Space in 3D - Immersive Analytics Module for a Python Based Visual Analytics Software

This was my bachelors thesis that I did together with Andreas Ambühl in 2021.

Flowchart of the glue plugin. From astro cloud data to mesh to AR.

The Seamless Astronomy Group at the Harvard-Smithsonian Center for Astrophysics developed an open-source visualization software written in Python called Glue. It can create linked data visualizations and is highly extendable via plugins for all sorts of data. Viewing and exploring 3D data on a 2D screen has some limitations and results in hidden structures. Recent developments in AR, i.e., its availability and affordability, increased the desire to view more content in AR.

To solve this, we wrote a Python plugin for Glue, which enables an AR-view of the data in Glue. A user only needs our plugin, a mobile-phone with the CoSpaces App installed and a Merge Cube, which serves as a 3D marker.

The 3D data in glue is exported as meshes and uploaded to CoSpaces its API. A room with the data will then be created and a QR code presented to the user that can be scanned with a mobile phone.

Our plugin for Glue, enabling an AR-view of the data using a phone and a Merge Cube

Appendix

Code Repository: Github Repo
Poster: Download Poster