High Resolution and High-Speed Live Optical Flow Velocimetry

Abstract

Particle Image Velocimetry (PIV) typically relies on cross-correlation,which makes it difficult to obtain instantaneous velocity fields that are both spatially dense and available in real time at high acquisition rates. Optical Flow Velocimetry (OFV) offers a per-pixel alternative. Here we demonstrate real-tome OFV that delivers dense velocity fields (one vector per pixel) with high effective spatial resolution at frequencies up to the kHz range. Using synthetic particle images for two benchmarks -- a Rankine vortex and a homogeneous isotropic turbulence DNS -- we show that, with suitable particle seeding, OFV can resolve strong displacement gradients down to small scales. We then achieve real-time performance through algorithmic refinements and GPU-focused optimizations, combined with practical choices of OFV parameters. With this implementation, 32 Mp fields are processed live at 90 Hz, 4 Mp fields up to 460 Hz, and 1 Mp fields up to 1400 Hz. The method is further validated experimentally on the flow past a circular cylinder, where dense instantaneous velocity fields support real-time computation of derived quantities over long durations. These capabilities enable in-experiment monitoring, recovery of low-frequency dynamics from sustained high-rate acquisition, and closed-loop-flow-control strategies based on OFV measurements while also accelerating conventional post-processing to reduce turnaround time and computational cost.