A neuromorphic camera for tracking passive and active matter with lower data throughput

TL;DR

This paper demonstrates that neuromorphic event-based cameras can effectively track passive and active microscopic particles, providing comparable results to traditional cameras while significantly reducing data size, thus benefiting long-term and edge computing applications.

Contribution

The study introduces the use of neuromorphic cameras for tracking microscopic matter, achieving similar accuracy to conventional cameras with much lower data throughput, especially useful for long-term studies.

Findings

Equivalent tracking accuracy to sCMOS cameras.

Up to two orders of magnitude reduction in data size.

Effective tracking of both passive and active matter.

Abstract

We demonstrate the merits of using a neuromorphic, or event-based camera (EBC), for tracking of both passive and active matter. For passive matter, we tracked the Brownian motion of different micro-particles and estimated their diffusion coefficient. For active matter, we explored the case of tracking murine spermatozoa and extracted motility parameters from the motion of cells. This has applications in enhancing outcomes for clinical fertility treatments. Using the EBC, we obtain results equivalent to those from an sCMOS camera, yet achieve a reduction in file size of up to two orders of magnitude. This is important in the modern computer era, as it reduces data throughput, and is well-aligned with edge-computing applications. We believe the EBC is an excellent choice, particularly for long-term studies of active matter.