Python Video Annotator

Documentation on ReadTheDocs

The software is an application written in python, which provides a GUI for users to analyse and to take notes of events occurred in the videos.

Features:

• Timeline for video navigation.
• Multiple events edition in the timeline.
• Graphs visualization over time.
• Objects tracking path edition.
• Plugins based app.

Installation & Running

How to install:

1. Create a Virtual Environment (highly recommended) on your preferred Python distribution with Python 3.6:

   • example with Anaconda in Anaconda Prompt

   conda create -n videoannotator python=3.6
   
   ...
   
   conda activate videoannotator
   

2. Make sure you are in the just created virtual environment and install PythonVideoAnnotator with pip:

   pip install python-video-annotator
   

Intel Mac installation

MacOS Catalina

1. Start with a fresh Virtual Environment (highly recommended) on your preferred Python distribution with Python 3.6:

   • example with Anaconda in Anaconda Prompt

   conda create -n videoannotator python=3.6
   
   ...
   
   conta activate videoannotator
   

2. Install some of the dependencies that might present issues as:

   pip install opencv-python-headless pyqt5==5.14.1 pyqtwebengine==5.14.0
   

3. Install Python Video Annotator

   pip install python-video-annotator
   

MacOS Mojave

1. Start with a fresh Virtual Environment (highly recommended) on your preferred Python distribution with Python 3.8:

   • example with Anaconda in Anaconda Prompt

   conda create -n videoannotator python=3.8
   
   ...
   
   conta activate videoannotator
   

2. Install PyFormsGUI 5 first

   pip install pyforms-gui==5
   

3. Downgrade PyQt5 and QScintilla:

   pip install --upgrade qscintilla==2.11.4 pyqt5==5.14.1 pyqtwebengine==5.14.0
   

4. Install missing packages

   pip install opencv-python-headless PyOpenGL_accelerate
   

5. Install Python Video Annotator, ignoring dependencies (will be installed afterwards since it has a conflict with opencv version)

   pip install python-video-annotator==3.306 --no-deps
   

6. Install python-video-annotator dependencies except the one that created conflicts (older version of opencv-python)

   pip install geometry-designer modular-computer-vision-api modular-computer-vision-api-gui pypi-xmlrpc python-video-annotator-models python-video-annotator-models-gui python-video-annotator-module-background-finder python-video-annotator-module-contours-images python-video-annotator-module-create-paths python-video-annotator-module-deeplab python-video-annotator-module-distances python-video-annotator-module-eventstats python-video-annotator-module-find-orientation python-video-annotator-module-import-export python-video-annotator-module-motion-counter python-video-annotator-module-path-editor python-video-annotator-module-path-map python-video-annotator-module-regions-filter python-video-annotator-module-smooth-paths python-video-annotator-module-timeline python-video-annotator-module-tracking python-video-annotator-module-virtual-object-generator scipy send2trash simplejson sklearn
   

How to run:

1. After the installation completes, you can start Python Video Annotator with:

   start-video-annotator
   

For developers

Follow the steps described here

Developers

• Ricardo Ribeiro - collaborator of the Scientific Software Platform of the Champalimaud Foundation.
• Hugo Cachitas working - collaborator of the Innate Behavior Lab of the Champalimaud Foundation.
• Carlos Mão de Ferro - collaborator of the Scientific Software Platform of the Champalimaud Foundation.
• Luís Teixeira - collaborator of the Scientific Software Platform of the Champalimaud Foundation.

Collaboration

This project was initially developed in collaboration with the Innate Behavior Lab to correct the result of a Computer Vision software to track flies in an arena