Comparison of Event-Based Alogrithms for Experimental Two-Dimensional, Two-Component Velocimetry

Event-based cameras, which operate without traditional frames and instead capture asynchronous changes in intensity, offer a transformative approach to flow diagnostics. This study investigates the application of event-based methods for velocimetry around an SD7003 wing and compares their performance to traditional particle image velocimetry (PIV). Four event-based algorithms [Weighted Tracking (Linear and Non-Linear fits), Kalman Filter, and Motion Compensation] are evaluated using a combination of freestream, near-surface, and wake flow scenarios. Results demonstrate that event-based methods successfully replicate global flow trends observed in PIV, including mean and median velocity estimates, while also capturing subtle flow features such as near-surface pressure gradients and wake asymmetry with higher resolution. Unlike PIV, which is constrained by frame rate and windowed processing, event-based methods track individual tracer particles, resolving flow features even in regions with steep gradients. The Weighted tracking methods exhibited the closest agreement with PIV, while Kalman Filter and Motion Compensation methods showed slight biases in U velocity, hypothesized to result from input parameter tuning. The findings highlight the potential of event-based methods as a viable candidate for flow diagnostics.