Non-motor vehicle and pedestrian behavior judgment method and system based on behavior state information

The invention relates to a non-motor vehicle and pedestrian behavior judgment method and system based on behavior state information. The method and system are used for solving the problem of understanding or predicting behaviors of other road users in a pedestrian and vehicle mixed driving environment of an automatic driving vehicle. According to the scheme, behavior state information of non-motor vehicles and pedestrians at the view angle of the vehicle is extracted by collecting pedestrian and vehicle mixed driving environment scene data at the view angle of the vehicle, and the behavior state information of the non-motor vehicles and the pedestrians comprises the speed, the position and the acceleration of the non-motor vehicles and the pedestrians and the speed and the position of the vehicle; non-motor vehicle target personalized behaviors are judged based on the non-motor vehicle behavior state information, the non-motor vehicle target personalized behaviors comprise non-motor vehicle left turning, righ