Method and Control Unit for Detecting a Lane Boundary

Abstract

A control unit for a vehicle is designed to detect movement sensor data with regard to a movement of at least one component of the vehicle. The movement is or has been effected by a roadway driven upon by the vehicle. In addition, the control unit is designed to detect a lane boundary of the roadway on the basis of movement sensor data and, in reaction thereto, to cause a functional reaction of the vehicle.

Claims

1.-10. (canceled)

11. An apparatus for a vehicle, comprising: a control unit operatively configured to: record motion sensor data in relation to a movement of at least one component of the vehicle, wherein the movement is or was brought about by a roadway being traveled by the vehicle; recognize a traffic lane boundary of the roadway on the basis of the motion sensor data; and bring about a functional response of the vehicle in response thereto.

12. The apparatus according to claim 11, wherein the motion sensor data are recorded by way of at least one of: a motion sensor, an acceleration sensor, and a tactile sensor of the vehicle; and the sensor comprises a level sensor and/or a wheel rotational speed sensor of a wheel of the vehicle.

13. The apparatus according to claim 11, wherein the motion sensor data display a temporal profile of a measured variable of the movement of the component of the vehicle; and the measured variable comprises at least one of: a rotational speed of a wheel of the vehicle, a deflection of a wheel suspension system of the vehicle, and an acceleration of the wheel rotational speed and/or of the deflection.

14. The apparatus according to claim 11, wherein the traffic lane boundary that is recognized has a profile perpendicular to a surface of the roadway being traveled by the vehicle that differs from a basic profile of the roadway; and the profile of the traffic lane boundary comprises: one or more spatially limited sections that are raised with respect to the roadway and/or a degree of roughness that is different from the basic profile.

15. The apparatus according to claim 11, wherein the control unit is further operatively configured to: determine a type of traffic lane boundary from a plurality of different types of traffic lane boundaries on the basis of the motion sensor data, wherein the plurality of different types of traffic lane boundaries comprises at least two of: a traffic lane marking, a shoulder next to the roadway being traveled by the vehicle, a border line, a guideline and cat's eyes arranged on the roadway.

16. The apparatus according to claim 11, wherein the control unit is further configured operatively to: determine a driving situation in relation to the traffic lane boundary from a plurality of different types of driving situations on the basis of the motion sensor data, wherein the plurality of different types of driving situations comprises at least two of: leaving a roadway traffic lane being traveled by the vehicle, changing between different traffic lanes of the roadway, and leaving the roadway.

17. The apparatus according to claim 11, wherein the control unit is further operatively configured to: determine position data in relation to a position of the vehicle; and recognize the traffic lane boundary on the basis of the position data and of digital map information, wherein the digital map information displays the position and/or a type of traffic lane boundaries for different roadways of a road network.

18. The apparatus according to claim 11, wherein the control unit is further operatively configured to: determine contactlessly recorded sensor data in relation to the roadway, wherein the contactlessly recorded sensor data comprise at least one of: image data from an image camera and reflection data from a laser sensor and/or radar sensor; and recognize the traffic lane boundary of the roadway on the basis of the contactlessly recorded sensor data.

19. The apparatus according to claim 11, wherein the control unit is further operatively configured to detect the traffic lane boundary by way of a machine-trained classifier.

20. The apparatus according to claim 14, wherein the degree of roughness is greater than that of the basic profile.

21. A method for providing a driving function in a vehicle, wherein the method comprises: recording motion sensor data in relation to a movement of at least one component of the vehicle, wherein the movement is or was brought about by a roadway being traveled by the vehicle; recognizing a traffic lane boundary of the roadway on the basis of the motion sensor data; and bringing about a functional response of the vehicle in response thereto.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0024] FIGS. 1a to 1c show exemplary traffic lane boundaries.

[0025] FIG. 2a shows exemplary components of a vehicle.

[0026] FIG. 2b shows an exemplary wheel of a vehicle on a traffic lane boundary.

[0027] FIGS. 2c and 2d show exemplary temporal profiles of sensor data from tactile sensors.

[0028] FIG. 3 shows a flowchart of an exemplary method for recognizing a traffic lane boundary.

DETAILED DESCRIPTION OF THE DRAWINGS

[0029] As set forth at the outset, the present document deals with the reliable and precise recognition of a traffic lane boundary of a roadway. In this connection, FIGS. 1a and 1b show a vehicle 100 that is driving on an exemplary roadway 110. The roadway 110 in FIG. 1a has lane markings 111, 112 painted onto the roadway 110, in particular a border line 111 and a guideline 112. The roadway 110 furthermore has a roadway edge 113 arranged next to the border line 111 (for example in the form of a tar verge or grass verge). The roadway 110 illustrated in FIG. 1b has what are known as cat's eyes 114 as a traffic lane boundary. The cat's eyes 114 are typically adhesively bonded to the roadway 110.

[0030] The traffic lane boundaries 111, 112, 113, 114 illustrated in FIGS. 1a and 1b typically each have, in the direction perpendicular to the roadway 110, a (height) profile that differs from the rest of the roadway 110. FIG. 1c shows the profiles of different traffic lane boundaries 111, 112, 113. FIG. 1c in particular shows the height profile 122 of a guideline 112 along the direction of travel of the roadway 110. FIG. 1c furthermore shows the height profile 121 of a border line 111 transverse to the direction of travel of the roadway 110. FIG 1c furthermore shows the height profile 123 of a roadway edge 113 transverse to the direction of travel of the roadway 110.

[0031] A traffic lane boundary, in particular a traffic lane marking, is nowadays recognized primarily on the basis of the optical recognition of a lane marking and/or on the basis of the reflection properties of the lane marking. For optical recognition purposes, a camera system is used and the recorded image data are evaluated by way of an object recognition algorithm in order to recognize a marking. The reflection properties of a lane marking may be evaluated using laser sensors and/or radar sensors that are configured so as to emit laser and/or radar beams that are reflected from a lane marking. It is however not always reliably possible to recognize traffic lane boundaries on the basis of the sensor data from image, radar and/or laser sensors, for example when traffic lane boundaries 111, 112, 113, 114 are soiled and/or covered.

[0032] As is apparent from FIG. 1c, traffic lane boundaries 111, 112, 113, 114 have height profiles 121, 122, 123 perpendicular to the surface of a roadway 110 that are able to be recorded by one or more tactile sensors of a vehicle 100, even when it is not possible to reliably record sensor data from image, radar and/or laser sensors. FIG. 2a shows a block diagram of a vehicle 100. The vehicle 100 may comprise one or more image, radar and/or laser sensors 202 that are configured so as to record sensor data in relation to a traffic lane boundary 111, 112, 113, 114. These sensor data are also referred to in this document as contactlessly recorded sensor data, since the sensor data are recorded using a contactless measurement method (image camera, laser beams and/or radar beams).

[0033] The vehicle 100 furthermore comprises one or more tactile sensors 201 that are configured so as to record a movement in the vehicle 100 brought about by the height profile 121, 122, 123. This is illustrated by way of example in FIG. 2b. FIG. 2b shows the wheel 212 of a vehicle 100 that is located on a marking section 211 at the times t1 and t2 and between two marking sections 211 of a traffic lane boundary 111, 112, 113, 114 at the time t1. The marking sections 211 bring about a movement 213 of the wheel 212. This movement 213 brought about by the traffic lane boundary 111, 112, 113, 114 may be recorded by one or more tactile sensors 201 of the vehicle 100. The level 223 of the wheel 212 may in particular be recorded as a function of time by way of a level sensor (for example on a wheel suspension system of the wheel 212 on the vehicle 100). FIG. 2c shows an exemplary temporal profile 224 of the level 223 for the marking sections 211 illustrated in FIG. 2b. As an alternative or in addition, the acceleration and/or the deceleration of a wheel 212 brought about by a traffic lane boundary 111, 112, 113, 114 may be recorded by way of a wheel rotational speed sensor. FIG. 2d by way of example shows the temporal profile 234 of the wheel rotational speed 233 for the marking sections 211 illustrated in FIG. 2b. As an alternative or in addition, the one or more tactile sensors 201 may comprise an acceleration sensor that is configured so as to record an acceleration and/or deceleration of a component, in particular a wheel 212, of the vehicle 100 brought about by a traffic lane boundary 111, 112, 113, 114.

[0034] The sensor data recorded by a tactile sensor 201 are also referred to as motion sensor data in this document, since the sensor data display a movement 213 of a component, in particular a wheel 212, of the vehicle 100 brought about by a traffic lane boundary 111, 112, 113, 114.

[0035] The vehicle 100 comprises a control unit 200 that is configured so as to recognize a traffic lane boundary 111, 112, 113, 114 on the basis of the (contactlessly recorded and/or motion) sensor data. The sensor data from different sensors 201, 202 of the vehicle 100 may in particular in this case be fused in order to reliably and precisely recognize a traffic lane boundary 111, 112, 113, 114. To this end, a machine-trained classifier may be used that is configured so as to recognize a traffic lane boundary 111, 112, 113, 114 and possibly a type of traffic lane boundary 111, 112, 113, 114 on the basis of the sensor data.

[0036] The control unit 200 is furthermore configured so as to bring about a functional response of the vehicle 100 in response to the recognition of a traffic lane boundary 111, 112, 113, 114. One or more actuators 203 (for example a steering system and/or a braking device) of the vehicle 100 may in particular be actuated for this purpose.

[0037] The control unit 200 may be configured so as to take into account position data from a position sensor 202 of the vehicle 100 when a traffic lane boundary 111, 112, 113, 114 is recognized, in particular in combination with digital map information that displays, for different positions, the respective traffic lane boundary 111, 112, 113, 114 that should be present for a roadway 110 at the respective position. The reliability and the quality of the recognition of traffic lane boundaries 111, 112, 113, 114 is thereby able to be further improved.

[0038] It is therefore proposed in this document to recognize a traffic lane boundary 111, 112, 113, 114 using sensors 201 that are already present and used in other ways in a vehicle 100, in particular using tactile sensors of the vehicle 100, such as the wheel rotational speed sensors and the level sensors. When driving over raised and/or structured traffic lane boundaries 111, 112, 113, 114, the wheel rotational speeds and/or the fully suspended travel of a wheel 212 of a vehicle 100 change in such a way that allows unambiguous identification of the event (that is to say driving over a traffic lane boundary 111, 112, 113, 114). As a result, in addition to painted traffic lane markings, it is also possible to reliably recognize further traffic lane boundaries. In addition, by evaluating motion sensor data, it is possible to draw a distinction between different driving situations in relation to a traffic lane boundary 111, 112, 113, 114, for example leaving a traffic lane, changing between different traffic lanes and/or leaving a roadway 110. The control unit 200 may thus be configured so as to recognize a particular driving situation in relation to a traffic lane boundary 111, 112, 113, 114 on the basis of the motion sensor data. It is then possible to bring about a suitable functional response of the vehicle 100 depending on the detected driving situation.

[0039] FIG. 3 shows a flowchart of an exemplary method 300 for providing a driving function in a vehicle 100 and/or for recognizing a traffic lane boundary 111, 112, 113, 114 of a roadway 110 being traveled by the vehicle 100. The driving function may comprise for example a driver assistance system and/or an at least partly autonomous driving function of the vehicle 100. The method 300 may be executed by a control unit 200 of the vehicle 100.

[0040] The method 300 comprises recording 301 motion sensor data in relation to a movement 213 of at least one component 212 of the vehicle 100, wherein the movement 213 is or was brought about by the roadway 110 being traveled by the vehicle 100. In other words, the movement 213 of the component 212 may be brought about by the roadway 110 when the vehicle 100 is driving on the roadway 110. In this case, the movement 213 that is brought about typically depends on the driving speed of the vehicle 100. The movement 213 of the component 212 may be brought about by a height profile 121, 122, 123 of the roadway 110 (perpendicular to the surface of the roadway 110). In this case, sensor data in relation to a movement 213 of a wheel 212 of the vehicle 100 that is in contact with the roadway 110 may in particular be recorded.

[0041] The method 300 furthermore comprises recognizing 302 a traffic lane boundary 111, 112, 113, 114 of the roadway 110 on the basis of the motion sensor data (typically taking into account the driving speed of the vehicle 100). The motion sensor data may in particular display a temporal profile 224, 234 of the movement 213 of the component 212 of the vehicle 100. A traffic lane boundary 111, 112, 113, 114 may in this case have a (height) profile 121, 122, 123 different from the rest of the roadway 110. As a result, the temporal profile 224, 234 of the movement 213 of the component 212 may differ depending on whether or not the vehicle 100 is located at least in part on the traffic lane boundary 111, 112, 113, 114. By way of example, reference profiles of the movement 213 of the component 212 may be stored for one or more types of traffic lane boundaries 111, 112, 113, 114. The recorded temporal profile 224, 234 of the movement 213 of the component 212 may then be compared with the one or more reference profiles of the movement 213 of the component 212 in order to determine whether or not the vehicle 100 is located at least in part on a traffic lane boundary 111, 112, 113, 114 and/or in order to recognize a traffic lane boundary 111, 112, 113, 114.

[0042] The method 300 may furthermore comprise bringing about 303 a functional response of the vehicle 100 in response to recognizing a traffic lane boundary 111, 112, 113, 114. A functional response of the vehicle 100 may in particular take place in the context of a driver assistance system (for example a lane keep assistant and/or a lane change assistant and/or an emergency braking function) and/or in the context of an at least partly autonomous driving function. One or more actuators 203 of the vehicle 100 (for example a braking device and/or a steering device) may be actuated for this purpose.

[0043] The measures described in this document make it possible to use an existing vehicle sensor system 201 to recognize traffic lane boundaries 111, 112, 113, 114. Efficient recognition of traffic lane boundaries 111, 112, 113, 114 is thereby made possible. In this case, reliable recognition of traffic lane boundaries 111, 112, 113, 114 is able to be made possible, even without a relatively expensive contactless sensor system 204 (such as an image camera and/or a radar sensor). The reliability when recording and modeling the surroundings of a vehicle 100 (that is to say when creating a surroundings model) is furthermore able to be increased by taking into account motion sensor data. In this case, it is possible to draw a distinction between different types of traffic lane boundaries 111, 112, 113, 114. A plausibility check may furthermore be performed on different driving situations and/or events (for example lane departure, lane change and/or road departure events). The motion sensor data may possibly, in addition to contactlessly recorded sensor data, be taken into account when recognizing traffic lane boundaries 111, 112, 113, 114 in order to further increase the reliability and the quality of the recognition. The motion sensor data may furthermore be used as a fallback level for recognizing traffic lane boundaries 111, 112, 113, 114 (for example if the contactless sensor system 204 fails). The reliability and the safety of driver assistance systems and/or of autonomous driving functions are thereby able to be further increased.

[0044] The present invention is not restricted to the exemplary embodiments that are shown. It should in particular be borne in mind that the description and the figures are intended only to illustrate the principle of the proposed methods, devices and systems.