX-XX)
------------------

* Based on labelImg project
* Added modern UI improvements
* Enhanced multi-language support
* Added brightness adjustment feature
* Improved annotation workflow
* Added GitHub Actions CI/CD

Note: This project is a fork and enhancement of the original labelImg project.