Aerial data enhancement methods enhanced by simulation for autonomous driving, non-temporal data storage media readable by computers for enhanced aerial data enhancement by simulation for autonomous driving, and aerial data enhancement systems enhanced by simulation for autonomous driving.

Abstract

The invention relates to a method for enhancing aerial data by simulation for autonomous driving, which includes the following steps: acquiring aerial sensor information of a field of view containing multiple ground vehicles and road objects; analyzing aerial sensor information, according to a first recognized driving scenario, to provide a realistic behavioral profile of road traffic objects in the first recognized driving scenario; generating, based on the analysis through simulation of different road traffic objects with corresponding realistic behavioral profiles, simulated aerial information containing simulated behavioral information related to road traffic objects in the first recognized driving scenario; enhancing sensor information by supplementing simulated aerial information, to produce enhanced aerial information related to the first recognized driving scenario;and provides enhanced aerial information for use in autonomous flight, wherein at least one of the steps of analysis, generation, and enhancement is performed by a machine learning process. The invention also relates to a computer-readable, non-temporal data storage medium for enhancing simulated enhanced aerial data for autonomous flight, and a simulation-enhanced enhanced aerial data system for autonomous flight.

Images