/Yolo_Annotator

Primary LanguagePython

YOLO Annotator - Image Annotation Tool

Author: Rukesh Duwal Date: January 30th, 2024 Occupation: Computer Vision Engineer at Crimson Tech License: MIT License

Description

The YOLO Annotator is a simple image annotation tool developed in Python using Tkinter for graphical user interface elements.

Key Features

  • Annotation of bounding boxes.
  • Keyboard shortcuts for quick navigation.

Main Program and Credits

Usage

  1. Run python main.py.
  2. Click LoadImage, select a folder that contains a list of images.
  3. For multi-class tasks, modify 'class.txt' with your class candidates. Before labeling a bounding box, choose the 'Current Class' in the Combobox or by pressing 1-9 on your keyboard.
  4. To create a new bounding box, left-click to select the first vertex. Move the mouse to draw a rectangle and left-click again to select the second vertex.
    • To delete the existing bounding box while drawing, press right click inside respective bbox or select it from the listbox and click Clear
    • To delete all existing bounding boxes in the image, click ClearAll or press r.
  5. After finishing one image, click Next or press d to advance. Click Prev or press a to reverse. Alternatively, input the index and click Go to navigate to an arbitrary image.
    • The labeling result will be saved in Labels/[folder name]/.. only if the 'Next' button is clicked.
    • The checkpoint of the last Image Number will be saved when the 'Next' button is clicked.
  6. Click Skip to skip an unwanted image from the directory and skip the annotation for that image (the skipped image path will be saved in log/skip.txt).
  7. Click ConvertYOLO or press c to convert the labeling result to YOLO format. The result will be saved in Result_YOLO/[folder name]/...
  8. Use the 'Add Class' and 'Delete Class' buttons in the UI to manage classes dynamically.

Additional Features

  • Bounding Box Count: The tool provides the count of the number of bounding boxes drawn inside image.
  • Color Assignment: Each class is assigned a unique color for better visualization.
  • Paste Last Bounding Box: Press 'v' to paste the last drawn bounding box where the mouse pointer is located.
  • Scroll Support: Users can scroll horizontally and vertically for larger images.

How to Run

  1. Make sure you have Python installed on your system.
  2. Clone this repository: git clone https://github.com/iamrukeshduwal/Yolo_Annotator.git
  3. cd Yolo_Annotator
  4. python main.py

License

This project is open-source and currently does not have a specific license. You are free to view, use, and modify the code for personal or educational purposes. If you plan to distribute or use this project in a commercial product, please consider adding an appropriate open-source license.