Method for aligning a map section of a digital road map

Abstract

The invention relates to a method for aligning a first map section of a digital road map with a second map section of the digital road map that at least partially overlaps the first map section, comprising the following steps: determining that a first relative rotation between the two map sections cannot be uniquely and unambiguously ascertained, determining a second relative rotation between a third map section of the digital road map and a fourth map section of the digital road map that at least partially overlaps the third map section, the third map section and the fourth map section being adjacent to the first map section and the second map section, and aligning the first map section with the second map section, the aligning comprising ascertaining a relative rotation between the first map section and the second map section on the basis of the ascertained second relative rotation. The invention relates to a device, a computer program and a machine-readable storage medium.

Description

Technical Field

[0001] The invention relates to a method, a device, a computer program and a machine-readable storage medium for aligning a first map section of a digital road map with a second map section of the digital road map which at least partially overlaps the first map section. Background Art

[0002] Publication CN 1 13 192 142 A discloses a map architecture method.

[0003] Publication CN 1 13 066 105 A discloses a positioning and mapping system.

[0004] Publication CN 1 15 077 519 A discloses a positioning and mapping method.

[0005] In the case of simultaneous localization and mapping (SLAM), sensor data must be regularly aligned and fused with map data from different trips. It is in the case of camera-based SLAM systems involving automobiles that road markings play a decisive role. Road markings constitute a basic element in HD mapping based on their good visibility and wide availability. At the same time, vehicles can be relatively accurately and mainly laterally located based on road markings. In order to accurately construct a map, it is often not enough to only use the route's travel, and instead the map becomes a so-called crowdsourced map composed of an entire fleet of motor vehicles. For this purpose, the individual maps that have been created by a motor vehicle are aligned and fused with other maps that have been created by other motor vehicles. However, the basic problem in mapping and positioning is that only a local part of the map is used, but only one road marking is identified in multiple road areas. This can occur due to perceived errors. However, it is also possible that only one road marking exists in a road section for a long time, for example, the right lane marking is the only road marking in some roads. Since many algorithms in HD (High Definition) road mapping are based on calculating relative rotations and translations between two map sections, the critical case is often the case in which only one road marking exists or can be matched. In this case, a unique and unambiguous roll angle cannot be determined. Summary of the invention

[0006] The object on which the present invention is based can be seen as providing a concept for aligning a first map section of a digital road map with a second map section of the digital road map which at least partially overlaps the first map section.

[0007] This object is achieved by means of the solution according to the invention. Advantageous embodiments of the invention are the subject of further developments.

[0008] According to a first aspect, a method for aligning a first map section of a digital road map with a second map section of the digital road map at least partially overlapping the first map section is provided, the method comprising the following steps:

[0009] It is determined that a first relative rotation between the two map sections cannot be determined uniquely and unambiguously,

[0010] determining a second relative rotation between a third map segment of the digital road map and a fourth map segment of the digital road map that at least partially overlaps the third map segment, wherein the third map segment and the fourth map segment are adjacent to the first map segment and the second map segment,

[0011] The first map segment and the second map segment are aligned, wherein the alignment includes ascertaining a relative rotation between the first map segment and the second map segment based on the ascertained second relative rotation.

[0012] According to a second aspect, an apparatus is provided, which is configured to carry out all the steps of the method according to the first aspect.

[0013] According to a third aspect, there is provided a computer program comprising instructions which, when executed by a computer, for example by an apparatus according to the second aspect, cause the computer to carry out the method according to the first aspect.

[0014] According to a fourth aspect, there is provided a machine-readable storage medium having stored thereon a computer program according to the third aspect.

[0015] The invention is based on and includes the following recognition: The above-mentioned object is achieved by using information from a neighboring map segment of the two map segments to be aligned, here a second relative rotation, in the presence of degenerate conditions, i.e., if a first relative rotation between two map segments cannot be determined uniquely and unambiguously. Therefore, according to the approach described here, the determined second relative rotation between the third map segment and the fourth map segment is used to align the first map segment with the second map segment. Now, according to the approach described here, the first relative rotation, which was previously not uniquely and unambiguously determined, is determined using the determined second relative rotation.

[0016] The first map segment and the second map segment can therefore also be aligned when degradation conditions are present.

[0017] A degenerate situation is a situation in which it is determined or confirmed that a first relative rotation between two map segments cannot be uniquely and unambiguously ascertained. This is, for example, the case in which only one road marking is present or detected in one of the map segments, whereas in another of these map segments, a plurality of road markings are present. This is, for example, the case in which only one road marking is present or detected in each of the two map segments.

[0018] If the relative rotation between the two map sections to be aligned cannot be determined uniquely and unambiguously, there is a so-called roll angle ambiguity.

[0019] In order to resolve such roll angle ambiguity, according to the approach described here, additional information is used from adjacent map segments: a second relative rotation between two adjacent map segments, ie, a roll angle between two adjacent map segments.

[0020] For example, it is provided that a roll angle, ie a relative rotation, between two adjacent map segments is propagated to the two map segments to be aligned.

[0021] According to one specific embodiment of the method, it is therefore provided that the ascertainment of the relative rotation between the first map segment and the second map segment includes the propagation of the ascertained second relative rotation into the first map segment and into the second map segment.

[0022] This achieves, for example, the technical advantage that the relative rotation between the first map section and the second map section can be efficiently ascertained.

[0023] In one embodiment of the method, it is provided that a relative delta pose between a first map segment and a second map segment is determined, wherein a set of inliers is determined based on the determined delta pose, wherein a principal axis transformation of the determined set of inliers is performed, wherein a first relative rotation between the two map segments is determined which cannot be uniquely determined based on the principal axis transformation.

[0024] This achieves, for example, the following technical advantage: the presence of a degradation condition can be efficiently identified or determined.

[0025] In one embodiment of the method, it is provided that the principal axis transformation comprises performing a singular value decomposition of a set of interior points In order to ascertain the principal axis, it is determined based on the singular value decomposition that a first relative rotation between the two map sections cannot be ascertained uniquely and unambiguously.

[0026] This achieves the technical advantage that, for example, it is possible to efficiently determine when a degradation condition exists.

[0027] In one embodiment of the method, it is provided that if two of the three singular values ​​obtained corresponding to the singular value decomposition are less than or equal to a predetermined limit, it is confirmed that all internal points are located on the main axis, wherein it is determined that a first relative rotation between the two map segments cannot be uniquely and unambiguously determined.

[0028] This achieves, for example, the technical advantage that degradation conditions can be efficiently detected.

[0029] In one embodiment of the method, it is provided that a set of inner points is obtained according to a nearest neighbor heuristic using a distance limit, in such a way that if there is a point in the second map segment that is within the distance limit relative to a corresponding point of the road marking represented by the second map segment, each point of the road marking represented by the first map segment is confirmed to be an inner point.

[0030] This achieves, for example, the following technical advantage: a set of interior points can be efficiently determined.

[0031] Device features are derived analogously from corresponding method features and vice versa.

[0032] For example, the method is carried out with the aid of the device.

[0033] For example, the method is a computer-implemented method.

[0034] The alignment between two map segments includes, for example, determining a relative delta attitude between the two map segments.

[0035] In the sense of this description, the third and / or fourth map segment is, for example, a map segment that is directly adjacent with respect to the first map segment and / or with respect to the second map segment.

[0036] For example, it is confirmed that the relative rotation between the first map segment and the second map segment is the same as the second relative rotation between the third map segment and the fourth map segment.

[0037] The relative rotation may also be called the roll angle.

[0038] Determining the relative Delta attitude includes, for example, determining the relative rotation and the relative translation.

[0039] A map section in the sense of this description is already determined or is determined, for example, using surroundings sensor data of one or more surroundings sensors of the motor vehicle.

[0040] An environmental sensor in the sense of this description is, for example, one of the following environmental sensors: a radar sensor, a lidar sensor, an image sensor, in particular an image sensor of a video camera, an ultrasonic sensor, an infrared sensor and a magnetic field sensor.

[0041] Thus, a map segment in the sense of this description can be generated or created, for example, using lidar data and / or radar data and / or video camera data and / or infrared data and / or magnetic field data and / or ultrasonic data.

[0042] Environmental sensor data in the sense of this specification describe the surroundings of the motor vehicle.

[0043] A road marking in the sense of this specification is, for example, one of the following road markings: longitudinal marking, zone marking Boundary markings, transverse markings, parking area markings.

[0044] Road markings may also be called lane markings.

[0045] The longitudinal markings include, for example, one of the following longitudinal markings: guide line, warning line, lane boundary, lane boundary, one-sided lane boundary, departure line, double lane boundary.

[0046] The longitudinal markings may include, for example, the following: a continuous line, a discontinuous line, a double line which is both continuous and discontinuous, a continuous double line, a discontinuous double line.

[0047] The embodiments and implementations described here can be combined with one another in any manner, even if this is not explicitly described.

[0048] For example, the device is programmed to execute the computer program. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] The present invention will be further explained below based on preferred embodiments. Herein, it is shown:

[0050] Figure 1 : A flow chart of the method according to the first aspect;

[0051] Figure 2 : According to the device of the second aspect,

[0052] Figure 3 : A machine-readable storage medium according to the fourth aspect,

[0053] Figure 4 : Two map sections of the digital road map before adjustment,

[0054] Figure 5 : After adjustment, Figure 4 Two map sections of

[0055] Figure 6 : Two map sections of the digital road map before adjustment,

[0056] Figure 7 : After adjustment, Figure 6 Two map sections of

[0057] Figure 8 : Two map sections of the digital road map before calibration.

[0058] In the following, the same reference signs may be used for the same features. DETAILED DESCRIPTION

[0059] Figure 1 A flow chart showing a method for aligning a first map section of a digital road map with a second map section of the digital road map which at least partially overlaps the first map section, comprising the following steps:

[0060] Determination 101: A first relative rotation between two map sections cannot be uniquely determined.

[0061] determining 103 a second relative rotation between a third map segment of the digital road map and a fourth map segment of the digital road map that at least partially overlaps the third map segment, wherein the third map segment and the fourth map segment are adjacent to the first map segment and the second map segment,

[0062] The first map segment is aligned 105 with the second map segment, wherein the alignment includes ascertaining 107 a relative rotation between the first map segment and the second map segment based on the ascertained second relative rotation.

[0063] Figure 2 A device 201 is shown which is configured to carry out all steps of the method according to the first aspect.

[0064] Figure 3 A machine-readable storage medium 301 is shown, on which a computer program 303 is stored. The computer program 303 comprises instructions which, when the computer program 303 is executed by a computer, cause the computer to carry out the method according to the first aspect.

[0065] Figure 4 Two map sections of a digital road map to be aligned are shown. Figure 4 A first map section 401 and a second map section 403 are shown in detail.

[0066] The first map section 401 includes a first road 405 . The second map section 403 includes a second road 407 .

[0067] The first road 405 has two solid lines 409 , 411 as lane boundary markings. In addition, a dashed line 413 extends as a center line between the two lane boundaries 409 , 411 .

[0068] For the second map section 403 , three solid lines 415 , 417 , 419 are identified, as well as a dashed line 421 and a short line 423 , which appears to belong to the dashed line 421 .

[0069] In this case, the two solid lines 415 , 417 are very close to each other.

[0070] Solid lines 415, 417 and 419 form the boundary lines or lane boundaries of road 407. Dashed lines 421 are medians. Short lines 423 are, for example, misidentifications.

[0071] The two map sections can be aligned using an alignment algorithm known per se, also called “alignment” in English. Figure 5 This is shown by way of example.

[0072] Figure 6 A first map section 601 and a second map section 603 to be aligned are shown.

[0073] Different from according to Figure 4 The first map section 401, about Figure 6 In the first map segment 601 of FIG. 4 , only a dashed line 413 is present in order to align the two map segments. Figure 7 After the alignment of the two map segments in , the roll angle cannot be determined uniquely. This alignment has been shown to be degenerate. Therefore, a degenerate situation exists. The roll angle is Figure 7 This is illustrated by way of example by a circular arrow having the reference numeral 701 .

[0074] Figure 8 The first map segment 801 and the second map segment 803 are shown at least partially overlapping, so that they can be aligned based on their overlap.

[0075] According to the scheme described herein, in the presence of a degradation condition, which is Figure 6 and Figure 7 In the case shown by way of example - a setting is made in which information from adjacent map segments is used: the scroll angle between two adjacent map segments.

[0076] Therefore, the approaches described herein are based, inter alia, on identifying or detecting degradation conditions.

[0077] As part of the calibration, for example, lane markings of the map segment are linked.

[0078] The approach described here therefore detects degraded lane associations, in particular line associations, wherein the problem of the geometric underdetermination of the association is solved with the aid of additional information.

[0079] Association can also be called "matching".

[0080] The two map segments are aligned, for example, by an algorithm, wherein corresponding road markings, in particular lines, are associated. Such algorithms are known per se. In particular, it is provided that, based on the alignment result, it is determined whether a degradation condition exists. If such a degradation condition exists, the associated road markings are essentially all located on a straight line. For example, it can be determined based on a principal axis analysis whether the point cloud is a close approximation to a straight line.

[0081] In order to resolve the roll angle ambiguity, for example, one or more additional information from two adjacent map segments that at least partially overlap is used. In this case, for example, it is provided that the two adjacent map segments are aligned, wherein, for example, corresponding roadway markings are linked within the scope of the alignment. If no degenerate conditions exist in this case, the roll angle determined accordingly between the two adjacent map segments can be used in order to determine the roll angle between the map segments that are initially to be aligned.

[0082] For example, it is provided that the inliers are found based on the relative delta pose between the two map segments to be aligned. For example, the inliers are found based on the nearest neighbor heuristic, especially when using a distance limit. This means that if there is a point in the second map segment that is sufficiently close to a point from the second map, then every point (of the line) from the first map segment belongs to an inlier. Now, for example, consider the set of all inliers. For example, the principal axis is calculated based on the set of all inliers. For example, the principal axis can be estimated by singular value decomposition. If, for example, two of the three singular values ​​are very small, then all the inliers are located on one axis and this is a degenerate condition.

[0083] If a degradation condition is detected, this situation is eliminated by adding one or more additional information, in particular a second roll angle. Additional adjacent map segments are used for this. If no degradation condition is detected for these adjacent map segments within the scope of the alignment of the two map segments, the roll angle between the two adjacent map segments can be propagated, for example, to the map segment pair with the degraded alignment. In order to make this possible, the roll angle can be propagated further, for example, it is provided that the adjacent map segments are located within the same digital road map. For example, adjacent map segments with sufficient overlap within the same road map are used. If a digital road map has been generated or is to be generated with the help of additional information, such as an inertial sensor device, an odometer (Odometrie), etc., this additional information can also be used, for example, in order to determine the relative rotation between the first and second map segments, that is, in order to propagate the second roll angle, for example.

[0084] The approach described here does not necessarily have to focus on lines; point features such as columns or planes etc. can also be used for alignment. For example, a mixed use of geometric types is provided.

[0085] The map segment can be generated, for example, from camera data and, for example, additionally or alternatively, also from data from other surrounding sensors, for example also from lidar sensors or radar sensors and / or other surrounding sensors, as already described above.

Claims

1. A method for aligning a first map section (401, 601, 801) of a digital road map with a second map section (403, 603, 803) of the digital road map which at least partially overlaps the first map section (401, 601, 801), the method comprising the following steps: It is determined (101) that a first relative rotation between the two map sections cannot be determined uniquely and unambiguously, A second relative rotation is determined (103) between a third map section of the digital road map and a fourth map section of the digital road map which at least partially overlaps the third map section, wherein: The third map segment and the fourth map segment are adjacent to the first map segment (401, 601, 801) and the second map segment (403, 603, 803), Aligning (105) the first map segment (401, 601, 801) with the second map segment (403, 603, 803), wherein the alignment includes determining (107) a relative rotation between the first map segment (401, 601, 801) and the second map segment (403, 603, 803) based on the determined second relative rotation.

2. The method according to claim 1, wherein: Determining the relative rotation between the first map segment (401, 601, 801) and the second map segment (403, 603, 803) includes propagating the determined second relative rotation into the first map segment (401, 601, 801) and into the second map segment (403, 603, 803).

3. The method according to claim 1 or 2, wherein: A relative delta attitude between the first map segment (401, 601, 801) and the second map segment (403, 603, 803) is determined, wherein a set of inner points is determined based on the determined delta attitude, wherein a principal axis transformation of the determined set of inner points is performed, wherein a first relative rotation between the two map segments that cannot be uniquely determined is determined based on the principal axis transformation.

4. The method according to claim 3, wherein: The principal axis transformation includes performing a singular value decomposition of the set of interior points in order to determine the principal axis, wherein a first relative rotation between the two map segments which cannot be uniquely and unambiguously determined is determined based on the singular value decomposition.

5. The method according to claim 4, wherein: If two of the three singular values ​​ascertained according to the singular value decomposition are less than or less than or equal to a predetermined limit, it is determined that all inner points lie on the main axis, wherein it is determined that a first relative rotation between the two map sections cannot be uniquely and unambiguously ascertained.

6. The method according to any one of claims 3 to 5, wherein: The set of inner points is obtained according to the nearest neighbor heuristic method using a distance limit, in that if there is a point in the second map segment (403, 603, 803) that is within the distance limit relative to a corresponding point of a lane marking (415, 417, 419, 421, 423) represented by the second map segment, then each point of the lane marking (409, 411, 413) represented by the first map segment is confirmed to be an inner point.

7. A device (201) arranged to carry out all the steps of the method according to any one of the preceding claims.

8. A computer program (303) comprising instructions which, when the computer program (303) is executed by a computer, cause the computer to carry out the method according to any one of claims 1 to 6.

9. A machine-readable storage medium (301) on which a computer program (303) according to claim 8 is stored.

Images