Abstract: A method for constructing a lane-level navigation map is provided. The method may include: determining a change position of a lane in a road-level electronic map; extracting lane information of the change position of the lane, the lane information including a lane connectivity relationship; and storing the lane information in the road-level electronic map to obtain the lane-level navigation map.

Abstract: A method and apparatus for generating a direction identifying model, a device, a medium, and a program product are provided. The method includes: acquiring direction-targeted road test data corresponding to a target road, and a guide arrow sign and an accessible road-direction corresponding to the target road; and training a machine learning model by using the road test data and the guide arrow sign as an input of the direction identifying model, and using the accessible road-direction as an output of the direction identifying model, to obtain the direction identifying model.

Abstract: This application discloses a road data processing method, an apparatus, a device and a storage medium, and relates to the technical fields of intelligent traffic in data processing, big data and cloud computing. The specific implementation solution is: obtaining shape information of a target road and shape information of a reference road; generating a first segment sequence according to the shape information of the target road; generating a second segment sequence according to the shape information of the reference road; and recognizing parallel road segments in the target road and the reference road according to the first segment sequence and the second segment sequence. Thus, by converting the target road and the reference road into sequences of segments respectively, parallel road segments are recognized in the target road and the reference road based on the sequences of segments, the efficiency and accuracy are improved in recognizing parallel roads.