Method for guiding a motor vehicle laterally on a road and motor vehicle

Abstract

The present invention relates to a method for guiding a motor vehicle laterally on a road, in which a lane marking associated with a driver's side of the motor vehicle is associated with a lane type based on sensor data representing the surroundings of the motor vehicle, wherein the lane type characterizes: whether the lane marking is associated with a ego lane on which the motor vehicle is to be guided or with an adjacent lane adjacent to the ego lane; the lane marking associated with the driver's side of the motor vehicle is associated with the lane type based on cluster data received from another vehicle; if the lane type ascertained based on the sensor data and the lane type ascertained based on the cluster data are different, the lane type ascertained for the lane marking is determined to be the lane type ascertained based on the cluster data; and the motor vehicle is laterally controlled with the aid of a control device based on the determined lane type of the lane marking associated with the driver's side of the motor vehicle.

Description

Technical Field

[0001] The invention relates to a method for transversely guiding a motor vehicle on a road having at least two lanes and to a motor vehicle. Background Art

[0002] WO2021 / 043507A1 discloses a method for guiding a vehicle laterally, in which surrounding data of the vehicle is detected while driving on a road. In addition, stored surrounding data is obtained, which is detected by multiple other vehicles that are not currently driving the route while driving on the road. The rationality of these stored surrounding data is checked with the help of the detected surrounding data. The lateral guidance of the method is based on the surrounding data that has been checked for rationality. Using the surrounding data, at least one lane marking or lane delimitation on the left side from the perspective of the vehicle or a lane marking or lane delimitation on the right side from the perspective of the vehicle or at least one outer lane edge is detected as a surrounding characteristic.

[0003] Furthermore, US Pat. No. 10,955,855 B1 discloses a method for autonomous navigation, in which a travel segment is generated from a starting point to a destination. Using cameras and sensors, the method detects highway entrances, exit lanes, or highway dividers based on lane markings. A 3D model is generated in real time based on the data output by the cameras and sensors. This real-time 3D model is verified at the current location using a high-resolution map database. Based on the verified real-time 3D model, if the travel segment passes a highway entrance or exit, the vehicle maintains the current lane without exiting. Otherwise, the vehicle follows the highway entrance or exit. Summary of the Invention

[0004] The object of the present invention is to provide a solution which allows safe lateral guidance of a motor vehicle when the vehicle's cluster data and sensor data differ.

[0005] The object is achieved by a method for guiding a motor vehicle laterally on a road. In the method, lane markings associated with the driver's side of the motor vehicle are assigned to a lane type based on sensor data representing the surroundings of the motor vehicle, wherein the lane type characterizes whether the lane marking is assigned to a ego lane in which the motor vehicle is to be guided or to an adjacent lane adjacent to the ego lane; the lane markings associated with the driver's side of the motor vehicle are assigned to the lane type based on cluster data received from another vehicle; if the lane type ascertained based on the sensor data and the lane type ascertained based on the cluster data differ, the lane type ascertained based on the cluster data is determined as the lane type ascertained; and the motor vehicle is laterally controlled with the aid of a control device based on the determined lane type of the lane markings associated with the driver's side of the motor vehicle.

[0006] The present invention relates to a method for guiding a motor vehicle laterally on a road. The method is particularly advantageous on two-lane roads if the road centerline is missing or obscured, for example by dirt or snow, and therefore cannot be detected by a detection device of the motor vehicle. In particular, when there is oncoming traffic on this road having at least two lanes, it is important that the motor vehicle is safely guided in its lane in order to avoid a collision with the oncoming traffic.

[0007] The method provides for assigning lane markings associated with the driver's side of the vehicle to a lane type based on sensor data representing the vehicle's surroundings, wherein the lane type indicates whether the lane marking is assigned to the ego lane, in which the vehicle is to be guided, or to an adjacent lane adjacent to the ego lane. This means that lane markings associated with the driver's side of the vehicle are detected using a sensor device in the form of sensor data representing lane markings, which are located on the left side of the vehicle in right-hand traffic or on the right side of the vehicle in left-hand traffic. Based on the sensor data, in particular using an electronic computing device, it is determined whether the detected lane markings delimit the ego lane or the adjacent lane, and therefore whether they are assigned to the ego lane or the adjacent lane. The method also provides for assigning lane markings associated with the driver's side of the vehicle to a lane type based on cluster data received from another vehicle. Thus, based on the cluster data, it is determined whether the lane markings associated with the driver's side are assigned to the ego lane or the adjacent lane. The cluster data describes surrounding data previously recorded by another vehicle and provided to the vehicle. For this purpose, the cluster data can be provided, for example, by a server device for the motor vehicle.

[0008] Furthermore, the method provides that if the lane type of a lane marking assigned to the driver's side ascertained from the sensor data differs from the lane type ascertained from the cluster data, the lane type assigned to the lane marking is determined to be the lane type ascertained from the cluster data. An electronic computing device can perform a comparison between the lane type of the lane marking ascertained from the cluster data and the lane type of the lane marking ascertained from the sensor data. If the comparison determines that the lane type of the lane marking ascertained from the sensor data differs from the lane type ascertained from the cluster data, it is predefined that the lane type of the lane marking is the lane type ascertained from the cluster data. Consequently, the lane type of the lane marking ascertained from the sensor data is overcontrolled. In this case, it is assumed that there is an error in the sensor data, which results in a difference between the lane type of the lane marking ascertained from the sensor data and the lane type of the lane marking ascertained from the cluster data. This ensures that the vehicle can be safely controlled laterally even when there are errors in the sensor data.

[0009] Furthermore, the method provides for the vehicle to be laterally controlled with the assistance of a control device based on the lane type determined for lane markings associated with the driver's side of the vehicle. Depending on whether the lane type determined based on the cluster data is determined as delimiting the vehicle's lane or as delimiting an adjacent lane, the vehicle is laterally controlled accordingly to minimize the risk of a collision with oncoming traffic traveling in the adjacent lane. This allows for particularly safe operation of the vehicle. The vehicle can be laterally controlled based on the lane type determined for lane markings associated with the driver's side of the vehicle based on at least one predefined condition. For example, the distance of the vehicle from the lane markings associated with the passenger side of the vehicle is adjusted based on the determined lane type only if, as a predefined condition, the lane type determined based on the sensor data differs from the lane type determined based on the cluster data for the lane markings associated with the driver's side for at least a predefined traveled route and / or a predefined duration of travel. This prevents frequent, undesirable changes in the distance of the vehicle from the lane markings associated with the passenger side of the vehicle. As a result, the vehicle is guided centrally on the road, and the lane markings associated with the driver's side and the lane markings associated with the passenger side of the vehicle can each be determined as self-markings. The vehicle is guided laterally from one side of the road to the center after a predetermined route and / or a predetermined duration. The primary source for the lateral guidance of the vehicle is the shape of the lane markings associated with the passenger side of the vehicle, as detected in real time.

[0010] In one possible embodiment of the present invention, the vehicle is laterally controlled based on sensor data. The sensor data can be provided by a camera device of a sensor system, a radar device of a sensor system, a lidar device of a sensor system, and / or an ultrasonic device of a sensor system. Lateral control of the vehicle based on the sensor data allows for particularly safe control of the vehicle based on current data representing the detected surroundings of the vehicle. This allows the vehicle to be guided particularly safely along the road, particularly during short-term events.

[0011] In another possible alternative embodiment of the present invention, the vehicle is laterally controlled based on cluster data. This means that not only is the lane type determined based on the cluster data, but the vehicle's lateral control is also controlled using the cluster data by using the cluster data instead of the sensor data. In particular, controlling the vehicle laterally based on cluster data is particularly safe in the event of errors in the sensor data, especially since the cluster data has been validated on multiple vehicles.

[0012] In another possible embodiment of the present invention, if the lane type ascertained from the sensor data and the lane type ascertained from the cluster data differ for lane markings assigned to the passenger side of the motor vehicle, the lane type assigned to the lane marking is determined to be the lane type ascertained from the sensor data. In other words, if the lane type ascertained from the sensor data and the lane type ascertained from the cluster data differ and are therefore not identical for both lane markings assigned to the driver's side and lane markings assigned to the passenger side, the lane type assigned to the lane marking is determined to be the lane type ascertained from the sensor data. If it is determined that both the lane markings assigned to the driver's side and the lane markings assigned to the passenger side are assigned to different lane types by the cluster data and by the sensor data, then the highest probability of an error in the cluster data is determined, since the sensor data is generally considered more reliable than the cluster data. For example, there may be changes in the lane guidance system of the road that have occurred in the past and which are reflected in the sensor data but not yet in the cluster data. If the lane type determined for the lane marking assigned to the driver's side differs between the sensor data and the cluster data, but the lane type determined for the lane marking assigned to the passenger side is identical between the sensor data and the cluster data, the lane type assigned to the lane marking is determined as the lane type determined based on the cluster data. This allows the lane type of the lane marking assigned to the driver's side to be accurately determined with a particularly high probability, thereby keeping the risk of a collision of the motor vehicle with oncoming traffic or another vehicle particularly low.

[0013] In another possible embodiment of the present invention, the sensor data includes camera data, based on which lane markings associated with the driver's side of the vehicle are analyzed using an image recognition method. Furthermore, lane markings associated with the passenger side of the vehicle can be analyzed using an image recognition method based on the camera data. The camera data can particularly include corresponding images reflecting the vehicle's surroundings, particularly lane markings, which can be analyzed within the scope of the image recognition method for the presence of lane markings or for the type of lane markings and, therefore, the lane type. The camera data enables a particularly comprehensive recording of the vehicle's surroundings, thereby enabling a particularly informed determination of the lane type of the corresponding lane markings, particularly the lane type of the lane markings associated with the driver's side of the vehicle, to be made using an image recognition method, particularly with the aid of an electronic computing device.

[0014] In another possible embodiment of the present invention, if the cluster data meets at least one predefined quality criterion, the lane type assigned to the lane marking is determined as the lane type ascertained from the cluster data. If the cluster data does not meet the predefined quality criterion, the lane type assigned to the lane marking is determined as the lane type ascertained from the sensor data. The at least one quality criterion describes the reliability of the cluster data with respect to ascertaining the lane type of the lane marking. If the cluster data proves to be sufficiently safe, accurate, and reliable according to the at least one predefined quality criterion, the cluster data is then used only to determine the lane type of the lane marking assigned to the driver's side.

[0015] In this regard, another possible embodiment of the present invention may provide for a predefined quality criterion to be the rate of agreement between characteristic values ​​determined from cluster data and characteristic values ​​determined from sensor data for a defined route segment and / or the quality of the positioning of a vehicle on the road using the cluster data. This means that sensor data is compared with cluster data for a defined, predefined route segment, and the quality criterion is considered met if the cluster data and sensor data provide identical results within a predefined tolerance range for at least one predefined characteristic value. Such characteristic values ​​may, for example, be the distance of the vehicle from a predefined lane marking and / or the angular error of the vehicle. For example, a route segment of approximately 100 meters recently traveled by the vehicle may be considered the defined route segment. Therefore, cluster data is historically considered in comparison with sensor data. If the quality of the positioning of a vehicle on the road using the cluster data is predefined as the quality criterion, a check is performed to determine whether the quality of the vehicle's positioning based on the cluster data is sufficiently high. The positioning quality describes the accuracy with which the vehicle is located using the cluster data, particularly in a digital map. The positioning of the vehicle along the direction of travel may be considered. For example, an analysis can be performed to determine whether the vehicle can be located with meter accuracy, centimeter accuracy, and / or millimeter accuracy based on the cluster data. The positioning quality of the cluster data can be checked, for example, based on a corresponding curve in the road. In particular, the relative position of the vehicle on the road relative to the curve can be determined using sensor data, and a check can be performed to determine whether the relative position of the vehicle on the road relative to the curve determined based on the cluster data is consistent with this position, particularly within a predefined tolerance range. If the positioning quality of the vehicle's position determined based on the cluster data exceeds a predefined minimum positioning quality, the quality criterion is determined to be met. The predefined quality criterion ensures that the cluster data is used to determine the lane type of the corresponding lane marking only if the cluster data is of sufficient quality.

[0016] In another possible design of the present invention, it is provided that, during a preset time interval and / or while the motor vehicle passes a preset route and / or while the motor vehicle moves in the direction of travel at a speed within a preset speed range and / or when the road has a curvature above a preset curvature limit value and / or when a road with lane markings has a width within a preset width range and / or when the motor vehicle has a distance within a preset distance range relative to another lane marking assigned to the passenger side of the motor vehicle, the lane type assigned to the lane marking is determined to be the lane type ascertained based on the cluster data. This means that when the time interval has elapsed and / or the motor vehicle is located outside a predetermined route and / or is moving in or against the direction of travel at a speed outside a predetermined speed range and / or the curvature of the road is outside or equal to a predetermined curvature limit and / or the road with lane markings has a width outside a predetermined width range and / or the motor vehicle is spaced apart from another lane marking associated with the passenger side of the motor vehicle by a distance outside a predetermined distance range, the lane type assigned to the lane marking is determined to be the lane type ascertained from the sensor data. Therefore, corresponding boundary conditions for overruling sensor data using cluster data can be predefined when determining the lane type. If these boundary conditions are not met, the lane type ascertained from the sensor data for the lane marking associated with the driver's side of the motor vehicle cannot be overruled by the lane type ascertained from the cluster data. Therefore, clear limits can be predefined for when the lane type ascertained from the cluster data can overrule the lane type ascertained from the sensor data. Outside these limits, the lane type of the lane markings assigned to the driver's side is essentially predetermined by the sensor data, since the sensor data is assumed to be particularly accurate due to its currentness. This keeps the risk of the lane type of the lane markings assigned to the driver's side, ascertained from the sensor data, being incorrectly overruled by the lane type ascertained from the cluster data, particularly low.

[0017] In another embodiment of the present invention, if the lane type ascertained from the sensor data and the lane type ascertained from the cluster data are identical, the lane type associated with the lane marking associated with the driver's side of the vehicle is determined as the lane type ascertained from the sensor data. If the same lane type is ascertained for the lane marking associated with the driver's side of the vehicle both from the cluster data and from the sensor data, it is assumed that there is no error in the sensor data and that the sensor data can be used to ascertain the lane type. Therefore, regardless of whether the same lane type is ascertained from the sensor data and from the cluster data for the lane marking associated with the passenger side of the vehicle, if the same lane type is ascertained, it is determined for the lane marking associated with the driver's side that the sensor data should not be rejected for determining the lane type. In particular, since the sensor data is assumed to be particularly reliable for determining the lane type of lane markings assigned to the passenger side, the method provides that the cluster data overrides the sensor data in the evaluation of the lane type of lane markings assigned to the driver's side only if the evaluation of the lane type for lane markings assigned to the driver's side of the vehicle differs between the cluster data and the sensor data, but does not differ in the evaluation of the lane type for lane markings assigned to the passenger side of the vehicle. This allows the risk of an erroneous evaluation of the lane type of lane markings assigned to the driver's side of the vehicle to be kept particularly low. Consequently, the risk of a collision between the vehicle and other road users can be kept particularly low.

[0018] The present invention also relates to a motor vehicle that is designed to operate within the scope of the method described herein or an embodiment of the method according to the present invention. Advantages and advantageous embodiments of the method according to the present invention can be considered advantages and advantageous embodiments of the motor vehicle according to the present invention, and vice versa.

[0019] Further features of the invention can be derived from the following description of the drawings and with the aid of the drawings. The features and feature combinations mentioned above in the description of the drawings and the features and feature combinations shown individually in the following description of the drawings and / or in the drawings can be used not only in the respectively indicated combination but also in other combinations or alone without departing from the scope of the invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In the attached figure:

[0021] Figure 1A schematic top view of a road with two lanes is shown, on which a motor vehicle is traveling in its own lane and another road user is traveling in an adjacent lane, wherein the motor vehicle has a sensor device including a camera device, with the aid of which sensor data representing the surroundings of the motor vehicle can be recorded, and the motor vehicle receives cluster data from a higher-level server device, which allows the motor vehicle to be assisted in lateral control on the road;

[0022] Figure 2 A method diagram showing a method for laterally guiding a motor vehicle, wherein a lane type of lane markings associated with a driver's side of the motor vehicle is determined based on both sensor data and cluster data, and the motor vehicle is laterally controlled based on the respective determined lane type; and

[0023] Figure 3 A schematic diagram of a method for transversely guiding a motor vehicle on a road is shown. DETAILED DESCRIPTION

[0024] Identical or functionally identical elements are provided with the same reference symbols in the figures.

[0025] Figure 1 , a road 10 is shown in a top view, wherein road 10 currently has two lanes 12. Currently, there is no centerline separating lanes 12 from one another. Therefore, it may be difficult for a traffic user traveling on road 10 to distinguish how many lanes 12 road 10 has. Currently, a motor vehicle 14 and another traffic user 16 are traveling on road 10. Motor vehicle 14 is traveling in the ego lane, while another traffic user 16, as an oncoming vehicle, is traveling in the adjacent lane to the ego lane.

[0026] Motor vehicle 14 is configured to receive cluster data 18 from a higher-level server device 20. This cluster data 18 describes driving data of another vehicle that was moving along road 10 at an earlier point in time, particularly in the vehicle's lane. Furthermore, motor vehicle 14 includes a camera device 22 as part of a sensor system, which is configured to record camera data representing the surroundings of motor vehicle 14 as sensor data 32. To this end, a field of view 24 of camera device 22 is directed toward the surroundings of motor vehicle 14, particularly toward the surroundings that lie ahead of motor vehicle 14 in the direction of travel. Thus, camera device 22 can detect lane markings 26 that delimit road 10 outward. A first lane marking of lane markings 26 (hereinafter designated by reference numeral 28) faces the side-bounding road 10 associated with the driver's side of motor vehicle 14. A second lane marking 30 faces the side-bounding road 10 associated with the passenger side of motor vehicle 14.

[0027] In order to avoid a collision between motor vehicle 14 and other road users 16 during at least assisted lateral control of motor vehicle 14, it is necessary for motor vehicle 14 to be able to distinguish whether road 10 has only a single lane 12 and therefore motor vehicle 14 can be guided centrally on road 10, or whether road 10 has multiple lanes 12 and motor vehicle 14 must therefore remain safely in its own lane to avoid a collision between motor vehicle 14 and other road users 16 traveling in adjacent lanes. To make this determination, motor vehicle 14 includes an electronic computing device that can determine, based on sensor data 32 and / or cluster data 18, whether first lane marking 28 is assigned to an adjacent lane or to the own lane. If it is determined that first lane marking 28 is assigned to the own lane, motor vehicle 14 can be guided safely and centrally on road 10 because road 10 has only one lane 12. If the electronic computing device determines that first lane marking 28 is assigned to an adjacent lane, motor vehicle 14 must remain safely in its own lane with assisted lateral guidance in order to prevent a collision between motor vehicle 14 and further road users 16 who may be located in the adjacent lane, since road 10 has at least the own lane and the adjacent lane.

[0028] In addition to real-time lane recognition, future driver assistance systems may also utilize cluster data 18. This cluster data 18 can include a significant amount of highly current information and can be stored in the cloud. Furthermore, cluster data 18 may be based on data captured by a current series of vehicles, particularly using a front-facing camera, and stored cumulatively in the backend. This allows numerous vehicles to identify specific lane demarcations and upload these identified lanes to the cloud. There, the data is analyzed and made available to the vehicle in use, currently motor vehicle 14. In addition to cluster data 18, motor vehicle 14 can utilize the real-time lanes identified primarily by camera system 22, represented by sensor data 32. Based on this, lateral guidance of motor vehicle 14 can be provided. Using real-time detection by camera system 22, it can be determined, particularly using a neural network, whether the identified lane demarcations belong to the ego lane or to an adjacent lane. In particular, a driver assistance system without cluster data 18 may require two lane markings 26 associated with the ego lane, based on which motor vehicle 14 can be kept in the center of lane 12. If first lane marking 28 assigned to the driver's side of motor vehicle 14 is identified as an adjacent marking assigned to an adjacent lane, lateral guidance of motor vehicle 14 can be abandoned, taking into account the hysteresis. This means that incorrectly identified or incorrectly assigned lane marking 26 can have a direct impact on the driving behavior of motor vehicle 14 before notification. Cluster data 18 can significantly improve this behavior.

[0029] Motor vehicle 14 has a sensor system which may include a camera system 22 and / or radar and / or ultrasonic sensors, wherein motor vehicle 14 can additionally receive cluster data 18. Furthermore, motor vehicle 14 is provided for lateral control assisted by a control device of motor vehicle 14.

[0030] Incorrect lane assignments to adjacent lane markings are problematic. Furthermore, on narrow country roads, the system can exhibit incorrect or fluctuating, incomprehensible driving behavior, depending on how effectively the real-time lane detection is functioning at the given moment. In the prior art, real-time identified traffic lanes 12 are not overruled using cluster data 18. This means that real-time lane detection based on sensor data 32, based on which the motor vehicle 14 is controlled, has a higher priority in the algorithm than cluster data 18. To address this issue, in certain situations, information detected in real time by the camera system 22 should be overruled using cluster data 18. Thus, using cluster data 18, incorrect lane detection by the camera system 22 can be stabilized.

[0031] If the adjacent lane is identified as the adjacent lane, motor vehicle 14 is correctly controlled, motor vehicle 14 being able to remain in the own lane based on sensor data 32 and / or cluster data 18 and being guided, for example, at a distance of approximately 15 centimeters from second lane marking 30 assigned to the passenger side of motor vehicle 14. If the adjacent lane is incorrectly identified as the own lane, motor vehicle 14 may behave incorrectly by driving in the middle of the lane, in the center of road 10.

[0032] exist Figure 2 shows a method diagram for a method for transversely guiding a motor vehicle 14. In a first method step V1a, first lane marking 28 associated with the driver's side of motor vehicle 14 is assigned to a lane type based on received cluster data 18. Furthermore, in first method step V1a, second lane marking 30 associated with the passenger side of motor vehicle 14 is assigned to a lane type based on cluster data 18. The lane type indicates whether the corresponding lane marking 26 is assigned to the ego lane (in which motor vehicle 14 is to be guided) or to an adjacent lane adjacent to the ego lane.

[0033] In a further first method step V1b of the method, sensor data 32 representing the surroundings of motor vehicle 14 is used to assign a first lane marking 28 associated with the driver's side of motor vehicle 14 and a second lane marking 30 associated with the passenger side of motor vehicle 14 to a lane type. In a second method step V2, a comparison of the respectively determined lane types is performed for the respective lane markings 26. In this case, a check is performed in the first step to determine whether the lane type determined associated with first lane marking 28 based on cluster data 18 matches the lane type determined for first lane marking 28 based on sensor data 32. Thus, in the case of right-hand traffic, a check is performed in the first step to determine whether the lane demarcations determined based on cluster data 18 and sensor data 32, including lane types and lane assignments, are identical for the left side and, therefore, for first lane marking 28. In the second comparison step, the lane type determined based on sensor data 32 and the lane type determined based on cluster data 18 for second lane marking 30 associated with the passenger side of motor vehicle 14 are compared. In the case of right-hand traffic, in the second step of the method, it is therefore checked whether the lane delimitations ascertained from cluster data 18 and from sensor data 32 , including lane type and lane assignment, are identical for right-hand and therefore second lane marking 30 .

[0034] If, within the scope of the comparison, in second method step V2, it is determined that there is an identity in each of the first and second method steps, then in a third method step V3, the lane type associated with first lane marking 28 is determined for motor vehicle 14 as the lane type ascertained from sensor data 32, and motor vehicle 14 is laterally controlled based on sensor data 32. If, within the scope of the comparison, it is determined that there is no identity in each of the first and second steps, then in third method step V3, motor vehicle 14 is also laterally controlled based on sensor data 32, the lane type ascertained from sensor data 32 being determined as the lane type for first lane marking 28. If, within the scope of the comparison, it is determined that there is an identity in the first step but not in the second step, then in a third method step V3, motor vehicle 14 is laterally controlled based on sensor data 32, the lane type associated with first lane marking 28 being determined as the lane type ascertained from sensor data 32.

[0035] If the comparison determines that there is no identity in the first step, but there is an identity in the second step, the comparison is followed by a fourth method step V4, in which the lane type assigned to first lane marking 28 is determined to be the lane type ascertained from cluster data 18. Motor vehicle 14 is then laterally controlled with the aid of a control device according to the determined lane type of first lane marking 28. In particular, motor vehicle 14 is laterally controlled according to sensor data 32. Alternatively, motor vehicle 14 can be laterally controlled according to cluster data 18.

[0036] After the predefined travel time limit has expired, following fourth method step V4, motor vehicle 14 can continue to be controlled within the scope of third method step V3. In fourth method step V4, the lane type assigned to first lane marking 28 is determined to be the lane type ascertained from cluster data 18 only if cluster data 18 meets at least one predefined quality criterion. Two quality criteria are currently predefined. One of the quality criteria is the required consistency of characteristic values ​​ascertained from cluster data 18 with characteristic values ​​ascertained from sensor data 32 for a defined route section. The second quality criterion is the required quality of the positioning of motor vehicle 14 on road 10 as determined from cluster data 18.

[0037] In particular, a time interval can be predefined as a travel time limit and / or a route to be traversed can be predefined as a travel time limit and / or a speed range for motor vehicle 14 can be predefined as a travel time limit and / or a curvature limit value can be predefined as a travel time limit and / or a width range for road 10 can be predefined as a travel time limit and / or a distance range for the distance between motor vehicle 14 and second lane marking 30 can be predefined as a travel time limit. This means that within the scope of fourth method step V4, during the predetermined time interval and / or while motor vehicle 14 traverses the predetermined route and / or while motor vehicle 14 moves in the direction of travel at a speed within a predetermined speed range and / or when road 10 has a curvature above a predetermined curvature limit value and / or when road 10 with lane marking 26 has a width within a predetermined width range and / or when motor vehicle 14 is spaced apart from second lane marking 30 by a distance within a predetermined distance range, the lane type assigned to first lane marking 28 is determined to be the lane type ascertained from cluster data 18.

[0038] Basically, good real-time data for lane recognition (sensor data 32 currently provided by camera device 22) is generally preferred over potentially outdated map data and thus cluster data 18. In one approach, the use of cluster data 18 for overruling real-time data is therefore limited, while the problem described is nevertheless solved.

[0039] exist Figure 3 In more detail Figure 2 . Here, a scenario is illustrated in which, based on cluster data 18, it is determined that first lane marking 28 is to be assigned to an adjacent lane, while based on sensor data 32, it is determined that first lane marking 28 is to be assigned to the ego lane. This is determined within the scope of the comparison in second method step V2. Then, in fourth method step V4, the lane type ascertained based on sensor data 32 is rejected by cluster data 18 if the confidence measure of cluster data 18 is sufficiently high and cluster data 18 therefore meets at least one predefined quality criterion. Using multidimensional feature map 34, corresponding application parameters for the rejection of sensor data 32 by cluster data 18 can be determined for determining the lane type. Thus, a maximum route and / or a maximum time and / or speed range and / or a minimum confidence measure and / or a curvature range and / or a distance range can be predefined as application parameters, wherein the application parameters determine whether the lane type ascertained based on sensor data 32 is rejected by the lane type ascertained based on cluster data 18. These application parameters can predefine travel time limits. Until a situation 36 in which a travel time limit is reached, in particular exceeded, motor vehicle 14 is laterally controlled with the aid of a control device, depending on the lane type of first lane marking 28 determined based on cluster data 18. Motor vehicle 14 can also be additionally oriented during the lateral control according to second lane marking 30. If a situation 36 in which a travel time limit is reached occurs, motor vehicle 14 is then laterally controlled in a third method step V3 in which motor vehicle 14 is controlled according to sensor data 32, in particular according to the lane type of first lane marking 28 ascertained based on sensor data 32.

[0040] In summary, the present invention shows how to implement a method for stabilizing real-time lane detection by a front camera using cluster data 18 .

[0041] Reference Signs List

[0042] 10 Road

[0043] 12 lanes

[0044] 14 Motor Vehicles

[0045] 16 Traffic Participants

[0046] 18 Cluster Data

[0047] 20 Server Devices

[0048] 22 Camera Devices

[0049] 24 Field of View

[0050] 26 Lane Markings

[0051] 28 First lane marking

[0052] 30 Second lane marking

[0053] 32 sensor data

[0054] 34 Comprehensive characteristic curve

[0055] 36 situations

[0056] V1a to V4 Corresponding method steps.

Claims

1. A method for laterally guiding a motor vehicle (14) on a road (10), wherein: - assigning a lane marking (28) associated with the driver's side of the motor vehicle (14) to a lane type based on sensor data (32) representing the surroundings of the motor vehicle (14), wherein: The lane type characterizes whether the lane marking (28) is assigned to a self-lane on which the motor vehicle (14) is to be guided, or to an adjacent lane adjacent to the self-lane; - assigning a lane marking (28) associated with the driver's side of the motor vehicle (14) to the lane type based on cluster data (18) received from another vehicle; - if the lane type determined based on the sensor data (32) and the lane type determined based on the cluster data (18) are different, determining the lane type assigned to the lane marking (28) as the lane type determined based on the cluster data (18); and - laterally controlling the motor vehicle (14) with the aid of a control device in accordance with a determined lane type of a lane marking (28) associated with the driver's side of the motor vehicle (14).

2. The method according to claim 1, wherein The motor vehicle (14) is laterally controlled based on the sensor data (32).

3. The method according to claim 1, wherein The motor vehicle (14) is laterally controlled based on the cluster data (18).

4. The method according to any one of claims 1 to 3, wherein If the lane type determined based on the sensor data (32) and the lane type determined based on the cluster data (18) are different for another lane marking associated with the passenger side of the motor vehicle (14), the lane type associated with the lane marking (28) is determined to be the lane type determined based on the sensor data (32).

5. The method according to any one of claims 1 to 3, wherein The sensor data (32) include camera data, based on which lane markings (28) associated with the driver's side of the motor vehicle (14) are analyzed by image recognition methods.

6. The method according to any one of claims 1 to 3, wherein If the cluster data (18) meets at least one predefined quality criterion, the lane type assigned to the lane marking (28) is determined to be the lane type ascertained from the cluster data (18).

7. The method according to claim 6, wherein: The consistency rate of the characteristic values ​​determined from the cluster data (18) and the characteristic values ​​determined from the sensor data (32) for a defined route section is preset as a quality criterion and / or the positioning quality of the positioning of the motor vehicle (14) on the road (10) based on the cluster data (18) is preset as a quality criterion.

8. The method according to any one of claims 1 to 3, wherein - at predetermined intervals, and / or - during the passage of the motor vehicle (14) through the predetermined route, and / or - during movement of the motor vehicle (14) in the direction of travel at a speed within a predetermined speed range, and / or - when the road (10) has a curvature higher than a preset curvature limit value, and / or - when the road (10) having the lane marking (28) has a width within a predetermined width range, and / or - when the motor vehicle (14) is spaced apart from a further lane marking associated with the passenger side of the motor vehicle by a distance within a predetermined distance range, The lane type assigned to the lane marking (28) is determined as the lane type ascertained from the cluster data (18).

9. The method according to any one of claims 1 to 3, wherein If the lane type determined based on the sensor data (32) and the lane type determined based on the cluster data (18) are the same, the lane type associated with the lane marking (28) associated with the driver's side of the motor vehicle (14) is determined as the lane type determined based on the sensor data (32).

10. A motor vehicle (14) configured to carry out the method according to any one of claims 1 to 9.

Images