labelme: Image Polygonal Annotation with Python

Labelme is a graphical image annotation tool inspired by http://labelme.csail.mit.edu.
It is written in Python and uses Qt for its graphical interface,
and supports annotations for semantic and instance segmentation.

Requirements

• Ubuntu / macOS / Windows
• Python2 / Python3
• PyQt4 / PyQt5

Installation

There are options:

• Platform agonistic installation: Anaconda, Docker
• Platform specific installation: Ubuntu, macOS

Anaconda

You need install Anaconda, then run below:

# python2
conda create --name=labelme python=2.7
source activate labelme
conda install pyqt
pip install labelme
# if you'd like to use the latest version. run below:
# pip install git+https://github.com/wkentaro/labelme.git

# python3
conda create --name=labelme python=3.6
source activate labelme
# conda install pyqt
pip install pyqt5  # pyqt5 can be installed via pip on python3
pip install labelme

Docker

You need install docker, then run below:

wget https://raw.githubusercontent.com/wkentaro/labelme/master/scripts/labelme_on_docker
chmod u+x labelme_on_docker

# Maybe you need http://sourabhbajaj.com/blog/2017/02/07/gui-applications-docker-mac/ on macOS
labelme_on_docker examples/single_image/apc2016_obj3.jpg -O examples/single_image/apc2016_obj3.json
labelme_on_docker examples/semantic_segmentation/data_annotated

Ubuntu

# Ubuntu 14.04 / Ubuntu 16.04
# Python2
# sudo apt-get install python-qt4 pyqt4-dev-tools  # PyQt4
sudo apt-get install python-pyqt5 pyqt5-dev-tools  # PyQt5
sudo pip install labelme
# Python3
sudo apt-get install python3-pyqt5 pyqt5-dev-tools  # PyQt5
sudo pip3 install labelme

macOS

# macOS Sierra
brew install pyqt  # maybe pyqt5
pip install labelme  # both python2/3 should work

Usage

Run labelme --help for detail.
The annotations are saved as a JSON file.

labelme  # just open gui

# single image example
cd examples/single_image
labelme apc2016_obj3.jpg  # specify image file
labelme apc2016_obj3.jpg -O apc2016_obj3.json  # close window after the save
labelme apc2016_obj3.jpg --nodata  # not include image data but relative image path in JSON file
labelme apc2016_obj3.jpg \
  --labels highland_6539_self_stick_notes,mead_index_cards,kong_air_dog_squeakair_tennis_ball  # specify label list

# semantic segmentation example
cd examples/semantic_segmentation
labelme data_annotated/  # Open directory to annotate all images in it
labelme data_annotated/ --labels labels.txt  # specify label list with a file

For more advanced usage, please refer to the examples:

• Single Image Example
• Semantic Segmentation Example
• Instance Segmentation Example

Screencast

Testing

pip install hacking pytest pytest-qt
flake8 .
pytest -v tests

Acknowledgement

This repo is the fork of mpitid/pylabelme, whose development has already stopped.