Machine-learning techniques for 3D particle reconstruction in dusty plasmas

Dusty plasmas provide an interesting system to study fundamental processes in many-particle systems since the particles can be imaged and followed on the kinetic individual-particle level.

We have performed experiments with dusty plasmas on parabolic flights using a stereoscopic camera system with four cameras. Under microgravity conditions the dust particles form a dense dust cloud, and a small fraction of the dust cloud is imaged by the four cameras.

In this contribution, techniques to reconstruct the three-dimensional position of the dust particles from the stereoscopic images with the help of machine-learning methods are reviewed and tested. This is important for a future application in the Compact facility planned for the ISS [1].

[1] C. Knapek et al., ”COMPACT - A new complex plasma facility for the ISS”, Plasma Phys. Control. Fusion 64 (2022) 12400