US-12626520-B2 - Detection of intersection points of road marking lines

US12626520B2US 12626520 B2US12626520 B2US 12626520B2US-12626520-B2

Abstract

A method for detecting at least one intersection point of road marking lines, wherein a computing unit of a transportation vehicle receives a camera image and identifies a first road marking line based on the camera image. Using the computing unit, a search region is automatically defined in an environment of the first road marking line, and an automatic search limited to the search region is performed. At least one intersection point of the first road marking line with at least a second road marking line is identified as the result of the search.

Inventors

Timo Iken
Stefan Wappler
Roland KUBE
Niklas Koch

Assignees

VOLKSWAGEN AKTIENGESELLSCHAFT

Dates

Publication Date: 20260512
Application Date: 20220429
Priority Date: 20210507

Claims (20)

1 . A method for detecting at least one intersection point of road marking lines, wherein, by use of a computing unit of a transportation vehicle, the method comprising: receiving at least one camera image depicting a road surface in an environment of the transportation vehicle from a camera system of the transportation vehicle; identifying a first road marking line within the at least one camera image by use of the computing unit to cluster lines within the at least one camera image based on line orientation to form the first road marking line; automatically determining a search region in a predefined environment of the identified first road marking line, wherein the search region is determined using predefined rules for constructing the search region that determine position, orientation and size of the constructed search region that is specific to the identified first road marking line; and use of the computing unit to perform an automatic search restricted to the constructed search region to identify at least one intersection point of the first road marking line with at least one second road marking line in the constructed search region specific to the identified first road marking line to identify intersecting road marking lines that are related and intersect to signify an area of the road surface in the environment of the transportation vehicle.
2 . The method of claim 1 , wherein a line recognition algorithm for recognizing lines is applied to image data based on the at least one camera image to identify the first road marking line.
3 . The method of claim 2 , wherein the line recognition algorithm includes a random sample consensus algorithm and/or a Hough transformation algorithm.
4 . The method of claim 1 , wherein, for carrying out the automatic search, a feature recognition algorithm for recognizing corners and/or for recognizing other predefined characteristic intersection point descriptors is applied to the search region.
5 . The method of claim 4 , wherein the feature recognition algorithm includes a Harris detector algorithm.
6 . The method of claim 1 , further comprising: identifying a plurality of potential intersection points of the first road marking line with the at least one second road marking line as a result of the automatic search; selecting the at least one intersection point of the first road marking line from the plurality of potential intersection points based on at least one predefined rule.
7 . The method of claim 6 , wherein the at least one predefined rule contains a first rule concerning a target distance between two of the plurality of potential intersection points.
8 . The method of claim 6 , wherein the at least one predefined rule contains a second rule concerning an angle formed between the first road marking line and a straight line connecting two of the plurality of potential intersection points to one another.
9 . A method for self-localization of a transportation vehicle, the method comprising: generating at least one camera image depicting a road surface in an environment of the transportation vehicle by a camera system of the motor transportation vehicle; use of a computing unit of the transportation vehicle, to perform the method of claim 1 to detect the at least one intersection point of the road marking lines on the road surface; and use of the computing unit to compare the at least one intersection point of the first road marking line with a predefined digital map and to determine a position of the transportation vehicle in a predefined reference coordinate system based on a result of the comparison.
10 . A method for at least partially automatic control of a transportation vehicle, the method comprising: performing the method for the self-localization of the transportation vehicle of claim 9 ; and generating at least one control signal for the at least partially automatic control of the transportation vehicle based on the position of the transportation vehicle in the predefined reference coordinate system.
11 . A non-transitory computer readable medium including a computer program product comprising instructions which, when executed by a computer system, cause the computer system to perform the method of claim 1 .
12 . A detection system for a transportation vehicle for detecting at least one intersection point of road marking lines, the detection system comprising: a camera system configured to generate at least one camera image depicting a road surface in an environment of the transportation vehicle; and a computing unit configured to identify a first road marking line within the at least one camera image to cluster lines within the at least one camera image based on line orientation to form the first road marking line; wherein the computing unit is further configured to: automatically determine a search region in a predefined environment of the identified first road marking line, wherein the search region is determined using predefined rules for constructing the search region that determine position, orientation and size of the constructed search region that is specific to the identified first road marking line; and perform an automatic search restricted to the constructed search region, and identify at least one intersection point of the first road marking line with at least one second road marking line in the constructed search region specific to the identified first road marking line to identify intersecting road marking lines that are related and intersect to signify an area of the road surface in the environment of the transportation vehicle as the result of the search.
13 . The detection system of claim 12 , wherein the computing unit is configured to compare the at least one intersection point of the first road marking line with a predefined digital map and determines a position of the transportation vehicle in a predefined reference coordinate system based on a result of the comparison.
14 . An electronic vehicle control system for a transportation vehicle, the electronic vehicle control system comprising: the detection system of claim 13 ; and a control unit configured to generate at least one control signal for at least partial control of the transportation vehicle depending based on the position of the transportation vehicle in the predefined reference coordinate system.
15 . A transportation vehicle comprising the electronic vehicle control system of claim 14 .
16 . The detection system of claim 12 , wherein, for identifying the first road marking line, a line recognition algorithm for recognizing lines is applied to image data based on the at least one camera image.
17 . The detection system of claim 16 , wherein the line recognition algorithm includes a random sample consensus algorithm and/or a Hough transformation algorithm.
18 . The detection system of claim 12 , wherein, for carrying out the automatic search, a feature recognition algorithm for recognizing corners and/or for recognizing other predefined characteristic intersection point descriptors is applied to the search region.
19 . The detection system of claim 18 , wherein the feature recognition algorithm includes a Harris detector algorithm.
20 . The detection system of claim 12 , wherein: a plurality of potential intersection points of the first road marking line with the at least one second road marking line are identified as the result of the search; the at least one intersection point of the first road marking line is selected from the plurality of potential intersection points based on at least one predefined rule.

Description

PRIORITY CLAIM This patent application is a U.S. National Phase of International Patent Application No. PCT/EP2022/061454, filed 29 Apr. 2022, which claims priority to German Patent Application No. 10 2021 204 639.8, filed 7 May 2021, the disclosures of which are incorporated herein by reference in their entireties. SUMMARY Illustrative embodiments relate to a method for detecting at least one intersection point of road marking lines, wherein by use of a computing unit of a transportation vehicle, at least one camera image depicting a road surface in an environment of the transportation vehicle is received from a camera system of the transportation vehicle, and a first road marking line is identified on the basis of the at least one camera image. Illustrative embodiments also relate to a corresponding method for the localization of a transportation vehicle, a method for the at least partially automatic control of a transportation vehicle, a detection system for a transportation vehicle, an electronic vehicle control system, a transportation vehicle, and a computer program product. BRIEF DESCRIPTION OF THE DRAWINGS Exemplary embodiments are described below with reference to the figures, in which: FIG. 1 shows a schematic illustration of a transportation vehicle with an exemplary embodiment of a disclosed detection system; FIG. 2 shows a schematic flow diagram of an exemplary embodiment of a disclosed method for detecting at least one intersection point of road marking lines; FIG. 3 shows a schematic visualization of an action of the method of FIG. 2; FIG. 4 shows a schematic visualization of a further action of the method of FIG. 2; FIG. 5 shows a schematic visualization of a further action of the method of FIG. 2; and FIG. 6 shows a schematic visualization of a further action of the method of FIG. 2. DETAILED DESCRIPTION Driver assistance systems and systems for autonomous driving may use methods for the self-localization of the transportation vehicle, for example, for automated trajectory planning and the like. Many approaches for localization are based on the detection of landmarks in combination with a map comparison. Road marking lines are suitable for this, for example, particularly in applications for automated or partially automated parking. It is beneficial, in particular, to use intersection points of road marking lines for localization purposes. The document DE 10 2018 131 106 A1 describes a method for recognizing a parking region. That involves receiving from a camera of the transportation vehicle an image describing a parking area marking, in which image a corner element and a line element are recognized. The corner element and the line element are entered as separate representations in a digital environment map. Disclosed embodiments specify a possibility to enable intersection points of road markings to be detected with lower computational complexity for computation resources of the transportation vehicle. The disclosed embodiments are based on the concept of a multistage detection in which, firstly, a road marking line is identified and, in an automatically generated search region around the road marking line, an automatic search is carried out to identify an intersection point with a further road marking line. In accordance with at least one exemplary embodiment, a method for detecting at least one intersection point of road marking lines, in particular, of a first road marking line and a second road marking line, is specified. In this case, by use of a computing unit of a transportation vehicle, at least one camera image depicting a road surface in an environment of the transportation vehicle is received from a camera system of the transportation vehicle. By use of the computing unit, a first road marking line is automatically identified on the basis of the at least one camera image and, by the computing unit, a search region is automatically determined in a predefined environment of the first road marking line. By use of the computing unit, an automatic search restricted to the search region is carried out, and at least one intersection point of the first road marking line with at least one second road marking line is identified as the result of the search. The camera system can include one or more cameras mounted optionally at different positions of the transportation vehicle. The at least one camera image can be exactly one camera image, a plurality of camera images, or a combined image from different cameras. In this case, the output data of the camera system can be communicated to the computing unit without being changed, or can be partially preprocessed by the camera system itself before they are communicated to the computing unit. Alternatively or additionally, the computing unit can preprocess the data obtained from the camera system to generate the at least one camera image. The preprocessing by the camera system and/or the computing unit can include, for example, an equalization,