SimpleAnnotate

A napari plugin for simple image and video annotation that provides three main annotation workflows:

1. Bounding Box Annotation (YOLO format): For object detection training data on images
2. Video Bounding Box Annotation: For object detection training data on video files
3. Image Classification Labeling: For image classification training data

---

This napari plugin was generated with Cookiecutter using @napari's cookiecutter-napari-plugin template.

Installation¶

You can install napari-simpleannotate via pip:

pip install napari-simpleannotate

To install latest development version :

pip install git+https://github.com/hiroalchem/napari-simpleannotate.git

How to use¶

Bounding Box Annotation (for images)¶

1. Opening Files or Directories:

   • Click the Open File button to open an image file.
   • Click the Open Directory button to open a directory containing images.
   • If there's a class.yaml in the directory of the selected file or within the selected directory, it will be automatically detected. A popup will appear, giving you the option to load it.

2. Class Management:

   • Enter the class name in the textbox and click the Add class button to add a class. When adding a class name, a number is automatically assigned to it. This number will be used when saving annotations.
   • Select a class from the class list and click the Delete selected class button to remove it.

3. Annotating Images:

   • Use napari's rectangle tool to annotate the images. If you have a class selected, the annotation will automatically be assigned to that class.
   • For existing rectangles, you can change their class by selecting the rectangle and then choosing a different class from the list.

4. Saving Annotations:

   • Click the Save Annotations button to save the annotations in YOLO format.
   • Along with saving the annotations, the class.yaml will also be saved. If a class.yaml already exists and its content is different from the current one, a popup will appear asking for confirmation to overwrite it.

Video Bounding Box Annotation¶

1. Opening Videos:

   • Click the Open Video button to open a video file (supports MP4, AVI, MOV, MKV, WMV, FLV, WebM formats).
   • The video will be loaded and displayed in napari's time-aware viewer.

2. Navigation:

   • Use napari's time slider to navigate between frames.
   • Current frame information is displayed in the widget.

3. Class Management:

   • Same as image annotation: add/delete classes with automatic ID assignment.
   • Classes are saved to class.yaml in the video directory.

4. Annotating Videos:

   • Navigate to the desired frame using the time slider.
   • Use napari's rectangle tool to create bounding boxes.
   • Each annotation automatically includes the current frame number.
   • Annotations are frame-aware and will only be visible on their respective frames.

5. Saving Annotations:

   • Click Save Annotations to save annotations in extended YOLO format.
   • Annotations are saved as video_name.txt with format: class_id frame x_center y_center width height
   • Frame images are automatically extracted and saved alongside annotations.

Image Classification Labeling¶

1. Opening Directory:

   • Click the Open Directory button to select a directory containing images.
   • Supported formats: PNG, TIF, JPG, JPEG, TIFF (recursive search).

2. Display Options:

   • Check Split Channels to display multi-channel images as separate layers.
   • Contrast settings are preserved when switching between images.

3. Class Management:

   • Enter class names in the text box and press Enter to add/remove classes.
   • Classes are automatically saved to class.txt in the target directory.

4. Labeling Workflow:

   • Select an image from the file list to display it.
   • Click on a class name to assign that label to the current image.
   • Labels are automatically saved to labels.csv in real-time.

5. Resume Capability:

   • Previous labels and classes are automatically loaded when reopening a directory.
   • The workflow can be resumed from where you left off.

Contributing¶

Contributions are very welcome. Tests can be run with tox, please ensure the coverage at least stays the same before you submit a pull request.

License¶

Distributed under the terms of the BSD-3 license, "napari-simpleannotate" is free and open source software

Issues¶

If you encounter any problems, please file an issue along with a detailed description.