Among other things, we describe techniques for automatic annotation of environmental features in a map during navigation of a vehicle. The techniques include receiving, by the vehicle located within an environment, a map of the environment. Sensors of the vehicle receive sensor data and semantic data. The sensor data includes a plurality of features of the environment. A geometric model of a feature of the plurality of features is generated. The feature is associated with a drivable area within the environment. A drivable segment is extracted from the drivable area. The drivable segment is segregated into a plurality of geometric blocks, wherein each geometric block corresponds to a characteristic of the drivable area and the geometric model of the feature includes the plurality of geometric blocks. The geometric model is annotated using the semantic data. The annotated geometric model is embedded within the map.

A method for automatically marking a U-turn lane line includes: obtaining information about an intersection entry point, information about an intersection exit point, and information about an obstacle at a to-be-marked intersection; and marking a U-turn lane line with a first depth. The first depth is determined based on a second depth of the obstacle. The first depth of the U-turn lane line is determined based on the second depth of the obstacle, so that a marked U-turn lane line does not collide with the obstacle, to mark a more proper U-turn lane line.

A driving control method comprises: acquiring a destination of a vehicle; referring to a first map that includes identification information of a travel lane and a second map that does not include the identification information of the travel lane; calculating a route from a current position of the vehicle to the destination; when traveling along a first route included in the route and belonging to the first map, setting first driving control, while when traveling along a second route included in the route and belonging to the second map, setting second driving control with a lower level of autonomous driving than that of the first driving control; and creating a driving plan for the vehicle to travel along the route with contents of the set driving control. A driving control apparatus is based on the method.

A loss of map information is verified in a short time. Provided is a map verification method in which an arithmetic operation device verifies map data distributed to a control device mounted on a vehicle, wherein the arithmetic operation device executes: a step of searching for an end point of a link included in the distributed map data and existing within a predetermined range in a vicinity of the vehicle; and a step of determining whether there is a loss in the distributed map data in accordance with a result of collation as to whether the searched end point is a true end point.

The present disclosure relates to a vehicular electronic device including a power supply configured to supply power, an interface configured to receive HD map data on a specific area and data on the location of a vehicle from a server through a communication device, and at least one processor configured to continuously generate electronic horizon data on a specific area based on the high-definition (HD) map data in the state of receiving the power, wherein, upon determining, based on the data on the location of the vehicle, that the traveling lane of the vehicle located at an intersection is different from a lane corresponding to a main path included in the electronic horizon data, the at least one processor changes the electronic horizon data based on the traveling lane.

A system and method including identifying lane line data associated with a road within the sensor data; modelling a geometry of the road as a sequence of road segments, each road segment being defined by parameters including a curvature rate and a road grade rate; generating, based on a mathematical representation of the modelled road geometry, an approximation of each road segment; and generating, based on the generated approximation of each road segment, a three-dimensional representation of the road including the sequence of segments.

At a computing system comprising one or more processors and memory, the computing system receives road data collected on a moving vehicle along a road, the road data comprising a two-dimensional streetscape image, a three-dimensional point cloud, and inertial navigation data, identifies, within the two-dimensional streetscape image, a ground region image corresponding to the road based on a spatial position relation of the two-dimensional streetscape image and the three-dimensional point cloud according to the inertial navigation data, and detects at least one target road traffic marking in the ground region image, determining three-dimensional coordinates of the at least one target road traffic marking based on the spatial position relation of the two-dimensional streetscape image and the three-dimensional point cloud.

A computer-implemented method for validating the existence of roadwork is provided. The method comprises, for example, retrieving information for at least one segment of a road captured by a plurality of vehicles. The information comprises at least two of lane marking data, speed funnel presence data, and traffic behavior change data. The method also comprises generating a confidence score based on analysis of the retrieved information. The method further comprises validating the existence of the roadwork on the at least one segment of the road based on the generated confidence score.

An approach is provided for confirming road vector geometry based on aerial image(s). For example, the approach involves retrieving a feature and a vector representation of a road link. The approach also involves processing one or more aerial images depicting the road link to extract a list of spectral pixel values corresponding to the vector representation. The approach further involves determining a degree of misalignment between the spectral pixel values and a spectral signature of the feature of the road link. The approach further involves initiating a confirmation of a geometry of the vector representation based on the degree of misalignment. The approach further involves providing the confirmation as an output.

An apparatus for generating a lane network includes a processor configured to calculate, for each of individual points in a predetermined region connecting an entry lane and an exit lane, a probability distribution indicating a probability that the point lies on a standard trajectory, based on trajectories of one or more vehicles that traveled through the region, generate two virtual borders each connecting the positions of respective edges of the entry lane and the exit lane, set, of the individual points in the predetermined region, at least two points that lie between the two virtual borders and at which the variance of the probability distribution is not greater than a predetermined variance threshold, as control points, and generate a Bézier curve, based on the at least two control points, as a line connecting two lanes in the predetermined region in a lane network.