Method for spatial characterization of at least one vehicle image

摘要

A method for the spatial characterization of at least one vehicle image (30) of image information (200), the image information (200) comprising the vehicle image (30) of an external vehicle (1) and an environment image (40) of an environment (6) of the external vehicle (1), The following steps are carried out: - determining a bounding frame (230) for the vehicle image (30) in order to use the bounding frame (230) for delimiting the vehicle image (30) from the image of the surroundings (40), - determining a dividing line (240 ) for the bounding frame (230) in order to use the dividing line (240) for a division of the vehicle image (30) into at least two vehicle sides (20, 21), - determining the spatial characterization on the basis of the bounding frame (230) and the dividing line (240 ), wherein for determining the bounding frame (230) and the dividing line (240) at least one evaluation means based on machine learning, in particular a neural network (210) is applied.