Device and method for determining a lane boundary on a roadway section

Abstract

A device for determining lane boundaries on a roadway segment is described. The device is configured to determine a reference path for the roadway segment and, based on sensor data, to detect lane markings within that segment. Furthermore, the device is configured to determine arrangement information regarding the lane markings relative to the reference path and, based on this arrangement information, to determine whether the lane markings constitute lane boundaries or not.

Description

[0001] The invention relates to methods and corresponding devices which each make it possible to determine information relating to lanes of a roadway.

[0002] In order to partially or fully automate the longitudinal and / or transverse control of a vehicle, it is advantageous to have a highly accurate geographical map of the surrounding area. A single-precision map (SD: "standard definition"), which can be used, for example, to guide the vehicle to a predetermined destination, typically has an accuracy in the range of approximately one to ten meters. A high-precision map (HD: "high definition") should typically differ from reality by less than approximately one meter. HD and SD maps can differ even further beyond their accuracy, for example, in the information they contain.

[0003] For example, if the vehicle is traveling on a roadway with multiple lanes (also referred to as lanes), the HD map can indicate lane boundaries to distinguish which lane the vehicle is traveling in. Visual indicators, such as lane markings, can be used to determine lane boundaries. A lane marking typically consists of a line applied directly to the road surface (which can be detected visually, e.g., by a camera, and / or by sensors).

[0004] Creating an HD map using a surveying vehicle is laborious. Alternatively, an HD map can be created based on sensor-based lane markings, which are determined by a fleet of vehicles already traveling on the roadway.

[0005] An HD map can thus indicate the course of lane markings between one or more lanes on a roadway. This information can be used by a driving function to guide a vehicle at least partially or fully automatically along the roadway, in particular along a lane, longitudinally and / or transversely. For this purpose, a target trajectory for the vehicle can be determined based on the map data from the HD map. The target trajectory can, for example, run centrally between the lane markings indicated by the map data. The vehicle can then be guided longitudinally and / or transversely along the determined target trajectory.

[0006] Using the course of the lane markings to determine a target trajectory can result in a target trajectory being determined that runs between different lane markings, but does not run on an available lane (as may be the case, for example, on a motorway without a structural separation). Furthermore, in certain sections of a roadway, there may be no or insufficient number of sensor-detectable lane markings (for example, in front of a toll system, within an intersection, or on a country road). As a result, it may not be possible to determine a realistic target trajectory, or at least not a realistic one. Furthermore, a target trajectory determined based on the course of sensor-detected lane markings may be perceived as uncomfortable by the user of an automated vehicle.

[0007] This document deals with the technical task of efficiently and precisely determining map data from a digital (HD) map for a roadway that enables reliable and comfortable automated longitudinal and / or lateral guidance of a vehicle.

[0008] The problem is solved by each of the independent claims. Advantageous embodiments are described, among other things, in the dependent claims. It should be noted that additional features of a patent claim dependent on an independent patent claim can form a separate invention, independent of the combination of all features of the independent patent claim, without the features of the independent patent claim or only in combination with a subset of the features of the independent patent claim, which invention can be made the subject of an independent claim, a divisional application, or a subsequent application. This applies equally to technical teachings described in the description, which can form an invention independent of the features of the independent patent claims.In particular, it should be noted that the features of the devices and methods described in this document can be combined with one another in any way. In particular, the features of a first method or device can also be applied individually or in combination to another second method or device.

[0009] According to one aspect, a device for merging reference profiles for a road section is described. The road section can have one or more different lanes. A road section can, for example, have a length of between 20 and 50 meters. Furthermore, a road section can have a uniform direction of travel or two opposing directions of travel. In particular, the lanes of the road section can each have a uniform direction of travel or at least partially opposing directions of travel.

[0010] A roadway section can thus correspond to an area of ​​a (paved or unpaved) roadway that has one or more (possibly marked or unmarked) lanes. The individual lanes may be separated from each other, at least partially, by a structural measure. Furthermore, the individual lanes may at least partially have the same direction of travel and / or at least partially opposite directions of travel.

[0011] The device is configured to determine a first and a second reference profile for the road section. The individual reference profiles can each be determined using one of the methods described in this document. The individual reference profiles can each be determined based on measured travel paths (as described in this document).

[0012] The first and second reference profiles can be associated with different section sequences. The different section sequences can each comprise the roadway section and at least one surrounding roadway section arranged directly in front of and / or directly behind the roadway section in the direction of travel. Furthermore, the different section sequences can differ from one another in at least one surrounding roadway section. The different section sequences can be determined, for example, on the basis of a digital map (e.g., an SD card) in relation to the roadway network in the vicinity of the roadway section. For example, two or more, or five or more different section sequences can be considered.

[0013] The device is further configured to identify at least a first subsection of the roadway section in which the first and the second reference course satisfy a predefined parallelism criterion (and thus run at least almost parallel to each other).

[0014] The device can be configured to divide the first and second reference profiles of the roadway section into a sequence of reference point planes. The individual reference point planes can each be arranged perpendicular to the respective reference profile. Furthermore, the sequence of reference point planes (of the respective reference profile) can be arranged at a corresponding sequence of points along the first and second reference profiles. The successive points can follow one another at a uniform distance, approximately between 1 and 3 meters.

[0015] The parallelism criterion can depend on the relative orientation of corresponding support point planes of the first reference profile and the second reference profile. In particular, the device can be configured to identify, as a first subsection, a subsequence of points from the sequence of points for which the orientations of the corresponding support point planes of the first and second reference profiles deviate from each other by at most a predefined angular threshold value. The angular threshold value can be, for example, 10° or less.

[0016] The first reference path and the second reference path can be associated with opposite directions of travel. When considering the orientations of the corresponding reference point planes of the first and second reference paths, the directions of travel are preferably disregarded. In particular, a rotation of the orientation by 180° (due to an opposite direction of travel) is preferably disregarded.

[0017] Alternatively or additionally, the device can be configured to subdivide (only) the first reference profile of the roadway section into or by a sequence of support point planes. The individual support point planes can each be arranged perpendicular to the first reference profile. Furthermore, the sequence of support point planes can be arranged at a corresponding sequence of points along the first reference profile.

[0018] The device can be configured to determine, for each support point plane from the sequence of support point planes, an intersection point of the second reference profile with the respective support point plane. The parallelism criterion can then be determined in a particularly precise and robust manner based on the orientation of the individual support point planes relative to the orientation of the second reference profile at the corresponding intersection points.

[0019] The device can be configured, in particular, to identify, as a first subsection, a subsequence of points from the sequence of points for which the orientations of the corresponding support point planes and the orientations of the second reference path at the corresponding intersection points deviate from each other by at most a predefined angular threshold. The angular threshold can be 10° or less. When determining the orientation, the direction of travel of the respective reference path can be disregarded (in particular, for the determination of the parallelism criterion, it can be assumed that the two reference paths each have the same direction of travel).

[0020] The device is further configured to determine a fused reference profile for the first subsection on the basis of the first and second reference profiles and to use it instead of the first and second reference profiles for the first subsection of the roadway section.

[0021] The fused reference course can, for example, be provided as map data relating to the first sub-section of the roadway section for a digital map. Alternatively or additionally, the fused reference course can be used for automated longitudinal and / or lateral guidance of a vehicle (in the roadway section). Alternatively or additionally, one or more road markings in the first sub-section of the roadway section can be detected based on the fused reference course (e.g. using one of the methods described in this document). Alternatively or additionally, one or more lane markings and / or one or more lanes in the first sub-section of the roadway section can be detected based on the fused reference course.

[0022] By merging reference courses in one or more sections of a road section, map data for an HD map can be determined in a particularly resource-efficient manner.

[0023] The device can be configured to identify at least a second subsection of the roadway section in which the first and second reference paths do not meet the predefined parallelism criterion. The first and second reference paths can then be retained for the second subsection.

[0024] The device can thus be set up to determine a first and a second composite reference profile for the road section, the • have the common merged reference course for the first subsection; and • for one or more other sections of the roadway section (in which the parallelism criterion is not met) have the first or second reference course.

[0025] A first composite reference history can thus be provided, which • has the common merged reference history for the first subsection; and • for one or more other sections of the road section (in which the parallelism criterion is not met) has the first reference course.

[0026] Furthermore, a second composite reference history can be provided, which • has the common merged reference history for the first subsection; and • for one or more other sections of the road section (in which the parallelism criterion is not met) has the second reference course.

[0027] These composite reference trajectories may be provided as map data and / or used in any of the methods described in this document.

[0028] As already explained above, the device can be configured to divide the first reference profile in the first subsection into a sequence of support point planes. For the first reference profile and the second reference profile, a sequence of intersection points of the respective reference profile with the corresponding sequence of support point planes can be determined.

[0029] Furthermore, for each of the sequences of support point planes, a support point can be determined based on the determined intersection points with the respective support point plane. The support point for a support point plane can be determined as the mean, in particular as a weighted mean, of the determined intersection points with the respective support point plane.

[0030] The fused reference profile can then be determined in a particularly precise manner on the basis of the sequence of support points for the corresponding sequence of support point levels, in particular by sequentially connecting the support points by path segments (as exemplified in connection with Fig. 2d).

[0031] The device can be configured to determine a first number of measured travel paths used to determine the first reference profile and a second number of measured travel paths used to determine the second reference profile. The weights for determining the weighted average can be determined based on the first number and the second number. This allows the accuracy of the merged reference profile to be further increased.

[0032] Alternatively or additionally, the device can be configured to iteratively merge at least a partial section of each of two reference profiles from a set of reference profiles for the roadway section until a termination criterion is reached, in order to determine a fused reference profile in each case. The fused reference profile determined during one iteration can then be included in the set of reference profiles for the (directly) subsequent iteration.

[0033] For example, in a first step, a set of three or more reference profiles can be determined for the road section. Two reference profiles can then be selected during each iteration to merge at least a portion of the two reference profiles. The resulting merged reference profile can be included in the set of reference profiles for the subsequent iteration. In this way, the reference profiles can be further merged iteratively, e.g., until the parallelism criterion is no longer met for any possible pair of reference profiles.

[0034] It may happen that in one iteration, a fused reference profile is determined from a fused reference profile determined in a previous iteration. In particular, due to the iterative approach, it may happen that a fused reference profile is determined from more than two original reference profiles.

[0035] The device can be configured to determine the number of (original, unfused) reference curves from which the reference curves to be fused in an iteration were determined. The determined numbers can then be used as weights when merging the reference curves to be fused. This can further increase the accuracy of the fused reference curves.

[0036] According to a further aspect, a device for adapting reference profiles at a transition between two consecutive roadway sections is described. The device is configured to determine a first reference profile for a first roadway section and to determine a second reference profile for a second roadway section directly following the first roadway section (e.g., in each case using one of the methods described in this document).

[0037] The device is further configured to adapt the first reference profile and / or the second reference profile such that the first adapted reference profile and the second adapted reference profile have a uniform orientation at the transition.

[0038] The device can be configured, in particular, to determine a first orientation of the first reference profile at the transition and to determine a second orientation of the second reference profile at the transition. A target orientation can then be determined based on the first orientation and the second orientation. The target orientation can be determined particularly precisely based on an average, in particular an average value, of the first orientation and the second orientation.

[0039] Furthermore, the first reference profile and / or the second reference profile can be adapted such that the first adapted reference profile and the second adapted reference profile each have the target orientation at the transition.

[0040] The device can be configured to provide the first adapted reference profile as map data relating to the first road section and the second adapted reference profile as map data relating to the second road section for a digital map. Alternatively or additionally, the device can be configured to use the first adapted reference profile and / or the second adapted reference profile for automated longitudinal and / or transverse guidance of a vehicle (in the respective road section). Alternatively or additionally, the device can be configured to detect and / or determine one or more road markings in the first road section based on the first adapted reference profile and / or in the second road section based on the second adapted reference profile (e.g. using one of the methods described in this document).Alternatively or additionally, the device can be configured to detect one or more lane markings and / or one or more lanes in the first roadway section on the basis of the first adapted reference profile and / or in the second roadway section on the basis of the second adapted reference profile.

[0041] By adjusting the orientation of the reference courses at transitions between road sections, the quality of the map data obtained can be further increased.

[0042] The device can be configured to determine the first orientation of the first reference profile at an end point of the first reference profile and to determine the second orientation of the second reference profile at a start point of the second reference profile. A transition point can then be determined based on the end point and the start point, in particular as the center point of the end point and the start point (or as the center point determined from the end point and the start point). Furthermore, the first reference profile and / or the second reference profile can be adapted such that the first adapted reference profile at the end point and the second adapted reference profile at the start point are each arranged perpendicular to a plane passing through the transition point, wherein the normal vector of the plane has the target orientation. In this way, the quality of the adaptation of the reference profiles can be further increased.

[0043] It should be noted that the terms "end point" and "start point" are preferably independent of the direction of travel of the respective reference path. Rather, the end point of the first reference path is the final point of the first reference path that faces the initial point (i.e., the final point) of the second reference path.

[0044] The device can be configured to adapt the first reference profile and / or the second reference profile such that the first adapted reference profile and the second adapted reference profile lie at the transition in a uniform plane having the uniform orientation (in particular the target orientation).

[0045] The device can, in particular, be configured to adapt (e.g., shift) the end point of the first reference profile and the starting point of the second reference profile such that the end point of the first adapted reference profile and the starting point of the second adapted reference profile are arranged in the plane passing through the transition point. This allows the quality of the adaptation of the reference profiles to be further increased.

[0046] The device can be configured to determine a plurality of first reference profiles for the first roadway section and / or a plurality of second reference profiles for the second roadway section. The plurality of first reference profiles can be associated with different section sequences, wherein the different section sequences each comprise the first roadway section and at least one surrounding roadway section arranged directly in front of and / or directly behind the roadway section in the direction of travel. The different section sequences can differ from one another in at least one surrounding roadway section.

[0047] The device can be configured to adapt the plurality of first reference profiles and / or the plurality of second reference profiles such that all adapted reference profiles have a uniform orientation at the transition. Alternatively or additionally, the plurality of first reference profiles and / or the plurality of second reference profiles can be adapted such that all adapted reference profiles lie at a common transition point in a uniform plane (which has the target orientation). This can further increase the quality of the adaptation of the reference profiles.

[0048] The device can further be configured to adapt the first reference profile and / or the second reference profile such that the first adapted reference profile and the second adapted reference profile merge smoothly, and in particular continuously, into one another at the transition point. This allows the quality of the adaptation of the reference profiles to be further increased.

[0049] Alternatively or additionally, the device can be configured to adapt the first reference curve and / or the second reference curve such that the first adapted reference curve and the second adapted reference curve have a matching second derivative and / or a matching curvature at the (common) transition point. This can further increase the quality of the adaptation of the reference curves.

[0050] According to a further aspect, a device for determining a lane boundary on a road section is described. Based on one or more determined lane boundaries, one or more lanes can be determined. A lane is typically defined by two lane boundaries. Furthermore, groups of one or more lanes can be formed (so-called lane groups), which run parallel to one another, for example. Furthermore, the connectivity between multiple lanes can be determined. The connectivity can indicate whether or not a change between lanes is possible.

[0051] The device is designed to determine a reference course for the road section (e.g. using one of the methods described in this document).

[0052] The device is further configured to detect a road marking in the road section based on sensor data (in particular based on image data). The sensor data may have been acquired using one or more sensors (e.g., one or more cameras, one or more lidar sensors, one or more radar sensors, etc.). The image data may have been acquired using at least one camera of a vehicle and / or a camera of an aircraft. A (possible) road marking can be recognized from the image data using an image recognition algorithm. In general, a (possible) road marking can be recognized from the sensor data using a recognition algorithm. The road marking may extend along the direction of travel of the road section. The road marking may be represented by a sequence of pixels along the direction of travel of the road section.The reference course (with a sequence of points) and the road marking (with a sequence of image points) can be represented in a common coordinate system and compared with each other.

[0053] The device is configured to determine arrangement information relating to the arrangement of the road marking relative to the reference profile. As arrangement information, it can be determined, in particular, whether the road marking runs (essentially) parallel to the reference profile or not. Alternatively or additionally, the distance between the road marking and the reference profile can be determined as arrangement information. In this context, the described parallelism criterion can be used to determine whether the road marking and the reference profile run (essentially) parallel or not.

[0054] The arrangement information can indicate how the lane marking is positioned and / or aligned relative to the reference path. The arrangement information can, in particular, indicate the orientation of the lane marking relative to the reference path.

[0055] The device is further configured to determine, based on the arrangement information, whether the lane marking is a lane boundary (relevant for the relevance course) or not. In particular, it can be determined that the lane marking is a lane boundary (relevant for the relevance course) if it has been determined that the lane marking runs (essentially) parallel to the reference course. On the other hand, it can be determined that the lane marking is not a lane boundary (relevant for the relevance course) if it has been determined that the lane marking does not run (essentially) parallel to the reference course. Alternatively or additionally, it can be determined based on the distance between the lane marking and the reference course whether the lane marking is a lane boundary (relevant for the relevance course) or not.

[0056] The reference course for a road section (e.g. determined on the basis of measured travel paths) can thus be used to determine in a robust and precise manner whether a road marking (detected by sensors, in particular optically) corresponds to a lane boundary or not.

[0057] The detected lane markings can be provided as map data for the road section for a digital map. Alternatively or additionally, the detected lane markings can be used for automated longitudinal and / or lateral guidance of a vehicle (in the road section).

[0058] As already explained above, the device can be configured to divide the reference course into a sequence of support point planes at a corresponding sequence of points along the reference course. The individual support point planes can each run perpendicular to the reference course at the respective point. A sequence of intersection points of the road marking with the corresponding sequence of support point planes can then be determined. In particular, the intersection point with the road marking can be determined for each of the individual support point planes, resulting in a corresponding sequence of intersection points for the sequence of support point planes. Furthermore, the arrangement information can be determined in a particularly precise manner based on the sequence of intersection points.

[0059] In particular, arrangement information can be determined for each intersection point (e.g., indicating the distance between the lane marking and the reference path at the respective intersection point and / or whether the lane marking and the reference path at the respective intersection point are essentially parallel to each other). It can then be determined for each intersection point whether the lane marking at this intersection point is a lane marking (relevant to this reference path) or not. This allows for a sequential check to determine whether a lane marking (detected by sensors) corresponds to a lane marking or not.

[0060] The device can be configured to interpolate the lane markings detected based on the lane markings based on the sequence of intersection points and / or to extrapolate them, particularly beyond the lane section. This can further improve the quality of the map data obtained.

[0061] The device can be configured to determine the average distance of the intersection points of the sequence of intersection points from the corresponding points of the reference course. The distance of the individual intersection points to the corresponding points is determined within the respective reference point plane. The device can also be configured to compare the average distance with a width value for the (typical) width of a lane and, based on the comparison, to determine whether the road marking is a lane marking or not. This allows a lane marking to be detected in a particularly robust manner.

[0062] The device can be configured to determine an intersection point with a plurality of (sensor-detected) road markings for each of the individual support point levels of the sequence of support point levels of the reference course in order to determine a plurality of sequences of intersection points for the corresponding plurality of road markings. Thus, the intersection points of the individual support point levels with several (possibly parallel) road markings can be determined.

[0063] Based on the multitude of sequences of intersection points, it can be robustly determined whether the multitude of lane markings corresponds to a corresponding multitude of lane boundaries for a (contiguous) group of multiple, parallel lanes. This allows lane groups to be detected efficiently and robustly.

[0064] The device can be configured to determine a plurality of measured travel paths of one or more vehicles for a corresponding plurality of passages through the road section, and to determine a reference travel path for the road section on the basis of the plurality of measured travel paths (e.g. based on the data stored in connection with Fig. 3). It can then be determined in a particularly robust manner on the basis of the reference travel path, in particular on the basis of the arrangement of the reference travel path relative to the lane marking, whether the lane marking is a lane boundary (relevant for the relevance course) or not.

[0065] The device can be configured to detect a first lane marking and a second lane marking in the lane section based on sensor data (in particular based on image data), and to determine arrangement information regarding an arrangement of the first and second lane marking relative to the reference profile and / or relative to each other. It can then be determined based on the arrangement information whether • the first and second carriageway markings are lane boundaries; and • the first and second lane markings are part of a group of several parallel lanes.

[0066] This allows multiple lane markings to be analyzed to robustly identify groups of lanes (so-called lane groups).

[0067] According to a further aspect, a device for detecting a lane on a road section is described. The device can be designed, in particular, to determine the width of the lane. Alternatively or additionally, the device can be designed to determine one or both lane boundaries (i.e., the course and / or position of one or both lane boundaries) of the lane.

[0068] The device is configured to determine a plurality of measured travel paths of one or more vehicles for a corresponding plurality of passages through the roadway section. For each passage, a measured travel path can be sent from the respective vehicle to the device via a (wireless) communication link and received by the device. For example, 5 or more, 10 or more, or 20 or more measured travel paths can be determined.

[0069] A measured travel path of a vehicle can comprise a sequence of measurement points of the vehicle's position, in particular the position of a specific reference point (e.g. the center of an axle) of the vehicle, as it travels through the road section. The measurement points can be provided with a specific spatial resolution, such as 1 measurement point per meter or more. Alternatively or additionally, a measured travel path of a vehicle can indicate the actual travel trajectory of the vehicle as it travels through the road section. Measured travel paths can thus be determined which indicate the trajectory along which the individual vehicles (manually) traveled through the road section. The measured travel paths can therefore be used to describe the actual driving behavior of vehicles in the road section.

[0070] Furthermore, the device is configured to determine a vertical dispersion metric (e.g., the standard deviation) for the dispersion of the plurality of measured travel paths in the vertical direction, vertical to the surface of the roadway section. The lane on the roadway section can then be determined in a particularly precise manner based on the plurality of measured travel paths and taking into account the vertical dispersion metric. The device can, in particular, be configured to determine the width of the lane, transverse to the direction of travel of the roadway section, based on the vertical dispersion metric, in particular such that • the width of the lane decreases with increasing dispersion of the plurality of measured travel paths in the vertical direction; and / or • the width of the lane increases with decreasing scatter of the multitude of measured paths in the vertical direction.

[0071] The determined lane (with the determined width) can be provided as map data for the road section for a digital map. Alternatively or additionally, the determined lane (with the determined width) can be used for automated longitudinal and / or lateral guidance of a vehicle.

[0072] Thus, the vertical scatter of the measured paths can be taken into account when determining a reference course and / or a lane and / or a lane boundary in order to increase the quality of the determined map data.

[0073] The device can be configured to determine a lateral dispersion metric (e.g., the standard deviation) for the dispersion of the plurality of measured travel paths in the lateral direction, horizontal to the surface of the roadway section and transverse to the direction of travel of the roadway section. The lane, in particular the width of the lane, on the roadway section can then be determined particularly precisely, also taking into account the lateral dispersion metric. The device can, in particular, be configured to determine the width of the lane, transverse to the direction of travel of the roadway section, based on the vertical dispersion metric and on the basis of the lateral dispersion metric, in particular based on a quotient of the vertical dispersion metric and the lateral dispersion metric. The quotient can have the vertical control metric in the numerator and the lateral dispersion metric in the denominator.In this case in particular, the width of the lane can be determined in such a way that. • the width of the lane decreases with increasing quotient; and / or • the width of the lane increases with decreasing quotient.

[0074] Alternatively, the quotient can have the vertical control metric in the denominator and the lateral dispersion metric in the numerator. In particular, in this case, the lane width can be determined such that • the width of the lane increases with increasing quotient; and / or • the width of the lane decreases with decreasing quotient.

[0075] As already explained above, the device can be configured to divide a reference course of the roadway section into a sequence of support point planes for a corresponding sequence of points along the reference course. The sequence of support point planes can be determined such that, along the reference course of the roadway section, directly successive support point planes each have a predefined distance from one another, approximately between 1 and 3 meters, and / or that the individual support point planes are each arranged perpendicular to the reference course of the roadway section (at the respective point).

[0076] Then, for each of the plurality of measured travel paths, a sequence of intersection points of the respective measured travel path with the corresponding sequence of support point levels can be determined, so that for the individual support point levels, a plurality of intersection points for the corresponding plurality of measured travel paths results (as, for example, in connection with the method from Fig. 3 described).

[0077] For each of the individual support point levels, a vertical dispersion metric can be determined for the dispersion of the plurality of intersection points in the vertical direction. Furthermore, for each of the sequences of support point levels, a support point can be determined based on the plurality of intersection points with the respective support point level and taking into account the vertical dispersion metric for the respective support point level.

[0078] Based on the sequence of interpolation points for the corresponding sequence of interpolation points, a reference path for the lane on the road section can be determined. The reference path can precisely describe the centerline (in the direction of travel) of the lane.

[0079] The device can thus be configured (in particular using a method described in this document) to determine a reference travel path for the lane based on the plurality of measured travel paths. The reference travel path can be used, for example, as the center line of the lane. Furthermore, the width of the lane can be determined based on the plurality of measured travel paths (in particular based on the lateral scatter metric of the plurality of measured travel paths) and taking into account the vertical scatter metric (with 50% of the width being arranged on each side of the reference travel path). In this way, a lane and in particular the lane boundaries of the lane can be determined in a particularly precise manner.

[0080] According to a further aspect, a device for determining a lane boundary on a road section is described. The device can be designed, in particular, to detect a lane that is bounded by two lane boundaries.

[0081] The device is configured to determine a reference path for the road section (e.g., using one of the methods described in this document). The reference path of the road section can be divided (as described) into a sequence of support point planes for a corresponding sequence of points along the reference path.

[0082] The device is further configured to determine at least one measurement profile for the lane marking. The measurement profile may comprise at least one measured travel path of a vehicle for driving through the road section. Alternatively or additionally, the measurement profile may comprise at least one lane marking detected in the road section based on sensor data (in particular image data). Alternatively or additionally, the measurement profile may comprise at least one reference travel path for the road section (which was determined, for example, using one of the methods described in this document, in particular based on a plurality of measured travel paths).

[0083] The device is configured to determine a sequence of intersection points of the measurement profile with the corresponding sequence of support point planes, and to determine input data (in particular a feature vector) for a machine-learned determination unit based on the sequence of intersection points. The determination unit preferably has at least one artificial neural network. The determination unit (in particular the at least one neural network) can have been trained in advance using training data and a machine learning algorithm (e.g., a backpropagation algorithm). The training data can comprise a plurality of training data sets (e.g., 1,000 or more or 10,000 or more training data sets), each with training input data and training output data. The training data can (statistically) describe the target function of the trained determination unit.

[0084] The device is further configured to determine output data of the determination unit for the input data (transferred to the determination unit). The lane markings on the road section can then be determined precisely and robustly based on the output data.

[0085] The determined lane markings can be provided as map data for the road section for a digital map. Alternatively or additionally, the determined lane markings can be used for automated longitudinal and / or lateral guidance of a vehicle (in the road section).

[0086] The device is preferably configured to sequentially determine input data for each reference plane from the sequence of reference point planes, which data comprise the intersection point, in particular at least one coordinate of the intersection point, of the measurement curve with the respective reference plane. Output data can then be determined using the determination unit, which data comprise an intersection point, in particular at least one coordinate of the intersection point, of the lane boundary with the respective reference plane. The coordinates can be efficiently specified relative to the coordinate system of the respective reference plane.

[0087] This allows the intersection points of the lane markings with the individual support point levels to be determined step by step, level by level. These intersection points represent points of the lane markings along the reference course and / or along the direction of travel of the road section. This allows the lane markings to be determined with particular precision.

[0088] The device can be configured to determine input data that includes one or more intersection points of the measurement curve with one or more additional support point planes from a defined neighborhood of the respective reference plane. The defined neighborhood of the respective reference plane can include, • one or more support point planes located directly before the respective reference plane in the sequence of support point planes; and / or • one or more support point levels that are arranged directly after the respective reference level in the sequence of support point levels.

[0089] This allows additional intersection points from the vicinity of the reference plane to be considered as additional input data. This further improves the quality of the determined lane markings.

[0090] The device can be configured to determine input data comprising the intersection point of the lane boundary with the reference point plane that is located (in particular directly) before the respective reference plane in the sequence of reference planes. Thus, the output data for a previous iteration (i.e., for a previous reference plane) can be used as input data for the current reference plane. This allows the quality of the determined lane boundary to be further increased.

[0091] The device can thus be configured to sequentially select a reference point level from the sequence of reference point levels, starting with the first reference point level from the sequence of reference point levels, in order to sequentially determine a sequence of output data for the corresponding sequence of reference point levels. The lane markings on the road section can then be determined particularly precisely based on the sequence of output data.

[0092] The device can be configured to determine a plurality of measurement profiles for one or more lane markings on the roadway section, and for each of the plurality of measurement profiles, to determine a sequence of intersection points of the measurement profile with the corresponding sequence of reference point planes. The input data for the determination unit can then be determined based on the plurality of sequences of intersection points. By considering multiple measurement profiles, the quality of the one or more determined lane markings can be further increased.

[0093] The determination unit can be configured to receive a predefined (specified) maximum number of intersections per interpolation level as input data. The device can be configured to include one or more placeholder intersections for a specific interpolation level in the input data if the number of intersections with the specific interpolation level is smaller than the predefined maximum number of intersections. This allows for a particularly robust determination of one or more lane boundaries.

[0094] According to a further aspect, a method for merging reference profiles for a roadway section is described. The method comprises determining a first and a second reference profile for the roadway section (e.g., in each case using one of the methods described in this document). Furthermore, the method comprises identifying a first subsection of the roadway section in which the first and the second reference profile satisfy a predefined parallelism criterion. The method further comprises determining, for the first subsection, based on the first and the second reference profile, a fused reference profile that is used instead of the first and the second reference profile for the first subsection of the roadway section.

[0095] According to a further aspect, a method for adapting reference profiles at a transition between two consecutive roadway sections is described. The method comprises determining a first reference profile for a first roadway section and determining a second reference profile for a second roadway section directly following the first roadway section (e.g., using one of the methods described in this document). Furthermore, the method comprises adapting the first reference profile and / or the second reference profile such that the first adapted reference profile and the second adapted reference profile have a uniform orientation at the transition.

[0096] According to a further aspect, a method for determining a lane boundary on a road section is described. The method comprises determining a reference course for the road section (e.g., using one of the methods described in this document). Furthermore, the method comprises detecting, based on sensor data (in particular image data), a road marking in the road section. The method further comprises determining arrangement information regarding an arrangement of the road marking relative to the reference course, and determining, based on the arrangement information, whether the road marking is a lane boundary (relevant for the relevance course) or not.

[0097] According to a further aspect, a method for detecting a lane on a road section is described. The method comprises determining a plurality of measured travel paths of one or more vehicles for a corresponding plurality of passages through the road section, as well as determining a vertical dispersion metric for the dispersion of the plurality of measured travel paths in a vertical direction, vertical to a surface of the road section. Furthermore, the method comprises determining a lane on the road section based on the plurality of measured travel paths and taking the vertical dispersion metric into account.

[0098] According to a further aspect, a method for determining a lane boundary on a road section is described. The method comprises determining a reference course for the road section (e.g., using one of the methods described in this document). Furthermore, the method comprises dividing the reference course of the road section into a sequence of reference point planes (e.g., using one of the methods described in this document).

[0099] The method further comprises determining a measurement profile for the lane marking. The measurement profile can include a measured travel path of a vehicle for passing through the lane section and / or a lane marking detected in the lane section based on sensor data (in particular image data). Alternatively or additionally, the measurement profile can include at least one reference travel path for the lane section.

[0100] The method also includes determining a sequence of intersection points of the measurement profile with the corresponding sequence of support point planes, and determining, based on the sequence of intersection points, input data for a machine-learned determination unit. The method further includes determining output data of the determination unit for the input data and determining the lane markings on the road section based on the output data.

[0101] According to another aspect, a software (SW) program is described. The SW program can be configured to be executed on a processor and thereby perform one or more of the methods described in this document.

[0102] According to a further aspect, a storage medium is described. The storage medium may comprise a software program configured to be executed on a processor and thereby to perform one or more of the methods described in this document.

[0103] It should be noted that the methods, devices, and systems described in this document can be used both alone and in combination with other methods, devices, and systems described in this document. Furthermore, any aspects of the methods, devices, and systems described in this document can be combined in a variety of ways. In particular, the features of the claims can be combined in a variety of ways. Furthermore, features listed in parentheses are to be understood as optional features.

[0104] The invention will be described in more detail below using exemplary embodiments. Fig. 1a shows an exemplary system for determining reference travel paths for a road section; Fig. 1b shows an exemplary road section with several effective lanes; Fig. 2a exemplary measured travel paths for the Fig. 1b shown road section; Fig. 2b exemplary intersection points and support points within a plane of a road section; Fig. 2c exemplary support points for the sequence of levels of a road section; Fig. 2d exemplary reference paths for the Fig. 1b shown road section; Fig. 3 is a flowchart of an exemplary method for determining a reference travel path for a road section; Fig. 4a exemplary measured travel paths at a roadway intersection point; Fig. 4b shows an exemplary sequence of reference travel paths for successive road sections; Fig. 5a is a flowchart of an exemplary method for determining a reference travel path for a road section; Fig. 5b is a flowchart of an exemplary method for determining a target trajectory for a vehicle; Fig. 6 exemplary support points for reference travel paths in consecutive road sections; Fig. 7 shows an exemplary sequence of levels along a reference course of a road section; Fig. 8a is a flowchart of an exemplary method for determining a reference travel path for a road section; Fig. 8b is a flowchart of an exemplary method for determining a reference course of a road section; Fig. 9a exemplary reference courses in a road section; Fig. 9b shows an example merged reference curve; Fig. 9c is a flowchart of an exemplary method for merging reference histories; Fig. 10a exemplary reference courses in successive road sections; Fig. 10b is a flowchart of an exemplary method for adapting reference curves; Fig. 11a exemplary road markings; Fig. 11b is a flowchart of an exemplary method for detecting a lane boundary; Fig. 12 is a flowchart of an exemplary method for detecting a lane; Fig. 13a an exemplary machine learning-based detection unit; and Fig. 13b a flowchart of an exemplary method for determining a lane boundary.

[0105] As stated at the beginning, this document deals with the efficient and reliable determination of one or more reference paths and / or reference courses and / or lane markings for a roadway section of a road network. For example, different reference paths and / or lane markings can be determined for different lanes of the roadway section. The determined reference paths and / or lane markings can be provided as map data in an HD map for the road network. In particular, the one or more determined reference paths and / or lane markings can be used to determine target trajectories for the automated longitudinal and / or lateral guidance of vehicles.

[0106] Fig. 1a shows an exemplary system 100 for determining at least one reference travel path for at least one roadway section of a roadway network. The system 100 comprises a central unit 120 configured to determine measurement data 115 from a plurality of vehicles 110 and / or for a plurality of passages through a roadway section. The measurement data 115 for a passage and / or from a vehicle 110 can be received from the respective vehicle 110 via a (wireless) communication connection 101. The measurement data 115 from a passage of a vehicle 110 can indicate a measured travel path of the vehicle 110 through the roadway section (e.g., as a sequence of measurement points along the travel path traveled by the vehicle 110).

[0107] The central unit 120 (also generally referred to as a device in this document) can be configured to determine at least one reference travel path for this road section based on the measurement data 115 from a plurality of vehicles 110 and / or for a plurality of passages through the road section. The reference travel path can describe the typical travel path of vehicles 110 when passing through the road section. Different reference travel paths can be determined for different lanes of the road section. The one or more determined reference travel paths for the road section can be provided by the central unit 120 as map data 125 of a digital (HD) map for the road section (for example, sent to one or more vehicles 110 via a communication connection 101).

[0108] A vehicle 110 may include a position sensor 114 configured to acquire data related to the position of the vehicle 110 using a global navigation satellite system (GNSS). Furthermore, the vehicle 110 may include one or more sensors 112 (such as a speed sensor, an inertial measurement unit (IMU), one or more cameras, a wheel speed sensor, a steering sensor, etc.) configured to acquire sensor data that may be used for an odometry-based determination of the position of the vehicle 110.

[0109] A control unit 111 of the vehicle 110 can be configured to determine the position of the vehicle 110 when traveling through a road section based on the data from the position sensor 114 and / or on the sensor data from one or more sensors 112. In particular, a sequence of measurement points of the position of the vehicle 110 can be determined along the course of the road section, wherein the sequence of measurement points describes the travel path actually traveled by the vehicle 110. This travel path is referred to in this document as a "measured travel path." The sequence of measurement points of the measured travel path can be provided by the vehicle 110 as measurement data 115 (e.g., via a communication unit 113 of the vehicle 110). The sequence of measurement points can, for example, have a sampling rate or a spatial resolution of 1 measurement point per meter or higher.

[0110] Fig. 1b shows exemplary measured travel paths 160 for a roadway 150 (which can be divided into one or more roadway sections). The individual measured travel paths 160 each comprise a sequence of measurement points 161. The roadway 150 can have lane markings 152 that visually and / or sensorially indicate the edge of the roadway and / or a division of the roadway 150 into several lanes 151. The lane markings 152 can each represent an actual lane boundary.

[0111] The travel path 160 actually traveled by a vehicle 110 may approximately correspond to the centerline of a lane 151 of the roadway 150, particularly if the roadway 150 has a straight course. On the other hand, substantial deviations may occur, for example, before a curve, between the centerline of a lane 151 and the travel path 160 actually traveled by a vehicle 110. Furthermore, certain roadways 150 do not have lane markings 152 for subdividing them into lanes 151, such as before a toll system or at an intersection. In this case, no centerlines of lanes 151 can be determined as reference travel paths based on sensor-detected lane markings 152.

[0112] As explained above, the center line of a lane 151 can be used as a target trajectory for the automated longitudinal and / or lateral guidance of a vehicle 110. As in Fig. As shown in Figure 1b, in certain road sections, no lane markings 152 are present for determining lanes 151, and thus no target trajectory can be determined. Furthermore, in certain road sections (e.g., before or after a curve), the center line of a lane 151 may be unsuitable as a target trajectory for achieving a driving behavior typical of manually driven vehicles 110 during automated driving and correspondingly comfortable.

[0113] Fig. 2a shows the road section from Fig. 1b determined multitude of measured travel paths 160. It is assumed that the Fig. 1b corresponds to a (single) roadway section. The central unit 120 can be configured to determine a course 270 of the roadway section based on a digital (SD) map. This course 270 can be referred to as the map course of the roadway section. The map course 270 can be used as a reference for subdividing the roadway section into a sequence of (support point) levels 271. In particular, levels 271 can be arranged perpendicular to the map course 270 at a specific distance (e.g., between 1 and 3 meters, approximately 2 meters), which are used to subdivide the measured travel paths 160 in a corresponding manner into a sequence of intersection points.

[0114] The individual levels 271 intersect the different measured travel paths 160. Thus, a measured travel path 160 can be divided into a corresponding sequence of intersection points by the sequence of levels 271 along the map course 270 of the roadway section.

[0115] Fig. Figure 2b shows the plurality of intersection points 272 for a corresponding plurality of measured travel paths 160 within an exemplary level 271 of the sequence of levels 271 for a roadway section. The central unit 120 can be configured to convert the plurality of intersection points 272 of a level 271 into a set of reference points 200 for a corresponding set of reference travel paths for the roadway section using a clustering algorithm. The DBSCAN algorithm, for example, can be used as a clustering algorithm.

[0116] The central unit 120 can be configured to determine a plurality of intersection points 272 for the sequence of levels 271 of the roadway section and, based thereon, a set of support points 200, as shown by way of example in Fig. 2c. The number of support points 200 in the different levels 271 can differ at least partially from one another. The change in the number of support points 200 in different levels 271 can be due, for example, to the fact that travel paths merge along the road section or that a travel path splits into several travel paths along the road section (as can be the case, for example, when approaching a toll station).

[0117] As from Fig. 2c, a set of one or more reference points 200 can be determined for each of the individual levels 271 of the sequence of levels 271 on the basis of the measured travel paths 160. The central unit 120 can be configured to determine one or more reference travel paths 210 for the roadway section on the basis of the sequence of sets of support points 200 (as shown by way of example in Fig. 2d). For this purpose, along the sequence of levels 271 (in the direction of travel of the roadway section), starting from the first level 271, a support point 200 from the respective level 271 can be sequentially connected by a path segment 211 to a support point 200 from the directly following level 271. Thus, at least one sequence of support points 20 for the corresponding sequence of levels 271 can be determined and sequentially connected to path segments 211 in order to determine a reference travel path 210. In the Fig. In the example shown in Figure 2d, a total of 5 different reference travel paths 210 can be determined, wherein the reference travel paths 210 overlap in sub-segments of the roadway section.

[0118] The method described in this document processes discrete measurement points 161 along the measured trajectories 160 (i.e., the measured driving paths) of actual passages through a road section. This allows a reference driving path 210 to be determined for each lane 151, regardless of the existence of lane markings 152. This reference driving path 210, representing natural driver behavior, can replace the lane centerline as a target value for trajectory planning.

[0119] For each passage, a single estimated and / or measured sequence 160 of vehicle positions 161 (i.e., a measured travel path 160) is available. A clustering algorithm can be used to assign the intersection point 272 of a measured travel path 160 with a plane 271 perpendicular to the road course 270 to the corresponding intersection points 272 of the other measured travel paths 160.

[0120] The individual passages along a lane or road section typically differ due to natural driver behavior and a variety of environmental influences. To improve the reliability and accuracy of determining a reference travel path 210 along a lane, measures can be implemented to reduce the influence of interference. In particular, as a measure, when summarizing the intersection points 272 of the measured trajectories 160 with the planes 271 perpendicular to the road course 270 (if necessary per coordinate), an average value adjusted for outliers (in particular a so-called trimmed average) can be formed. The central unit 120 can thus be configured to identify outliers when determining the reference points 200 and to disregard them in order to increase the accuracy of the determined reference points 200.

[0121] Fig. 3 shows a flowchart of a (possibly computer-implemented) method 300 for determining a reference travel path 210 for a roadway section 410 (in a roadway network). A roadway section 410 is shown by way of example in Fig. 4a. The method 300 comprises determining 301 a plurality of measured travel paths 160 of one or more vehicles 110 for a corresponding plurality of passages through the roadway section 410. The measured travel paths 160 can each indicate the trajectory that the individual vehicles 110 actually traveled during the respective passage through the roadway section 410.

[0122] The method 300 further comprises subdividing 302 a history 270, 700 (e.g., the map history 270 or a history associated with Fig. 7) of the roadway section 410 into a sequence of support point planes 271. In other words, support point planes 271 (each perpendicular to the path 270, 700) can be placed (at regular intervals) along the path 270, 700. The support point planes 271 can each be two-dimensional. The path 270, 700 of the roadway section 270 and / or the measured travel paths 160 can be described in three-dimensional space.

[0123] Furthermore, the method 300 comprises determining 303, for each of the plurality of measured travel paths 160, a sequence of intersection points 272 of the respective measured travel path 160 with the corresponding sequence of support point planes 271. In other words, it can be determined where the individual measured travel paths 160 each separate the support point planes 271, and thus form intersection points 272.

[0124] The method 300 further comprises determining 304, for each of the sequence of support point planes 271, a set of support points 200 based on the determined intersection points 272 with the respective support point plane 271. For this purpose, one or more clusters of intersection points 272 can be determined in each individual support point plane 271 using a clustering algorithm. Furthermore, a support point 200 can be determined based on each cluster (e.g., as the (possibly trimmed) mean of the intersection points 272 in the cluster).

[0125] Furthermore, the method 300 includes determining 305, based on the sequence of sets of support points 200 for the corresponding sequence of support point levels 271, at least one reference travel path 210 for the roadway section 410. For this purpose, one or more connected sequences of support points 200 can be determined along the sequence of support point levels 271 and each connected to one another via path segments 211. The individual connected sequences of support points 200 then each form a reference travel path 210.

[0126] When a roadway 150 branches (e.g., at an exit or at an intersection) into two different adjacent roadways 150 with different orientations, the measured travel paths 160 of vehicles 110 in a roadway section may differ depending on whether the vehicle 110 drives into the first adjacent roadway 150 or the second adjacent roadway 150 following the roadway section. For example, at a turn, the measured travel paths 160 of vehicles 110 traveling straight ahead may (statistically) differ from the measured travel paths 160 of vehicles 110 turning. This is exemplary for a turning situation 400 in Fig. 4a. In particular, Fig. 4a a (core) roadway section 410 which is followed (when driving straight ahead) by a subsequent first (surrounding) roadway section 410 and which is followed (when turning right) by a subsequent second (surrounding) roadway section 410.

[0127] Out of Fig. 4a, it is particularly evident that the measured travel paths 160 in the (core) roadway section 410 before the turn differ depending on whether a vehicle 110 is driving straight ahead or turning after this (core) roadway section 410. The measured travel paths 160 in a (core) roadway section 410 can thus depend on which one or more subsequent (surrounding) roadway sections 410 the respective measured travel path 160 runs in.

[0128] In a corresponding manner, the measured travel paths 160 for a (core) roadway section 410 may depend on one or more preceding (surrounding) roadway sections 410 through which the respective measured travel path 160 has passed.

[0129] The central unit 120 can be configured to determine, for a (core) roadway section 410, a plurality of different reference travel paths 210 for different preceding and / or subsequent (surrounding) roadway sections 410. In particular, for a specific core roadway section 410 (e.g., based on the data of an (SD) card), all possible section sequences of directly consecutive roadway sections 410 can be determined, which have a specific number N of surrounding roadway sections 410 before the specific core roadway section 410 and / or a specific number M of surrounding roadway sections 410 after the specific core roadway section 410. N and / or M can each be 1 or more, or 2 or more. For each possible section sequence of roadway sections 410 (each comprising the specific core roadway section 410), a reference travel path 210 in the specific core roadway section 410 can then be determined.If different possible section sequences of roadway sections 410 exist for the specific core roadway section 410 Q, Q different reference travel paths 210 can thus be determined (e.g. Q greater than 1, or Q greater than 2).

[0130] For this purpose, the central unit 120 can be configured to assign the plurality of available measured travel paths 160 for the specific core roadway section 410 to one of the Q sequences of roadway sections 410. In the Fig. In the example shown in Figure 4a, the measured travel paths 160 in the straight-ahead direction would thus be assigned to a first section sequence of roadway sections 410 and the measured travel paths 160 in the right-turn direction would be assigned to a second section sequence of roadway sections 410.

[0131] Thus, for each section sequence of roadway sections 410, a subset of measured travel paths 160 can be determined. Based on the respective subset of measured travel paths 160, a reference travel path 210 can then be determined (using the method 300 described in this document). The Q reference travel paths 210 can then be provided as map data 125 for the specific roadway section 410. During operation of an (automated driving) vehicle 110, the section sequence of roadway sections 410 of the travel route of the vehicle 110 can then be determined. The appropriate reference travel path 210 can then be selected for traversing the specific core roadway section 410 (and used to determine the target trajectory of the vehicle 110 for traversing the core roadway section 410).

[0132] For the linking of reference travel paths 210 of successive roadway sections 410, it may be advantageous (as exemplified in Fig. 4b), a distance 420 can be provided between the end point of the reference travel path 210 of a roadway section 410 and the end or end point 412 of the roadway section 410. In a corresponding manner, a distance 420 can also be provided between the start or starting point 411 of the roadway section 410 and the start or starting point of the reference travel path 210 of the roadway section 410. The distance 420 can, for example, correspond to half the distance between two consecutive levels 271. The start point of a reference travel path 210 can correspond to the support point 200 of the first support point level 271 and / or the end point of a reference travel path 210 can correspond to the support point 200 of the last support point level 271 (in each case with respect to the direction of travel through the roadway section 410).

[0133] As in Fig. As shown in Figure 4b, the last reference point 200 of the reference travel path 210 of a roadway section 410 has a specific distance 420 from the end 412 of the roadway section 410. Furthermore, the first reference point 200 of the reference travel path 210 of the adjoining roadway section 410 has a specific distance 420 from the beginning 411 of the adjoining roadway section 410. A (possibly linear) connecting segment 421 can then be inserted between these two reference points 200 of the reference travel paths 210 in order to determine a linked reference travel path for the sequence of the roadway sections 410.

[0134] Fig. 5a shows a flowchart of a (possibly computer-implemented) method 500 for determining a reference travel path 210 for a core roadway section 410 (i.e., for a roadway section 410 that is referred to as a core roadway section for the purpose of clearly describing the method 500). The method 500 comprises determining 501 a plurality of measured travel paths 160 of one or more vehicles 110 for a corresponding plurality of passages through the core roadway section 410.

[0135] Furthermore, the method 500 comprises assigning 502 each of the plurality of measured travel paths 160 to a respective section sequence from a set of different section sequences in order to determine a respective subset of measured travel paths 160 for each section sequence from the set of section sequences. The different section sequences can each comprise the core roadway section 410 and at least one surrounding roadway section 410 arranged directly in front of and / or directly behind the core roadway section 410 in the direction of travel. The section sequences can thus each comprise one or more roadway sections 410.

[0136] In addition, the method 500 comprises determining 503, for each section sequence from the set of section sequences, a reference travel path 210 for the core roadway section 410 on the basis of the respective subset of measured travel paths 160. For this purpose, for example, the method 300 can be used.

[0137] Fig. Figure 5b shows a flowchart of a (possibly computer-implemented) method 510 for at least partially automated longitudinal and / or lateral guidance of a vehicle 110 along a route through a road network. The method 510 can be executed by a control unit 111 of the vehicle 110.

[0138] The method 510 comprises determining 511, in particular based on map data 125 of a digital map of the road network (such as an HD map), a set of different reference travel paths 210 for a corresponding set of different section sequences for an upcoming passage through a core roadway section 410 of the road network on the travel route. Furthermore, the method 510 comprises identifying 512, based on the travel route, a section sequence from the set of different section sequences. In particular, the section sequence that corresponds to the travel route, i.e., along which the travel route leads, can be determined.

[0139] Furthermore, the method 510 comprises determining 513 a desired trajectory of the vehicle 110 for an at least partially automated longitudinal and / or transverse guidance of the vehicle 110 when driving through the core roadway section 410 on the basis of the reference travel path 210 for the identified section sequence.

[0140] The determination of reference travel paths 210 for a sequence of roadway sections 410 that have a relatively large curvature (for example, due to a turning situation) can lead to significant discontinuities at the transitions between the reference travel paths 210 for directly consecutive roadway sections 410. This is exemplified in Fig. 6. In particular, Fig. 6 shows the support points 200 of the reference travel path 210 for a first roadway section 410 (each shown as white circles) and the support points 200 of the reference travel path 210 for a subsequent second roadway section 410 (each shown as hatched circles). The course 270 of the second roadway section 410 is essentially perpendicular to the course 270 of the first roadway section 410. This results in the support point planes 271 for the second roadway section 410 being arranged essentially perpendicular to the support point planes 271 for the first roadway section 410.

[0141] As from Fig. As can be seen in Figure 6, due to the relatively strong divergence between the orientation of the reference point planes 271 and the course of the measured travel paths 160, overlaps and / or discontinuities may occur between one or more reference points 200 at the end 412 of the first roadway section 410 and one or more reference points 200 at the beginning 411 of the subsequent second roadway section 410. This, in turn, may lead to discontinuities at the transition between the reference travel paths 210 of the two roadway sections 410.

[0142] As exemplified in Fig. As shown in Figure 7, the central unit 120 can be configured to determine a reference profile 700 of the roadway section 410 based on the measured travel paths 160 for a roadway section 410. For example, the measured travel paths 160 can be averaged (possibly without taking one or more outliers into account) to determine the reference profile 700. The reference profile 700 can then be used instead of the map profile 270 of the roadway section 410 to determine the sequence of support point levels 200 for the roadway section 410. The individual support point levels 200 can each be arranged perpendicular to the reference profile 700.

[0143] This in connection with Fig. 3 and the Fig. The method 300 described in Figures 2a to 2d can then be used accordingly to determine the support points 200 for a reference travel path 210. Thus, the accuracy of the determined reference travel path 210 can be increased, particularly with respect to the transition to the reference travel path 210 for the subsequent roadway section 410.

[0144] As from Fig. 7, the beginning 411 (i.e., the starting point or the position of the starting point) of the map course 270 of the roadway section 410, for which a travel path 210 is to be determined, can differ from the beginning 701 (i.e., the starting point or the position of the starting point) of the corresponding reference course 700. Alternatively or additionally, the end 412 (i.e., the end point or the position of the end point) of the map course 270 of the roadway section 410 can differ from the end 702 (i.e., the end point or the position of the end point) of the corresponding reference course 700. The central unit 120 can be configured to determine the beginning 701 and / or the end 702 of the reference course 700 such that a specific distance criterion or distance measure is reduced, in particular minimized.The distance criterion may depend on the distance of the respective point 701, 702 of the reference course 700 from the corresponding point 411, 412 of the map course 270. Alternatively or additionally, the distance criterion may depend on the deviation of the orientation 720 of the beginning 701 or the end 702 of the reference course 700 from the orientation 720 of the beginning 411 or the end 412 of the map course 270. The beginning 701 and / or the end 702 of the reference course 700 can thus be determined such that the distance to the beginning 411 and / or the end 412 of the map course 270 is as small as possible and / or that the orientation 720 of the beginning 701 and / or the end 702 of the reference course 700 is as similar as possible to the orientation 720 of the beginning 411 and / or the end 412 of the map course 270.

[0145] Fig. 8a shows a flowchart of a (possibly computer-implemented) method 800 for determining a reference travel path 210 for a roadway section 410. The method 800 comprises determining 801 a plurality of measured travel paths 160 of one or more vehicles 110 for a corresponding plurality of passages through the roadway section 410. Furthermore, the method 800 comprises determining 802, based on the plurality of measured travel paths 160, a reference course 700 of the roadway section 410. This reference course 700 can then be used as course 270, 700 in the method 300 to determine at least one reference travel path 210 for the roadway section 410.

[0146] The method 800 can thus in particular comprise arranging 803 a sequence of support point planes 271 along the reference course 700, and determining 804, for each of the plurality of measured travel paths 160, a sequence of intersection points 272 of the respective measured travel path 160 with the corresponding sequence of support point planes 271. Furthermore, the method 800 can comprise determining 805, based on the plurality of determined sequences of intersection points 272 for the corresponding plurality of measured travel paths 160, at least one reference travel path 210 for the roadway section 410.

[0147] Fig. 8b shows a flowchart of a (possibly computer-implemented) method 810 for determining a reference course 700 of a core roadway section 410 (i.e., a roadway section referred to as the core roadway section). The method 810 includes determining 811 a plurality of measured travel paths 160 of one or more vehicles 110 for a corresponding plurality of passages through the core roadway section 410.

[0148] Furthermore, the method 810 includes assigning 812 a first subset of measured travel paths 160 from the plurality of measured travel paths 160 to a first section sequence from a set of different section sequences. The different section sequences can each comprise the core roadway section 410 and at least one surrounding roadway section 410 arranged directly in front of and / or directly behind the core roadway section 410 in the direction of travel.

[0149] The method 810 further includes determining 813, based (solely) on the first subset of measured travel paths 160 for the first section sequence, a first reference profile 700 of the core roadway section 410 for the first section sequence. As in this document, the first reference profile 700 can be determined (solely) on the first subset of measured travel paths 160.

[0150] Thus, when determining the reference path 700 for a core roadway section 700, it is already possible to consider the surrounding roadway section 700 from which a vehicle 100 enters the core roadway section 700 and / or the surrounding roadway section 700 into which a vehicle exits from the core roadway section 700. Thus, a direction-dependent reference path 700 can be determined. This allows the quality of the determined reference travel paths 210 to be further increased.

[0151] In particular, the different reference travel paths 210 for the different section sequences can each be determined using the specific reference profile 700 for the respective section sequence. This allows the quality of the determined reference travel paths 210 to be significantly increased.

[0152] The reference course 700 of the core roadway section 410 for a specific section sequence can be provided as map data for a digital map. The reference course 700 can then be displayed, for example, in a head-up display of a vehicle 110 when driving through the core roadway section 410 (e.g., as an augmented reality display).

[0153] It may happen that a plurality of different reference courses 700 are determined for a (core) roadway section 410, e.g., different section sequences for Q. This is exemplified in Fig. 9a. In particular, Fig. 9a a first reference course 901 (e.g. for driving straight ahead) and a second reference course 902 (e.g. for a turning maneuver) for the (core) roadway section 410 (each with a beginning 701 and an end 702).

[0154] The use of a plurality of different reference curves 901, 902 may be associated with a relatively high resource expenditure (e.g., for storage and / or computing resources) in the context of further data processing. The device 120 may be configured as shown, for example, in Fig. 9b, to identify at least a partial section 904 of the roadway section 410 in which the reference courses 901, 902 are oriented substantially the same (and as a result provide substantially identical support point planes 271). In the Fig. In the example shown in Figure 9b, this is the case for the section 904 between the beginning 701 and an intermediate point 903.

[0155] A fused reference profile 911 can be determined for the identified subsection 904, which results, for example, from averaging the reference profiles 901, 902 in the identified subsection 904. The originally determined reference profiles 901, 902 can then adjoin at the end 903 of the subsection 904. A local offset between the fused reference profile 911 and the adjoining reference profiles 901, 902 is typically irrelevant (for determining the support point planes 271). On the other hand, when determining the fused reference profile 911, it can be taken into account and / or ensured that the fused reference profile 911 and the adjoining reference profiles 901, 902 each have the same orientation at the transition point 903.

[0156] The separate treatment of the individual driving and turning directions when generating the road geometry (i.e., the reference courses 700) can thus lead to situations in which road courses of different driving directions run almost parallel before junctions and after junctions being stored separately in the digital map 125. The multiple representation of the almost identical road geometry (i.e., the almost identical reference course 700) increases the data volume required for storing and transmitting the map data 125. Furthermore, the localization of a vehicle 110 on a digital map of road courses or associated road geometries 700 can be made more difficult due to the relatively high number of possible assignments. Path planning based on the map (i.e.,Determination of a target trajectory) for a vehicle 110 typically has to evaluate a relatively large number of possible paths when determining the optimal path, thereby increasing the required amount of computing resources.

[0157] When using the road geometry 700 to generate cross-sectional planes 271 for determining localization-relevant landmarks, the multiple representation of the same road geometry 700 can lead to landmarks being determined multiple times and subsequently having to be sorted out.

[0158] After generating the road course geometries (i.e., the reference courses) 901, 902, the road course geometries 901, 902 can be divided into groups of potentially fused road course geometries 901, 902. The criteria for fusedness can include, in particular, belonging to the same map tile (an SD map), belonging to one of its neighbors, spatial proximity in general, and / or belonging to the same SD road section 410 or the same sequence of road sections.

[0159] Within a group, each pair of road geometries 901, 902 can be iteratively checked to determine whether they run approximately parallel to each other in one or more subsections 904 and can therefore be merged. If this is the case, the original road geometries 901, 902 are replaced by the fusion result 911. The (iterative) process can then be continued with the fused reference geometries 911 and the remaining road geometries. The fusion result 911 may include the joint, fused representation of the road geometries of the subsection 904 on which the two original road geometries 901, 902 run parallel to each other, and the (now duplication-free) subsections before and / or after.

[0160] The method can be continued until no more pairs of fused road geometry 901, 902 can be found. The testing and fusion of a first and second road geometry 901, 902 is carried out, for example, by first determining the lengths along the first road geometry 901 from which and up to which the second road geometry 902 runs approximately parallel to the first road geometry 901, i.e., while adhering to a parallelism criterion. If this is the case for a non-empty interval 904, two or more cross-sectional planes 271 (in particular at regular intervals) can be constructed orthogonal to the first road geometry 901 within this interval 904.For each cross-sectional plane 271, a support point 200 of the merged road course 911 can be determined from the intersection points 272 of the first and second road course geometries 901, 902 with the respective cross-sectional plane 271.

[0161] The determination of the support point 200 can advantageously be carried out by a weighted averaging of the intersection points 272 with the respective cross-sectional plane 271. The number of road course geometries 901, 902 that have already been averaged to generate the first and second road course geometries 901, 902 can advantageously be used as the weighting factor.

[0162] Fig. 9c shows a flowchart of a (possibly computer-implemented) method 930 for merging reference profiles 901, 902 for a roadway section 410. The method 930 comprises determining 931 a first and a second reference profile 901, 902 for the roadway section 410. The two reference profiles 901, 902 can, for example, have been determined for different section sequences (e.g., using the method 810).

[0163] The method 930 further comprises identifying 932 a first subsection 904 of the roadway section 410 in which the first and second reference paths 901, 902 satisfy a predefined parallelism criterion.

[0164] Furthermore, the method 930 includes determining 933, for the first subsection 904, based on the first and second reference profiles 901, 902, a fused reference profile 911, which is used instead of the first and second reference profiles 901, 902 for the first subsection 904 of the roadway section 410. The fused reference profile 911 can be used, for example, to determine a lane boundary 152 and / or to determine a reference travel path 210 (e.g., within the scope of the method 300).

[0165] The determination of different reference profiles 700 for different, consecutive roadway sections 410 can lead to a discontinuity, in particular a discontinuity with respect to the orientation of the reference profiles 700, occurring at the transition between the reference profiles 700 for two directly consecutive roadway sections 410. This is exemplified in Fig. 10a. In particular, Fig. 10a shows a first reference profile 1001 for a first roadway section 410 and a second reference profile 1011 for a second roadway section 410 directly following the first roadway section. At the transition at the end 702 of the first reference profile 1001 and the beginning 701 of the second reference profile 1011, both reference profiles 1001, 1011 have different orientations 1002, 1012.

[0166] The device 120 can be configured to adapt the reference curves 1001, 1011 such that the two reference curves 1001, 1011 have the same orientation 1002, 1012 at the transition.

[0167] As already explained above, for the determination of one or more reference travel paths 210 and / or localization-relevant landmarks, it is typically not the road geometry 1001, 1011 itself that is relevant, but rather the position and orientation of the cross-sectional planes 271 constructed thereby. Therefore, all road geometries 1001, 1011 entering and / or exiting a transition point can be adjusted such that they have the same orientation 1002, 1012 at the transition point itself, without necessarily merging into one another. With a continuous representation of the road geometries 1001, 1011, e.g., a spline, the continuity of the first derivative, but not necessarily the zeroth derivative, can be ensured.

[0168] The adaptation of the road course geometries 1001, 1011 can be carried out by determining a common adjustment plane, the reference point of which advantageously corresponds to the center of gravity of the start and end points 701, 702 of the road course geometries 1001, 1011 involved and the normal vector of which is advantageously determined as the mean value of the associated orientations, which may be mirrored in the case of opposite directions.

[0169] Subsequently, the involved road geometries 1001, 1011 can be adjusted such that their start and end points 701, 702 lie in the determined plane, respectively, and their tangent vector at this point runs parallel to the normal vector of the determined plane. The geometries can be adjusted by adjusting their support points 200, for example, using spline interpolation.

[0170] Optionally, a mean curvature can be determined that all road geometries 1001, 1011 should have at the determined plane. Optionally, it can be ensured that all road geometries 1001, 1011 pass through the reference point of the determined plane, thus achieving continuity in the zeroth derivative, even if this results in the direction of travel of the determined road geometries 1001, 1011 differing locally more significantly from the actual travel paths 160.

[0171] Fig. 10b shows a flowchart of a (possibly computer-implemented) method 1020 for adapting reference profiles 1001, 1011 at a transition between two consecutive roadway sections 410. The method 1020 comprises determining 1021 a first reference profile 1001 for a first roadway section 410, and determining 1022 a second reference profile 1011 for a second roadway section 410 directly following the first roadway section 410. The reference profiles 1001, 1002 can each be determined using the method 810.

[0172] The method 1020 further includes adapting 1023 the first reference profile 1001 and / or the second reference profile 1011 such that the first adapted reference profile and the second adapted reference profile have a uniform orientation at the transition. The first adapted reference profile and / or the second adapted reference profile can be used, for example, to determine a lane boundary 152 and / or to determine a reference travel path 210 (e.g., within the scope of the method 300).

[0173] Based on the camera data from one or more cameras of vehicles 110 and / or on the basis of satellite data, ie generally on the basis of image data, possible lane markings can be detected, which, for example, separate different lanes 151 of a roadway 150 from one another. The reference curves 700 described in this document for individual road sections 410 can be used to detect in a particularly reliable manner whether a possible lane marking (detected on the basis of sensor data, in particular image data) is also an actual lane boundary 152.

[0174] For this purpose, as exemplified in Fig. As shown in Figure 11a, the reference path 700 for a roadway section 410 can be divided into a sequence of support point planes 271 (for a corresponding sequence of points along the reference path 700). Directly consecutive support point planes 271 can be spaced at a constant distance (e.g., 1 to 3 meters) from one another.

[0175] The intersection points 272 of the sequence of support point planes 271 with one or more possible lane markings 1101, 1102 can then be determined. This results in a corresponding sequence of intersection points 272 for each possible lane marking 1101, 1102.

[0176] The sequence of intersection points 272 for a possible lane marking 1101, 1102 can be analyzed to determine whether the possible lane marking 1101, 1102 is an actual lane marking 152 or not. In particular, it can be analyzed how the individual intersection points 272 are arranged relative to the reference profile 700. For example, the (perpendicular) distance 1103 of the individual intersection points 272 to the reference profile 700 can be considered. If this distance 1103 is essentially constant for the sequence of intersection points 272 and / or has a certain value (which is typical for the width of a lane 151), it can be reliably concluded that the possible lane marking 1101 is an actual lane marking 152.

[0177] On the other hand, if the distance 1103 of the intersection points 272 changes and / or the distance 1103 has a value that is inappropriate for a lane 151, it can be concluded that the possible lane marking 1102 is not an actual lane boundary 152.

[0178] Lanes 151 that can be used to travel locally from one point in the road network to another, e.g., all left-turn lanes 151 or all straight-ahead lanes 151 at an intersection, generally run parallel to one another. Since vehicles 110 primarily move along lanes 151, the road course geometry determined from the measurement data 115 (i.e., the reference course 700) also runs parallel to all lanes 151 of a contiguous group of lanes 151. In the coordinate system of a given road course geometry 700, the associated lanes 151 therefore run parallel to the coordinate axis in the direction of travel, provided the configuration of the lanes 151, e.g., the width of the lanes 151, remains unchanged.

[0179] This can be used to determine from one or more possible lane markings 1101, 1102 (e.g., recognized on the basis of sensor data, in particular image data) those that are relevant for the currently considered lane section 410 (e.g., for a specific turning direction at an intersection), in order to determine a common lane group for these actual lane markings 152, e.g., in the sense of the NDS (Navigation Data Standard) data format.

[0180] The NDS data format groups adjacent lanes 151 on a roadway section 410 as a lane group. In combination with the lane marking type 152 (e.g., solid line or dashed line), this results in a compact representation of the lateral connectivity of the lanes 151, i.e., the information on which roadway sections 410 allow a change from one lane 151 to another.

[0181] With the method described in this document, a subsequent determination of the lateral connectivity of adjacent lanes 151 is not necessary, since the lateral connectivity is already given by the sorting of the coordinates of the intersection points 272 of the lane markings 1101, 1102 in the cross-sectional plane 271 along the coordinate axis running transversely to the direction of travel.

[0182] This also means that the intersection points 272 of the road markings 1101 typically change only minimally between the levels 271 in the coordinate system of the cross-sectional levels 271 perpendicular to the road geometry 700, enabling traceability of individual markings 1101 from one level 271 to a subsequent level 271, thereby closing gaps in the available observations of the road markings 1101 (e.g., in the interior of the intersection). This enables a seamless determination of the lanes 151, even in the interior of the intersection.

[0183] Due to the fact that only lane markings 1101 that run parallel to the associated road geometry 700 are used for the individual lane sections 410, the described method is robust against incorrectly detected lane markings 1102 that run transversely or diagonally to the direction of travel. Furthermore, the joint consideration of several parallel lane markings 1101 enables additional plausibility checks, for example, with regard to the assumption that the lane width is the same for all lanes 151. This can further increase the robustness against incorrect lane markings 1102.

[0184] To further reduce the risk of errors in the determined road model, observations of actual passages 160 of a lane 151, for example in the form of one or more previously determined reference travel paths 210, can be used for plausibility checks. If necessary, only those lanes 151 for which such a reference travel path 210 or such an observation 160 exists, or for which certain other criteria apply, can be included in the road model. Such criteria can relate, for example, to whether an adjacent lane 151 has been confirmed by a corresponding observation and what type of lane marking 152 the current lane 151 has.

[0185] Fig. 11b shows a flowchart of a (possibly computer-implemented) method 1110 for determining a lane marking 152 on a roadway section 410. The method 1110 comprises determining 1111 a reference profile 700 for the roadway section 410 (e.g., using method 810). Furthermore, the method 1110 comprises detecting 1112, based on sensor data, in particular image data, a road marking 1101, 1102 in the roadway section 410.

[0186] The method 1110 further comprises determining 1113 arrangement information relating to the arrangement of the lane marking 1101, 1102 relative to the reference profile 700. In addition, the method 1110 comprises determining 1114, based on the arrangement information, whether the lane marking 1101, 1102 is a lane boundary 152 or not.

[0187] As in connection with Fig. As shown in Figure 1b, when a road section 410 is traversed a plurality of times, measurement data 115 can be recorded with respect to the travel path 160 of the respective traversal. In other words, a plurality of measured travel paths 160 can be recorded for the road section 410. Based on the measured travel paths 160, one or more reference travel paths 210 can be determined for the road section 410. Alternatively or additionally, at least one reference profile 700 of the road section 410 can be determined.

[0188] The detection quality of the individual measured travel paths 160 can vary. In particular, the position accuracy of the individual measuring points 161 of measured travel paths 160 can vary. The position accuracy of the position sensor 114 of a vehicle 110 is typically independent of direction. In particular, the position sensor 114 typically has the same position accuracy in the horizontal direction (i.e., parallel to the surface of the roadway 150) as in the vertical direction (i.e., perpendicular to the surface of the roadway 150). The control of the measuring points 161 of a measured travel path 160 in the vertical direction (i.e., the vertical scatter) can thus be used as an indicator of the position accuracy in the horizontal direction.

[0189] The vertical dispersion of the measurement points 161 of a measured travel path 160 can be used, in particular, to determine how strongly (e.g., with what weighting) the respective measured travel path 160 is taken into account when determining a reference travel path 210 and / or when determining a lane boundary 152. Typically, the weighting increases with decreasing vertical dispersion. On the other hand, the weighting typically decreases with increasing vertical dispersion.

[0190] A reference travel path 210 determined from measurement data 115 can contain information regarding the variance of the individual passages used to generate the reference travel path 210, in particular the scatter of the trajectories 160 transverse to the direction of travel. This scatter can advantageously be used to determine the width of a (possibly virtual) lane 151 and thus to determine the position of the determined lane markings 152.

[0191] Since the localization quality of the determined trajectories 160 can vary due to the availability and / or uniqueness of localization-relevant landmarks, it is not always possible to determine with certainty whether a high scatter of the trajectories 160 results from a reduced localization quality or from the actual driving behavior during the individual passes. Taking advantage of the observation that a reduced localization quality affects all spatial directions or that there is a correlation between the vertical localization quality and the lateral localization quality, the vertical scatter of the trajectories 160 can be used to decide whether a relatively high lateral scatter is due to a reduced localization quality or to the actual driving behavior during the individual passes.

[0192] For example, the quotient of the lateral and vertical dispersion metrics of the measured paths 160 belonging to a reference path 210 can be used to determine the width of the (possibly virtual) lane 151. A lane 151 may only be widened if a relatively low vertical dispersion indicates that the actual driving behavior is scattered, and not if a relatively low localization quality results in inaccurate trajectories 160.

[0193] In addition to the variance of the measured travel paths 160 during the individual passages, further information can be used to determine the position of the lane markings 152, such as the position and variance of the measured travel paths 160 for one or more adjacent reference travel paths 210 and / or the position of lane markings, which were detected, for example, on the basis of sensor data, in particular image data, before or after the considered road section 410 along the reference travel path 210.

[0194] In a further embodiment, other metrics of lateral dispersion can be used alternatively or additionally, such as the minimum and the maximum and / or a predetermined quantile of lateral dispersion.

[0195] Fig. 12 shows a flowchart of a (possibly computer-implemented) method 1200 for detecting and / or determining a lane 151 on a roadway section 410. In particular, a reference travel path 210 and / or a lane boundary 152 for the lane 151 can be detected and / or determined.

[0196] The method 1200 comprises determining 1201 a plurality of measured travel paths 160 of one or more vehicles 110 for a corresponding plurality of passages through the roadway section 410. Furthermore, the method 1200 comprises determining 1202 a vertical dispersion metric for the dispersion of the plurality of measured travel paths 160 in the vertical direction, vertical to the surface of the roadway section 410. The method 1200 further comprises determining 1203 a lane 151 (in particular the width of the lane 151) on the roadway section 410 based on the plurality of measured travel paths 160 and taking into account the vertical dispersion metric.

[0197] As already explained above, the reference profile 700 for a roadway section 410 can be used to determine lane markings 152 for one or more lanes 151 of the roadway section 410. The reference profile 700 can be used, in particular, to divide the roadway section 410 into a sequence of reference point planes 271, wherein the individual reference point planes 271 are each arranged perpendicular to the reference profile 700. The individual reference point planes 271 can then be used to determine intersection points 272 of the respective reference point plane 271 with (imaged) road markings 1101, 1102 and / or measured travel paths 160. These intersection points 272 (per support point level 271) can be analyzed to determine a reference travel path 210 and / or a lane boundary 152 for a lane 151 of the roadway section 410.

[0198] The determination of a reference travel path 210 and / or a lane boundary 152 can be effected in a particularly efficient and precise manner using a neural network that has been trained in advance. Fig. 13a shows an exemplary (artificial intelligence and / or machine learning-based) determination unit 1300 configured to determine output data 1302 (describing the reference travel path 210 and / or the lane boundary 152) based on input data 1301 (based on the intersection points 272 with one or more support point levels 271).

[0199] The detection unit 1300 may be trained in advance using training data (e.g., using a learning algorithm). The detection unit 1300 may comprise one or more neural networks that have been trained in advance.

[0200] The sequence of interpolation points 271 can, for example, have K consecutive interpolation points 271 (e.g., K≥10 or K≥20). A window with a subset of interpolation points 271 with a specific number of directly consecutive interpolation points 271 can be considered (e.g., one or more, or two or more, or three or more interpolation points 271). This window can be successively shifted from the beginning 701 of the reference curve 700 to the end 702 of the reference curve 700 in order to successively determine the reference travel path 210 and / or the lane boundary 152.

[0201] The input data 1301 can be determined based on the intersection points 272 for a window of one or more support point planes 271. Based on this, output data 1302 for this window of one or more support point planes 271 (or for a reference plane 271 from the window) can then be determined using the determination unit 1300. The window can then be repeatedly shifted with a specific increment to the end 702 of the reference curve 700, and input data 1301 and output data 1302 can be determined for the shifted window. The increment can, for example, correspond to a respective support point plane 271.

[0202] A method is thus described with which the road model of a roadway section 410 can be determined using machine learning methods, in particular deep neural networks. For this purpose, the available information can be transformed in such a way that the trained machine learning model (i.e., by the determination unit 1300), represented by the function f, can map a set X of possible feature vectors of fixed length (i.e., of possible input data 1301) to a set Y of output vectors of fixed length (i.e., of output data 1302), from which the road model to be determined (in particular, the course of a lane marking 152) can be determined.

[0203] Based on the road geometry 700, the information relevant to the road model can be represented in the cross-sectional planes 271 perpendicular to the road geometry 700. To determine the road model, features can be represented as attributes of the individual cross-sectional planes 271 and / or associated coordinates can be transformed into the coordinate systems of the individual cross-sectional planes 271.

[0204] The input data 1301 (in particular the input vector) can include intersection points 272 of one or more road markings 1101, 1102, which were detected, for example, based on sensor data, in particular based on image data, with a reference cross-sectional plane 271 (as well as with one or more directly adjacent cross-sectional planes 271). A fixed set and / or number of intersection points 272 can be used per plane 271.

[0205] The individual intersection points 272 can each be represented by their coordinates. Advantageously, the coordinates of the intersection points 272 contained in the input data 1301 are transformed into the coordinate system of the respective plane 271 and / or normalized to a predefined value range (e.g., minus 10 to plus 10 meters).

[0206] Typically, the actual number of intersection points 272 is not fixed, but variable. To nevertheless convert these into a feature vector of fixed length, the intersection points 272 can be grouped according to the associated cross-sectional plane 271, sorted per plane 271 according to their respective coordinates transverse to the direction of travel, or according to their respective coordinates in the coordinate system of the respective plane 271, and then supplemented by a number of placeholder intersection points, where the number of placeholder intersection points corresponds to the difference between the number of actually existing intersection points 272 and the specified constant maximum number of intersection points 272 per plane 271.

[0207] The machine learning model, i.e., the determination unit 1300, may have been trained in advance such that the output data 1302, in particular the output vector, contains a set of coordinates per cross-sectional plane 271, which represents the one or more intersection points 272 of the one or more lane boundaries 152 of a consistent road model with the reference cross-sectional plane 271. Incorrectly detected lane markings 1102 may have been eliminated by the determination unit 1300.

[0208] The output data 1302, in particular the output vector, can (analogously to the input data 1302) have placeholder intersections if the number of intersections 272 actually specified by the consistent road model does not correspond to the specified maximum number of intersections 272.

[0209] For training the machine learning model, i.e., the determination unit 1300, a plurality of training data sets can be used, with each training data set comprising training input data and matching (desired) training output data. The training output data can be determined through actual measurements of lane markings 152 (where the measurements were performed, for example, using a measuring vehicle). Thus, supervised learning of the determination unit 130 can have been performed in advance.

[0210] The trained machine learning model, i.e., the determination unit 1300, can be applied to feature vectors (i.e., input data 1301) for a sequence of consecutive reference cross-sectional planes 271. Thus, the intersection points 272 of the lane markings 152 of a consistent road model with the individual reference cross-sectional planes 271 can be determined. Any errors and / or gaps in the road markings 1101, 1102 used for the input data 1301 can be corrected by the machine learning model, so that the individual intersection or support points 272, 200 can be precisely combined to form a coherent road model.

[0211] Alternatively or in addition to the use of placeholder intersections, in the event that the defined maximum number of intersections is not exhausted, a grid with a fixed number of cells can be placed over the respective level 271, with either no or a maximum of one intersection 272 being represented per cell.

[0212] Alternatively or additionally, the type of lane markings 152 can be taken into account in the input data 1301 and / or in the output data 1302, for example in the form of probabilities for the different possible types of lane markings 152.

[0213] Alternatively or additionally, the intersection points 272 of measured travel paths 160 with the respective levels 271 can be taken into account in the input data 1301.

[0214] Fig.13b shows a flowchart of a (possibly computer-implemented) method 1310 for determining a lane boundary 152 on a roadway section 410. The method 1310 includes determining 1311 a reference course 700 for the roadway section 410 (e.g., using method 810). Furthermore, the method 1310 includes subdividing 1312 the reference course 700 of the roadway section 410 into a sequence of support point planes 271 for a corresponding sequence of points along the reference course 700.

[0215] Furthermore, the method 1310 comprises determining 1313 at least one measurement profile 160, 1101, 1102 for the lane marking 152. The measurement profile 160, 1101, 1102 can be, for example, a measured travel path 160 of a vehicle 110 for traveling through the roadway section 410. Alternatively or additionally, the measurement profile 160, 1101, 1102 can be a road marking 1101, 1102 detected in the roadway section 410 based on sensor data, in particular image data. Alternatively or additionally, the measurement profile 160, 1101, 1102 can be a reference travel path 210 for the roadway section 410. The method 1310 further comprises determining 1314 a sequence of intersection points 272 of the measurement profile 160, 1101, 1102 with the corresponding sequence of support point planes 271.

[0216] In addition, the method 1310 includes determining 1315, based on the sequence of intersection points 272, input data 1301 for a machine-learned determination unit 1300, determining 1316 output data 1302 of the determination unit 1300 for the input data 1301, and determining 1317 the lane marking 152 on the roadway section 410 based on the output data 1302.

[0217] The measures described in this document enable the efficient and precise determination of reference paths 700 and / or lanes 151 and / or lane markings 152 for roadway sections 410 of a roadway network, which can be provided as map data 125 for an HD map. The determined reference paths 700 and / or lanes 151 and / or lane markings 152 can be used by an at least partially autonomously driving vehicle 110 to determine target trajectories for the vehicle 110. This can improve the quality of autonomously driving vehicles 110.

[0218] The present invention is not limited to the embodiments shown. In particular, it should be noted that the description and figures are intended only to illustrate the principle of the proposed methods, devices, and systems by way of example.

Claims

[1] Device (120) for determining a lane boundary (152) on a roadway section (410); wherein the device (120) is configured - to determine a reference course (700) for the road section (410); - to detect a road marking (1101, 1102) in the road section (410) on the basis of sensor data; - to determine arrangement information relating to an arrangement of the road marking (1101, 1102) relative to the reference course (700); and - to determine on the basis of the arrangement information whether the road marking (1101, 1102) is a lane boundary (152) or not. [2] Device (120) according to claim 1, wherein the device (120) is arranged - to determine as arrangement information whether the road marking (1101, 1102) runs substantially parallel to the reference course (700) or not; and - to determine that the lane marking (1101, 1102) is a lane boundary (152) if it has been determined that the lane marking (1101, 1102) runs substantially parallel to the reference course (700); and / or - to determine that the lane marking (1101, 1102) is not a lane boundary (152) if it has been determined that the lane marking (1101, 1102) does not run substantially parallel to the reference course (700). [3] Device (120) according to claim 1, wherein the device (120) is arranged - to determine a distance (1103) between the lane marking (1101, 1102) and the reference course (700) as arrangement information; and - to determine, on the basis of the distance (1103) between the lane marking (1101, 1102) and the reference profile (700), whether the lane marking (1101, 1102) is a lane boundary (152) or not. [4] Device (120) according to one of the preceding claims, wherein the device (120) is arranged - to divide the reference profile (700) into a sequence of support point planes (271) at a corresponding sequence of points along the reference profile (700); wherein the individual support point planes (271) each run perpendicular to the reference profile (700) at the respective point; - to determine a respective intersection point (272) with the road marking (1101, 1102) for each of the individual support point levels (271) of the sequence of support point levels (271) in order to determine a sequence of intersection points (272) of the road marking (1101, 1102) for the corresponding sequence of support point levels (271); and - to determine the arrangement information based on the sequence of intersection points (272). [5] Device (120) according to claim 4, wherein the device (120) is configured to interpolate the lane boundary (152) detected on the basis of the lane marking (1101, 1102) on the basis of the sequence of intersection points (272) and / or to extrapolate it, in particular beyond the lane section (410). [6] Device (120) according to one of claims 4 to 5, wherein the device (120) is arranged - to determine an average distance of the intersection points (272) of the sequence of intersection points (272) from the corresponding points of the reference profile (700) within the respective support point plane (271); - to compare the mean distance with a width value for the width of a lane (151); and - to determine on the basis of the comparison whether the road marking (1101, 1102) is a lane boundary (152) or not. [7] Device (120) according to one of claims 4 to 6, wherein the device (120) is arranged - for each of the individual support point levels (271) of the sequence of support point levels (271), to determine an intersection point (272) with a plurality of road markings (1101, 1102) in order to determine a plurality of sequences of intersection points (272) for the corresponding plurality of road markings (1101, 1102); and - to determine, based on the plurality of sequences of intersection points (272), whether the plurality of lane markings (1101, 1102) corresponds to a corresponding plurality of lane boundaries (152) for a group of a plurality of parallel lanes (151). [8] Device (120) according to one of the preceding claims, wherein the device (120) is arranged - to determine a plurality of measured travel paths (160) of one or more vehicles (110) for a corresponding plurality of passages through the roadway section (410); - to determine a reference travel path (210) for the road section (410) on the basis of the plurality of measured travel paths (160); and - to determine on the basis of the reference travel path (210), in particular on the basis of an arrangement of the reference travel path (210) relative to the lane marking (1101, 1102), whether the lane marking (1101, 1102) is a lane boundary (152) or not. [9] Device (120) according to one of the preceding claims, wherein the device (120) is arranged - to detect a first lane marking (1101) and a second lane marking (1101) in the lane section (410) on the basis of sensor data; - to determine arrangement information relating to an arrangement of the first and second road markings (1101) relative to the reference course (700) and / or relative to each other; and - to determine on the basis of the arrangement information whether - the first and second road markings (1101) are lane boundaries (152); and - the first and second lane markings (1101) are part of a group of several parallel lanes (151). [10] Device (120) according to one of the preceding claims, wherein the device (120) is arranged - to provide the detected lane boundary (152) as map data (125) relating to the road section (410) for a digital map; and / or - to use the detected lane markings (152) for automated longitudinal and / or transverse guidance of a vehicle (110). [11] Method (1110) for determining a lane boundary (152) on a roadway section (410); the method (1110) comprising - determining (1111) a reference course (700) for the road section (410); - detecting (1112), on the basis of sensor data, a road marking (1101, 1102) in the road section (410); - determining (1113) arrangement information relating to an arrangement of the road marking (1101, 1102) relative to the reference course (700); and - Determining (1114), based on the arrangement information, whether the lane marking (1101, 1102) is a lane boundary (152) or not.

Images