42dot Inc.

The present disclosure relates to a method and apparatus for generating a lane polyline by using a neural network model. The method according to an embodiment may extract a multi-scale image feature by using a base image obtained from at least one sensor loaded in a vehicle. According to the method, the multi-scale image feature is input to a first neural network model as input data and a BEV feature may be obtained as output data from the first neural network model. Also, according to the method, the BEV feature may be input to a second neural network model as input data and a polyline image with respect to a certain road may be obtained as output data from the second neural network model. In the present disclosure, a lane polyline obtained from the neural network may be utilized in vehicle control without going through an additional treatment process.

The present disclosure relates to a method and apparatus for performing multipath search. The method according to an embodiment of the present disclosure may calculate a distance between an origin point and a destination point in a group of destination points. In addition, the method may generate a travel path between the origin point and the destination point using a first path-finding method when the distance between the destination point and the origin point is less than a threshold value, and generate the travel path between the origin point and the destination point using a second path-finding method when the distance between the destination point and the origin point is greater than or equal to the threshold value.

An embodiment provides an apparatus for determining a vehicle boarding possible area for a driving image using an artificial neural network, including: an image segmentation module that obtains a driving image for a driving direction of a vehicle from a camera module and segments the driving image into a plurality of image strips; a pre-trained boarding availability classification artificial neural network module that uses the image strip as input information and boarding availability information for the image strip as output information; a feature extraction module that extracts an activation map including feature information on the image strip from the boarding availability classification artificial neural network module; and an area information generation module that generates boarding possible area information for the image strip based on the feature information included in the activation map.

The present disclosure relates to a method and apparatus for determining a movement path of a vehicle considering a movement flow of passengers. In the method according to an embodiment of the present disclosure, boarding/alighting information for each time slot for each of a plurality of boarding/alighting points may be obtained, and departure/arrival information for each passenger may be obtained based on the boarding/alighting information for each time slot. In addition, in the method according to an embodiment of the present disclosure, at least one passenger movement flow vector for each time slot may be determined based on the departure/arrival information for each passenger, and a movement path of a vehicle may be determined based on the passenger movement flow vector for each time slot.

Provided is a training method for determining a driving route of a vehicle. The training method includes extracting a latent vector based on trajectories corresponding to maneuver modes that a vehicle is capable of selecting in a driving situation and training a model for generating route distributions to determine a driving route of the vehicle, based on a driving dataset and the latent vector, in which the trajectories are generated by applying, to a route search algorithm, weights of indicators for maneuvers differently according to the maneuver modes.

The present disclosure relates to a method and apparatus for generating a travel route. The method according to an embodiment of the present disclosure may include obtaining first information related to first group points, calculating first estimated required times for moving between the first group points, based on the first information, and recording the first estimated required times in a first area of a weight matrix. In addition, the method may further include obtaining second information related to a second group point, calculating, based on the first information and the second information, second estimated required times including estimated required times for moving from the first group points to the second group point, and estimated required times for moving from the second group point to the first group points, and recording the second estimated required times in a second area of the weight matrix. In addition, the method may further include generating a travel route based on the first estimated required times and the second estimated required times.

The disclosure relates to a method and an apparatus for determining a lane centerline network. A lane centerline network determination method according to an embodiment of the disclosure may include obtaining a driving trajectory of a vehicle by using a global positioning system (GPS) mounted on the vehicle, matching a road map with the driving trajectory, generating road feature information considering the driving trajectory based on an image captured by an image capturing device mounted on the vehicle, determining a position of a section node dividing a road on the road map into a plurality of sections based on the road feature information, determining a lane centerline placement for each section by determining a lane centerline for the traveling lane of the vehicle among a plurality of lanes included in the section based on the driving trajectory, and estimating lane centerlines for remaining lanes, determining a final longitudinal position of the section node based on reliability of the section node for each of driving trajectories through which the vehicle has passed a same road repeatedly, determining a section network connection relationship between the section, a previous section, and a next section based on the lane centerline placement for each section, determining a geometry of the lane centerline placement for each section based on reliability of a section link connecting the section nodes to each other, and generating a lane centerline network based on the section network connection relationship and a lane centerline geometry within the section.

A vehicle control apparatus and method are disclosed. A vehicle control apparatus for autonomous driving includes a receiver configured to receive a local path based on a trajectory planner, and a controller configured to generate a steering command for lateral control of the vehicle based on the local path, generate a compensated steering command by compensating the steering command based on a movement state of a vehicle to be controlled, and control the vehicle based on the compensated steering command.

Disclosed are a method of generating a circular traveling route of a vehicle, a device for generating a circular traveling route of a vehicle, and a recording medium in which a program for executing the method of generating a circular traveling route of a vehicle. The present disclosure is implemented through a process of receiving a request for generating a circular traveling route including information about a list of roads to be circulated, information about the number of repeated travels on the roads to be circulated, information about a preferred traveling direction at a road branch point, and information about a travelable time of the vehicle, grouping the list of roads to be circulated into a plurality of groups, determining a vehicle traveling sequence in the plurality of groups, and generating the circular traveling route of the vehicle on the basis of a vehicle traveling sequence in the determined plurality of groups, the information about the number of repeated travels on the roads to be circulated, and the information about the travelable time of the vehicle, by a route generation device. According to the present disclosure, an optimal circular traveling route for a vehicle, which needs to travel through the entire list of predetermined roads for a specific purpose, such as building a road map, obtaining image data, or performing road maintenance work, may be efficiently generated. In addition, according to the present disclosure, an optimal circular traveling route for a vehicle, which needs to repeatedly travel a list of predetermined roads a certain number of times depending on the purpose of vehicle travel, may be generated.

The present disclosure relates to a method and apparatus for performing multipath search. The method according to an embodiment of the present disclosure may calculate a distance between an origin point and a destination point in a group of destination points. In addition, the method may generate a travel path between the origin point and the destination point using a first path-finding method when the distance between the destination point and the origin point is less than a threshold value, and generate the travel path between the origin point and the destination point using a second path-finding method when the distance between the destination point and the origin point is greater than or equal to the threshold value.