S5-1 - Real-Time Multi-Particle Tracking at 10 kHz: Leveraging Binary Frames for Event-Based Vision

Event
23. ITG/GMA-Fachtagung Sensoren und Messsysteme 2026
2026-06-09 - 2026-06-10
Nürnberg
Band
Vorträge
Chapter
Event-Based Vision
Author(s)
J. Baßler, M. Heizmann - Karlsruhe Institute of Technology, Karlsruhe, P. Bäcker - Fraunhofer IOSB, Karlsruhe
Pages
82 - 89
DOI
10.5162/sensoren2026/S5-1
ISBN
978-3-910600-11-9
Price
free

Abstract

While event cameras provide microsecond temporal resolution, purely asynchronous tracking algorithms remain strongly load-dependent, scaling poorly with both instantaneous event-rate bursts and the number of simultaneous particles. We present a deterministically scalable framework that converts this variable tracking workload into a bounded, GPU-parallel detection task by aggregating events into ultra-short 10⁰ µs binary frames. Validated in a physical free-fall sorting scenario, our architecture sustains a strict 10 kHz tracking It scales to 4,000 simultaneous trajectories within the 10⁰ µs deadloop on standard consumer hardware. (1) core detection latency to guarantee the execution determinism required for line, achieving an effective industrial automation. O