Realistic Noise Generation to Enhance Realism of Virtual Lidar Scans

Many real-world phenomena corrupt light detection and ranging (lidar) measurements, such as laser energy attenuation, variations in aerosol concentration and composition with height, and hard target returns. Accurate studies of lidar scans using virtual lidar methods should include some realistic model of these corrupting effects to generate more realistic simulations of lidar scans. We present a simple model that characterizes noise caused by energy attenuation and aerosol stratification. The model requires limited inputs and is developed for a Halo Photonics Streamline XR lidar but is readily generalizable for other lidar systems. A critical component of this model is a model of the standard deviation of measured wind speed as a function of the backscattered signal’s signal-to-noise ratio. We derive a general model for this behavior that can be adapted to different scan settings.