Event cameras provide microsecond-level temporal resolution for brightness changes, which makes them particularly suitable for event-based imaging velocimetry (EBIV). However, accurately estimating velocity vectors from event data is challenging because of its inherent asynchronous and sparse nature. This work proposes a Projection Concentration Maximization with Smooth Annealing (PCM-SA) method that estimates the velocity by maximizing the concentration of raw-event projections. Specifically, r