Passive cross-wind and turbulence sensing with neuromorphic camera based imaging

Abstract

Method comprising: detecting, using a neuromorphic camera, events indicating change of optical flux; determining, based on the events, an optical shift indicating a wind speed or turbulence profile; determining, based on the optical shift, an estimated trajectory of a projectile. Edge detection or user selection may be used to identify a contrast region in the camera’s field of view. Atmospheric turbulence along a path toward an object in the field of view may be determined. A rifle scope may direct light from the path to the camera sensor. Features in contrast regions may appear laterally shifted due to cross wind, triggering event data. Multiple regions may correspond to objects at different distances, enabling event detection at different focal planes. Spatio-temporal variations in wind velocity and direction may help identify optical distortions along the imaging path. Neuromorphic cameras may generate time-series event data for each camera pixel at MHz rate, with microsecond temporal resolution and high dynamic range. Events detected at each pixel include changes in intensity and optical flux above a threshold, resulting in generation of a timestamped positive or negative polarity voltage signal. The polarity depends on whether intensity has increased or decreased. Artificial intelligence may be used. [Fig. 1]