choreopath

Transform human movement into generative art. A Python tool that uses MediaPipe to track body poses from video and creates SVG visualizations of motion trajectories - perfect for pen plotters, laser cutters, and digital art.

Features

• Track 33 body landmarks using MediaPipe Pose
• Process video files with OpenCV to extract tracking data
• Export tracking data to CSV with coordinates and visibility scores
• Generate SVG visualizations with hierarchical body part grouping, ready to be used in Inkscape
• Analyze tracking data quality with detailed reports
• Create animated visualizations of pose tracking
• Command-line interface

Installation

choreopath is pushed to Pypi at each new release. You can pip install choreopath to install it. Beware though, some versions of Python are incompatible with it as MediaPipe isn't currently available for Python >=3.13.

Instead, it is suggested to use uv and run choreopath like this:

uvx --python 3.12 choreopath

Usage

Track poses from video

choreopath track video.mp4 tracking_data.csv

Track with custom confidence thresholds

choreopath track video.mp4 tracking_data.csv --min-detection-confidence 0.7 --min-tracking-confidence 0.7

Generate SVG visualization

choreopath draw tracking_data.csv output.svg

Generate SVG with custom dimensions

choreopath draw tracking_data.csv output.svg --width 1920 --height 1080 --min-visibility 0.7

Generate an overlay video

choreopath overlay video.mp4 video-overlay.mp4

Analyze tracking data

choreopath analyze tracking_data.csv --animation output.mp4 --fps 30

Origin Story

This project was created when working on a generative art project. I used generated videos of dancers to create SVG that I could then plot using a pen plotter.

See example files:

original video overlay video animated tracking data examples/fall-recovery-4.svg

Development

• Use uv
• Run smoke test by using make smoke-test

Changelog

See CHANGELOG.md.

License

This software is under a MIT license. See LICENSE.md.