METHOD FOR FORMING A LOCALIZATION LAYER OF A DIGITAL LOCALIZATION MAP FOR AUTOMATED DRIVING

Abstract

A method for forming a localization layer for a digital localization map for automated driving. The method includes: providing the localization layer for a defined region; providing a planning layer for the region; and extracting alignment features from the localization layer that is provided for an alignment with the planning layer, the alignment features being extracted in such a way from the localization layer that an inadmissible deformation of the localization layer may be recognized during the alignment of the planning layer with the localization layer.

Claims

1-8. (canceled)

9. A method for forming a localization layer of a digital localization map for automated driving, the method comprising the following steps: providing the localization layer for a defined region; providing a planning layer for the region; and extracting alignment features from the localization layer provided for an alignment with the planning layer, the alignment features being extracted in such a way from the localization layer that an inadmissible deformation of the localization layer may be recognized during the alignment of the planning layer with the localization layer.

10. The method as recited in claim 9, wherein the alignment features are configured in a structurally defined manner.

11. The method as recited in claim 9, wherein the alignment features are configured according to a defined pattern.

12. The method as recited in claim 9, wherein the alignment features are configured according to at least one defined statistical criterion.

13. The method as recited in claim 12, wherein the statistical criterion includes at least one of the following: density distribution or variance.

14. The method as recited in claim 9, wherein an extent of a maximally admissible allowed distortion of the alignment features during the alignment of the localization layer with the planning layer is defined via a threshold value.

15. A device for forming a localization layer of a digital localization map for automated driving, the device comprising: a provision device configured to provide the localization layer for a defined region and to provide a planning layer for the region; and an extraction device configured to extract alignment features from the localization layer that is provided for an alignment with the planning layer, the alignment features being extracted in such a way from the localization layer that an inadmissible deformation of the localization layer may be recognized during the alignment of the planning layer with the localization layer.

16. A non-transitory computer-readable data medium on which is stored a computer program including program code for forming a localization layer of a digital localization map for automated driving, the computer program, when executed by an electronic device, causing the electronic device to perform the following steps: providing the localization layer for a defined region; providing a planning layer for the region; and extracting alignment features from the localization layer provided for an alignment with the planning layer, the alignment features being extracted in such a way from the localization layer that an inadmissible deformation of the localization layer may be recognized during the alignment of the planning layer with the localization layer.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0031] FIG. 1 shows a schematic illustration of a provided device for forming a localization layer of a digital localization map for automated driving, in accordance with an example embodiment of the present invention.

[0032] FIGS. 2 through 4 are figures for schematically elucidating the provided arrangement of alignment features of a localization layer, in accordance with an example embodiment of the present invention.

[0033] FIG. 5 shows a schematic illustration of a flowchart of a provided device for forming a localization layer of a digital localization map for automated driving, in accordance with an example embodiment of the present invention.

DETAILED DESCRIPTION OF EXAMPLE EMBODIMENTS

[0034] One feature of the present invention is in particular to make possible an improved alignment of the localization layer with the planning layer of a digital localization map. In particular, it is intended to make it possible to detect errors during the alignment of the above-mentioned layers and thus to efficiently control the manufacturing process of the digital localization map.

[0035] During the alignment process, non-linear transformations are preferably applied to the localization layer in order to align same exactly with regard to the planning layer. It must be ensured that the applied transformations do not destroy the digital localization map or render it useless (error control). Here, statistical measures are preferably carried out prior to and following the alignment process to determine whether it is possible that the alignment process is carried out inadmissibly or the alignment is carried out unsuccessfully.

[0036] If the alignment features were derived in such a way that they have relevant statistical properties or parameters (for example, density distribution, variance, etc.) of the localization layer, from which they were derived, the alignment features themselves may be used to assess an error that was introduced as a result of the non-linear transformation.

[0037] Advantageously, it is thus not necessary to use the entire data material of the above-mentioned layers to check the alignment process. This complies with data protection interests, for example, if the above-mentioned map layers originate from different manufacturers that are interested in data protection with regard to their own supplied map layers.

[0038] With the aid of the provided method in accordance with the present invention, the alignment process may be advantageously accelerated, since the computation of cyclic measures based on the reduced number of alignment features is in general much faster than when the total amount of data of the localization layer is used.

[0039] During the alignment process of the localization layer, the localization layer is subjected to non-linear transformations that aim to sufficiently precisely ascertain the identified alignment features in both layers (localization layer and planning layer). Since identifying matching alignment objects or features is an automated process, errors may occur (for example the localization layer being erroneously shifted in a way that renders it useless for an instantaneous localization process).

[0040] According to the present invention, statistical measures, such as for example a density distribution or a variance prior to and following the above-named non-linear transformation, are thus ascertained for the purpose of extracting alignment features. An analysis of these statistical measures may be used to ascertain erroneous transformations (for example if a difference between the density of alignment features or localization objects in an imagined region exceeds a threshold value following the transformation).

[0041] It is provided to compute the alignment features in a way that allows the error control to be carried out directly by using alignment features 1a . . . 10, it not being necessary to use the entire localization layer for this. For this purpose, the alignment features are selected in a way that allows errors to be easily recognizable.

[0042] The following illustrations show in a qualitative manner the provided principle of extracting the alignment features.

[0043] FIG. 1 shows a block diagram of a device 100 for forming a localization layer of a digital localization map for automated driving. Device 100 includes a provision device 10 for providing localization layer 12 for a defined region and for providing a planning layer 11 for the region. Furthermore, device 100 includes an extraction device 20 for extracting alignment features 1a . . . 1n from localization layer 12 that is provided for an alignment with planning layer 11, alignment features 1a . . . 1n being extractable in such a way from localization layer 12 that an inadmissible deformation of localization layer 12 may be recognized during the alignment of planning layer 11 with localization layer 12.

[0044] FIG. 2 shows a schematic illustration of a localization layer 12 including landmarks that were detected with the aid of a radar sensor, for example.

[0045] FIG. 3 shows localization layer 12 including conventional alignment features 1a . . . 1d extracted from localization layer 12 from FIG. 2. Here, a statistical distribution of alignment objects 1a . . . 1d does not correspond to an arrangement of the landmarks from FIG. 2.

[0046] In FIG. 4, it is indicated that alignment features 1a . . . 1n are now situated in a structured manner according to the present invention, in the case of FIG. 4 merely qualitatively and in the form of squares by way of example, a statistical distribution of the alignment objects corresponding to an arrangement of the landmarks from FIG. 2 in this case.

[0047] It is apparent that by positioning alignment features 1a . . . 1n from FIG. 4 in an orderly manner, which is represented merely illustratively and by way of example, an inadmissible distortion of alignment features 1a . . . 1n of localization layer 12 with regard to planning layer 11 (not illustrated) is easily recognizable and thus an erroneous alignment of localization layer 12 with planning layer 11 may be easily established. As a result, a correct manufacturing process of the digital localization map is thus already supported at an early stage.

[0048] Advantageously, the surroundings data provided by the method according to the present invention may be used to create very accurate and robust digital localization maps that are as complete and highly available as possible. It is advantageous that the general principle is very simple and allows for many implementations and applications that are described above only by way of example.

[0049] Device 100 is preferably designed in the cloud as a computational device, on which a computer program product is carried out.

[0050] FIG. 5 shows a schematic flowchart of a provided method for forming a localization layer of a digital localization map for automated driving.

[0051] In a step 200, a provision of a localization layer for a defined region takes place.

[0052] In a step 210, a provision of a planning layer for the region is carried out.

[0053] Finally, in a step 220, an extraction of alignment features 10a . . . 10n from the localization layer that is provided for an alignment with the planning layer is carried out, alignment features 10a . . . 10n being extracted in such a way from the localization layer that an inadmissible deformation of the localization layer may be recognized during the alignment of the planning layer with the localization layer.

[0054] The method according to the present invention may be advantageously implemented as a software that preferably runs in a cloud-based manner on a server device. An easy adaptability of the method is thus facilitated.

[0055] Those skilled in the art will modify and/or combine the features of the present invention in a suitable way, without departing from the core of the present invention.