YOLO Image Annotation Tool

A lightweight YOLO image annotation tool for creating and editing bounding box annotations. It supports drawing, deleting, and classifying objects with an intuitive interface. This tool allows you to efficiently label images with rectangular bounding boxes and assign classes, making it perfect for preparing training data for YOLO-based object detection models.

YOLO.Image.Annotation.Tool.mp4

Features

• YOLO Format Compatible: Creates annotation files in the format required by YOLO models.
• Intuitive Interface: Left-click to draw boxes, right-click to delete boxes, middle mouse button to define class.
• Automatic File Handling: Automatically reads and displays existing annotations, or creates new ones as needed.
• Multiple Image Format Support: Works with various image formats.
• Organized Output: Creates a dedicated Results folder in the same directory as your images folder.
• Persistence: Saves annotation data in text files compatible with YOLO frameworks.

Installation

Prerequisites

• Python 3.6+
• Required Python packages:

pip install opencv-python tkinter

Setup

1. Clone this repository:

git clone https://github.com/AminNasiri63/YOLO-Image-Annotation-Tool.git

2. Navigate to the repository directory:

cd YOLO-Image-Annotation-Tool

3. Install dependencies:

pip install opencv-python

Note: tkinter usually comes pre-installed with Python, but if not, you may need to install it separately.

Usage

1. Run the main script:

python main.py

2. When prompted, select the folder containing the images you want to annotate.

3. The tool will open and display the first image in the folder.

Controls:

• Left Mouse Button: Draw a bounding box.
• Right Mouse Button: Delete a selected bounding box.
• Middle Mouse Button: Define class for the currently selected box.
• S: Save current annotations.
• N: Skip image without saving.
• F: Select new folder.
• Q: Quit tool.

File Format

Annotations are saved in YOLO format text files with the same name as their corresponding image file but with a .txt extension. Each line in the text file represents one bounding box in the format:

<class_id> <x_center> <y_center> <width> <height>

where:

• class_id: Integer representing the class ID (0-indexed).
• x_center, y_center: Normalized coordinates (0-1) of the center of the bounding box.
• width, height: Normalized width and height of the bounding box (0-1).

Folder Structure

Your initial folder structure:

📁 Parent_Directory/
│
└── 📁 Selected_Image_Folder/
    ├── 📄 image1.jpg
    ├── 📄 image1.txt (if already annotated)
    ├── 📄 image2.png
    └── ...

After running the tool, the structure will be:

📁 Parent_Directory/
│
├── 📁 Selected_Image_Folder/
│   ├── 📄 image1.jpg
│   ├── 📄 image1.txt (if already annotated)
│   ├── 📄 image2.png
│   └── ...
│
└── 📁 Results/
    │
    └── 📁 Selected_Image_Folder_Name_Final/
        ├── 📄 image1.jpg
        ├── 📄 image1.txt
        ├── 📄 image2.jpg
        └── ...

The tool reads images from the Selected_Image_Folder and saves annotation files in the Results/Selected_Image_Folder_Name_Final directory.

Limitations

• Box Drawing Constraints: It is impossible to start drawing a new bounding box inside an existing box. When attempting to click inside an existing box, the tool will select or interact with that box instead of starting a new one.

License

MIT License.

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.