Method and system for providing map data for operation of an automated and / or assistive system

Abstract

The invention relates to a method for creating map data for operating an automated and / or assistive system of a vehicle or robot, wherein a map (1) for a traffic route section of a traffic route is automatically compared with a reference map (3) for the traffic route section, wherein the map (1) is only automatically used to create the map data if the map (1) does not deviate from the reference map (3) by a predetermined amount with regard to at least one predetermined quality measure.

Description

[0001] The invention relates to a method for providing map data for operating an automated and / or assistive system of a vehicle or robot. Furthermore, the invention relates to a system for providing map data for operating an automated and / or assistive system of a vehicle or robot. State of the art

[0002] Maps used for self-localization of vehicles, so-called high-definition maps, form a particularly important basis for autonomous driving. Without such maps, a vehicle's navigation in traffic is significantly more difficult, as the basis for self-localization is missing. Above all, the maps must be highly accurate and up-to-date with regard to the position of road features, such as lane markings, street signs, etc.

[0003] Map creation is typically fully automated via an algorithmic chain based on data from a fleet of vehicles already in use on the road, i.e., data obtained through crowdsourcing from a fleet of vehicles. This data is subject to significant noise due to the vehicle sensors used to generate it. As a result, a map accuracy check is mandatory after the map has been created. This accuracy check, or verification of a map, is currently a laborious manual process, resulting in considerable financial and personnel expenditure.

[0004] US 2022 / 0 341 750 A1 discloses monitoring the condition of a high-definition map to determine whether or not there is an inaccuracy in the map. If an inaccuracy is identified in the map that indicates a deterioration in the map's condition, updated data from one or more vehicles regarding an area of ​​the map containing the error can be crowdsourced, and the updated data can be used to update, verify, and validate the map. Disclosure of the invention

[0005] The invention relates to a method for creating map data for operating an automated and / or assistive system of a vehicle or robot, with which a first automated accuracy check of a freshly generated map can be carried out by automatically or electronically comparing the map for a traffic route section with a reference map for the traffic route section, and only using the map automatically to create the map data if the map does not deviate from the reference map by more than a predetermined amount with regard to at least one predetermined quality measure. This automation thus makes it possible to greatly reduce the manual effort required to create map data.The accuracy check of the map can, for example, be performed immediately after the map has been created to ensure the initial quality of the map data, which ultimately improves the map data quality. The present invention therefore serves to initially ensure the quality of the map data and not to revise map data already in use.

[0006] The map contains spatially resolved or local map data, which can be provided as a point cloud. Maps or local map datasets assigned to a specific traffic area can be combined using a scan matching method (also known as a map alignment method) to improve map accuracy or map timeliness. Traditionally, this involves determining an approximate transformation between two local map datasets, which can originate from any sensor, such as lane markings detected by a vehicle camera or radar echoes. This approximate transformation is usually a central first step in the processing of multiple, partially overlapping local map datasets. In the context of the present application, such a local map dataset is referred to as a map, which is compared with the corresponding reference map.

[0007] The invention can be used to analyze map data generated by at least one vehicle sensor. The vehicle sensor can be used, for example, to detect the vehicle's surroundings, with the map data being provided in the form of sensor signals, which can be provided, for example, as digital images, such as video images, radar data, LiDAR data, ultrasound data, motion data, or thermal images.

[0008] Road features can be represented on maps using polylines or point clouds. The accuracy of a local map can therefore be verified by comparing these structures with a local reference map parameterized in the same way. This map verification can be carried out, for example, using a threshold value of at least one quality measure that can be specified in advance by the user. One conceivable approach is to use the mean square error with a specific threshold value. This threshold value would then establish an upper bound on the tolerable local deviation of the map from the reference map. Using heuristic limits for quality measures or error measures specified by the user, erroneous sub-areas of the map can be identified.

[0009] The map data created using the method according to the invention can be made available to an automated vehicle system (AD system) and / or an assisted vehicle system (DA system), thereby providing the respective system with more precise inputs, enabling the respective system to provide more precise outputs, thus providing improved functionality. This is particularly advantageous with regard to autonomous driving of vehicles.

[0010] According to an advantageous embodiment, an indication signal is generated and output if the map deviates from the reference map by more than the specified amount with regard to the specified quality measure. If the map deviates from a corresponding feature of the reference map by more than the specified amount or a specified limit value with regard to at least one feature contained in the map, the user, in particular a developer who is processing the generated map, can be made aware of this at least one deviation area of ​​the map by means of the perceptible indication signal. A manual inspection of the map can then be carried out, limited only to this deviation area, which is possible with significantly less personnel effort than the conventional complete manual inspection of a map. This embodiment is therefore particularly useful for the map development process.

[0011] According to a further advantageous embodiment, the map is created using data obtained through crowdsourcing from a vehicle fleet. The map can be produced based on the crowdsourcing data using an algorithmic chain.

[0012] According to a further advantageous embodiment, the reference map is created using sensor data acquired during a drive along the road section using high-precision vehicle sensors. The reference map can thus be obtained based on measurements from at least one vehicle with high-precision environmental sensors.

[0013] According to a further advantageous embodiment, the map is only used automatically to create the map data if the quality of the map meets at least one additional evaluation criterion. Whether the additional evaluation criterion is met or not can be checked, for example, by applying a plausibility check within a data layer of the map. For example, as part of a plausibility check, it can be checked whether a recorded number of lanes and / or a recorded width of a lane (with a specified tolerance) and / or a recorded distance between line markings (with a specified tolerance) and / or a recorded relationship between different landmarks (for example, that traffic signs must not be located on a lane marking) is plausible.The additional evaluation criterion can, for example, be given as an error function in the form of a linear combination of various additional evaluation criteria, for example the various additional evaluation criteria mentioned above.

[0014] Alternatively or additionally, the additional evaluation criterion can be derived from information outside the map's data layer. For example, an additional evaluation criterion can be specified that a road center line correctly derived from lane markings must never extend outside the road. This can then be verified as part of a plausibility check. A "behavior map" can provide an indication of an invalid lane marking if, according to the "behavior map," this lane marking is always ignored by a driver. As an additional evaluation criterion, the map can therefore be checked for the presence of an invalid lane marking.

[0015] Another possibility for implementing an additional evaluation criterion is the use of measurements and / or values ​​that arise during the creation of the map data. For example, the extent of an error in a global pose optimization of local scan matches can be used as an additional evaluation criterion. Alternatively or additionally, the ratio of the number of scan matches with low errors to the total number of scan matches used can be used as an additional evaluation criterion. Alternatively or additionally, an inconsistency measure of map sections that overlap after a scan matching ("map is aligned, but inconsistent at overlapping seams") can be used as an additional evaluation criterion. Alternatively or additionally, the number of required corrections during an optimization process can be used as an additional evaluation criterion.

[0016] A further subject of the invention is a system for providing map data for operating an automated and / or assistive system of a vehicle or robot, wherein the system is configured to compare a map of a traffic route section of a traffic route with a reference map of the traffic route section and to use the map to create the map data only if the map does not deviate from the reference map by a predetermined amount with respect to at least one predetermined quality measure. Advantageously, the system is configured to carry out the method according to one of the above-mentioned embodiments or a combination of at least two of these embodiments.

[0017] The system offers the advantages mentioned above with regard to the method. Advantageous embodiments of the method may correspond to advantageous embodiments of the system, even if this is not explicitly mentioned below.

[0018] In the following, the invention is explained by way of example with reference to the attached figures using preferred embodiments, wherein the features explained below can represent an advantageous and / or further developing aspect of the invention both individually and in different combinations of at least two of these features with one another. Short description of the characters

[0019] It shows: Fig. 1 schematic representations of an embodiment of a map with defective areas and an embodiment of a reference map; and Fig. 2 a flowchart of an embodiment of a method according to the invention. Detailed description of the characters

[0020] In the figures, identical or functionally identical components are provided with the same reference numerals. A repeated description of such components may be omitted in detail to avoid unnecessary repetition.

[0021] Fig. 1 shows schematic representations of an embodiment of a local map 1 with five faulty areas 2 and an embodiment of a corresponding local reference map 2. The faulty areas 2 have been determined by comparing the map 1 with the reference map 3.

[0022] Fig. 2 shows a flowchart of an embodiment of a method according to the invention for creating map data for operation of an automated and / or assistive system of a vehicle or robot.

[0023] In process step 10, data on a traffic route section is obtained and collected by crowdsourcing using a vehicle fleet.

[0024] In method step 20, a local map of the traffic route section is created using the data obtained in method step 10.

[0025] In process step 30, the map created in process step 20 is automatically compared with a reference map of the traffic route section. For this purpose, the reference map was created in advance using sensor data obtained during a drive along the traffic route section using high-precision vehicle sensors.

[0026] The map is only used automatically to create the map data if it is determined in method step 30 that the map does not deviate from the reference map by more than a predetermined amount with regard to at least one predetermined quality measure.

[0027] Furthermore, in method step 30, an indication signal is generated and output if the map deviates from the reference map by more than the specified amount with respect to the specified quality measure. A faulty or inaccurate area of ​​the map can then be manually checked in method step 40 before the map is used to create the map data.

[0028] In method step 30, it can further be checked whether a quality of the map meets at least one additional evaluation criterion, wherein the map is only used automatically to create the map data if the quality of the map meets the at least one additional evaluation criterion.

Claims

[1] Method for creating map data for operation of an automated and / or assistive system of a vehicle or robot, characterized by that a map (1) for a traffic route section of a traffic route is automatically compared with a reference map (3) for the traffic route section, wherein the map (1) is only automatically used to create the map data if the map (1) does not deviate from the reference map (3) by a predetermined amount with regard to at least one predetermined quality measure. [2] Method according to claim 1, characterized by that an indication signal is generated and output when the card (1) deviates from the reference card (3) by more than the specified amount with regard to the specified quality measure. [3] Method according to claim 1 or 2, characterized by that the map (1) is created using data obtained through crowdsourcing using a vehicle fleet. [4] Method according to one of the preceding claims, characterized by that the reference map (3) is created using sensor data obtained during a journey along the traffic route section by means of a high-precision vehicle sensor system. [5] Method according to one of the preceding claims, characterized by that the map (1) is only used automatically to create the map data if the quality of the map (1) meets at least one additional evaluation criterion. [6] System for providing map data for operation of an automated and / or assistive system of a vehicle or robot, characterized bythat the system is set up to compare a map (1) for a traffic route section of a traffic route with a reference map (3) for the traffic route section and to use the map (1) to create the map data only if the map (1) does not deviate from the reference map (3) by a predetermined amount with regard to at least one predetermined quality measure.

Images