Abstract
We present a robust, non-invasive strategy to optimize the detection and tracking of Brownian microparticles using event-based cameras inspired by neuromorphic vision, enhancing their functionality beyond the internal sensor settings. By introducing artificial sway into the sensor plane with a steering mirror, we significantly increase the event recording rate, thereby improving the spatiotemporal resolution of the tracked particles. From the spatial distribution of detected events, we identify the positions of isolated particles without prior knowledge of their shapes, bypassing the limitations of tracking algorithms based on particle centroids. In our experiment, we modulated the mirror at 1 kHz, achieving up to a 400-fold enhancement in temporal resolution. To test our method, we characterize the Brownian motion of a microparticle by calculating the variance of its position and estimating the diffusion coefficient of the medium at various temperatures through the mean-square displacement calculation. Using permutation entropy, we confirm that our modulation does not affect the stochastic nature of the particle movement.