Event-based Lagrangian particle tracking for turbulent shear flows

Kurzfassung

The application of event-based cameras in particle tracking velocimetry enabled cost- and energy efficient measurement setups with low latency and superior temporal resolution, naturally meeting the high demands of measuring turbulence and shear flow phenomena. However, handling the asynchronously recorded event-stream poses new challenges on data processing. In this work, we evaluate the performance of different processing strategies by using the modular framework for spatiotemporal event-based Lagrangian particle tracking (STELLA) recently developed. Besides deterministic algorithms, artificial neural networks tailored for object detection based on dense representations of the event-stream are used for particle detection. To reveal the versatile applicability of event-based particle tracking velocimetry to turbulent fluid flows, a synthetic as well as different experimental datasets are derived, spanning from the turbulent flow in the wake of a cylinder up to the shear flow in the turbulent boundary layer of a high-Reynolds pipe flow.