Method for creating a digital map of the surroundings of a vehicle using a mapping system, and mapping system

Abstract

Method for creating a digital map of the surroundings of a vehicle by means of a mapping system (1), comprising the steps: - Providing a basic factor graph (5) comprising nodes (6) and factors (7) generated on the basis of individual trip data from a previous trip; - Providing further individual journey data relating to a further journey of a vehicle which is different from the previous journey; - extending the basic factor graph (5) by generating further nodes (8) and further factors (9) and inserting the further nodes (8) and the further factors (9) into the basic factor graph (5) on the basis of the further individual trip data provided, so that an extended factor graph (10) is generated by expanding the basic factor graph (5); - identifying nodes (8) of the further nodes (8) that have changed by more than a predetermined deviation compared to the nodes (6) of the provided basic factor graph (5); - Creating a digital map based on the extended factor graph (10) taking into account the identified nodes (8).

Description

[0001] One aspect of the invention relates to a method for creating a digital map of a vehicle's surroundings using a mapping system. Another aspect of the invention relates to a mapping system.

[0002] From WO 2020 / 016385 A1 a method and a system for determining a position of a vehicle is known.

[0003] Graph-based SLAM methods are known that generate a factor graph and then a digital map from the factor graph. The generated digital map may be inaccurate.

[0004] The invention is based on the object of creating a digital map in an improved manner, in particular of updating it.

[0005] The object is achieved by the subject matter of the independent claims. Advantageous developments of the invention are defined by the dependent claims, the following description, and the figures.

[0006] One aspect of the invention relates to a method for creating a digital map of a vehicle's surroundings using a mapping system. In particular, the method comprises the following steps: - In particular, providing a basic factor graph comprising nodes and factors generated on the basis of individual trip data from a previous trip; - In particular, providing further individual journey data relating to a further journey of a vehicle which is different from the previous journey; - In particular, extending the basic factor graph by generating further nodes and further factors and inserting the further nodes and the further factors into the basic factor graph on the basis of the further individual trip data provided, so that an extended factor graph is generated by expanding the basic factor graph; - In particular, identifying nodes of the further nodes that have changed by more than a specified deviation compared to the nodes of the provided basic factor graph; - In particular, creating a digital map based on the extended factor graph taking into account the identified nodes.

[0007] This method enables improved creation of a digital map. In particular, an improved digital map can be created. By expanding the basic factor graph and, in particular, only subsequently creating the digital map, changes, for example, can be better incorporated into the created digital map. This is particularly advantageous compared to methods that merely compare maps that have already been created. The method allows data from the basic factor graph to be used to generate the digital map, even if parts of the data are outdated. In particular, the method takes into account structural changes in the vehicle's surroundings, particularly in a parking garage.

[0008] In particular, the method is a computer-implemented method. For example, the base factor graph and the additional individual trip data are provided to an evaluation unit, which then carries out the subsequent steps of the method. In particular, the evaluation unit provides the generated digital map to other vehicles.

[0009] The method is preferably carried out for multiple journeys. The multiple journeys are, for example, carried out by the same vehicle or by different vehicles. In particular, fleet data from a vehicle fleet is therefore also used. In this case, both multiple journeys by a vehicle count as fleet data and individual journeys by different vehicles count as fleet data. The individual journey data can, for example, also be referred to as a maplet or a data packet. In particular, exactly one data packet is recorded per journey. However, it is also possible for a data packet to relate to several further journeys of a vehicle. A data packet includes, for example, self-motion data and mapping object observations. The self-motion data contains, for example, several items of position data for the vehicle, such as a spatial coordinate and, if applicable, an orientation of the vehicle in space. The map generated is, in particular, a global map.Preferably, the spatial coordinate refers to a global origin of the digital map.

[0010] In particular, the sequence of steps is not mandatory and can, in particular, be at least partially overlapping. The process can, for example, be implemented as a so-called offline SLAM (simultaneous localization and mapping) process. This means, in particular, that the map is only created after the end of the rest of the journey. In particular, the digital map is created for a parking garage and / or a parking facility with multiple parking zones; in particular, the digital map represents the parking garage and / or the parking facility.

[0011] The digital map can be understood as a data set that spatially relates objects within a given spatial area to each other and / or to a predefined reference point or coordinate origin. Optionally, the digital map can contain semantic information about the objects.

[0012] Optionally, the method can be referred to as graph-based SLAM. It is possible for there to be a fleet of vehicles that collects environmental data using vehicle sensors while driving through relevant parking spaces and generates so-called maplets from this. A maplet, for example, is a data container that contains the vehicle's own movement, which is in particular a fleet vehicle of the vehicle fleet, as well as mapping object observations from the vehicle's surroundings relative to an observer position of the vehicle. Mapping objects can be walls, pillars, posts, other vehicles, in particular parked vehicles, ground surfaces, open spaces, markings and / or parking spaces. These objects are formed, for example, in the vehicle from the sensor data from radar, camera, ultrasonic sensors and / or lidar and combined into a maplet by a vehicle component.Depending on the availability of a communication interface, a maplet is uploaded to a central evaluation unit, in particular a backend component (cloud), or temporarily cached, or transferred to another instance for uploading, for example via ad-hoc WLAN or C2X. In particular, the backend component receives maplets of individual fleet vehicles over time. The method according to the invention relates in particular to a computer-implemented method which generates a global map, in particular from the maplets. In particular, for each node in the extended factor graph, information is processed as to which point in time a node can be assigned. Based on this information, an analysis is carried out if necessary to identify which nodes have been removed over time, which nodes have been newly created, and which nodes have changed.For example, a node is identified as removed if a corresponding mapping object in the parking space has been removed. A node is identified as created if a corresponding mapping object has been inserted into the parking space, for example, if a new parking space is provided by painting a new floor marking.

[0013] In one embodiment, the provided basic factor graph has, as nodes, self-position nodes and landmark nodes of detected landmarks in the surroundings of the vehicle. The provided basic factor graph has, in particular, landmark factors and self-position factors as factors. The self-position nodes represent, for example, traveled positions of a vehicle, in particular of the vehicle. The landmark nodes represent, for example, traveled and, in particular, detected landmarks. The self-position factors preferably indicate a local relationship between self-position nodes. The landmark factors preferably indicate a local relationship between self-position nodes and landmark nodes. The local relationships are specified, for example, in meters.

[0014] It is possible that the further nodes are further landmark nodes and further self-position nodes and that the further factors are further landmark factors and further self-position factors.

[0015] In particular, the additional self-position nodes are generated based on the self-motion data from the individual trip data. If necessary, the additional landmark nodes are generated based on the mapping object observations. Landmarks are, for example, mapping objects in the vehicle's surroundings, which are detected, for example, by a detection system of the mapping system, in particular the vehicle. This allows information about the vehicle's own positions and the positions of the mapping objects to be recorded.

[0016] A landmark or a mapping object can be understood as features and / or patterns in an environment that can be identified and to which at least one piece of location information or position information can be assigned. These can be, for example, characteristic points or objects arranged at specific positions in the environment. A landmark can be assigned a landmark type, in particular based on one or more geometric and / or semantic properties of the landmark. For example, road markings, lane markings, other ground marking lines, building edges or corners, masts, posts, traffic signs, information signs or other signs, elements of vegetation, structures or parts thereof, parts of traffic control systems, and two-dimensional codes can each be defined as landmark types. A landmark can also be assigned to multiple landmark types.

[0017] In one embodiment, to identify the nodes, landmark nodes of the further nodes are compared with landmark nodes of the nodes of the provided basic factor graph.

[0018] In particular, it is not necessary to compare the self-position nodes and the other self-position nodes. This allows the process to be carried out faster and, above all, more efficiently.

[0019] For example, a distance is determined between each landmark node and the other landmark nodes. If the determined distance is less than or equal to a specified maximum value, the landmark nodes to be compared are identified as matching. In particular, the landmark nodes to be compared are identified as different if they exceed the specified maximum value. This enables unambiguous and rapid identification.

[0020] The maximum value is, for example, a fixed value, for example between 0 centimeters and 50 cm, in particular between 5 centimeters and 15 centimeters, in particular 10 centimeters. Optionally, the maximum value is specified depending on the sensor type of the detection system. The greater the accuracy of the sensor type used to determine the respective landmark node, the smaller the maximum value. If necessary, other parameters for determining the landmark nodes are taken into account, such as the distance of the landmark to the vehicle, as this may influence the accuracy of the determined distance between the landmark and / or the landmarks. This takes the accuracy of the sensor types used into account. This means that even qualitatively less than ideal self-motion data can be implicitly upgraded.

[0021] In particular, the maximum value is determined using a machine learning algorithm. For this purpose, the method is performed several times, for example, and the extended factor graph is analyzed after each execution. This allows the magnitude of typical deviations, for example those occurring on average, to be determined. The maximum value is preferably determined as a value that is, for example, 50%, in particular 20%, in particular 10% greater than the determined typical deviation.

[0022] In one embodiment, a change detection algorithm is implemented to identify the nodes. The change detection algorithm determines the distances of the landmark nodes to the other landmark nodes using the factors and the additional factors. One advantage of this embodiment is that misassociations of mapping objects can be reduced. In particular, the misassociations of mapping objects are reduced compared to methods in which the mapping object observations are assigned only after the digital map has been created.

[0023] In particular, the nodes and / or the other nodes are evaluated depending on the specified distances. For example, the nodes and the other nodes, in particular the landmark nodes and the other landmark nodes, are evaluated as inserted, deleted, or moved depending on the specified distances from each other.

[0024] If necessary, the change detection algorithm is performed for each additional landmark node. For example, the distances of all or only the surrounding landmark nodes of the base factor graph to the respective additional landmark node are compared with the specified maximum value. Matching landmark nodes are optionally identified as already described. For example, it is possible that no matching landmark node of the base factor graph is found for the respective additional landmark node. In this case, the landmark node in question is evaluated as inserted, for example. However, it is also possible that in this case a further comparison with an additional maximum value that is greater than the maximum value is performed. For example, in the described case, the additional landmark node is only evaluated as inserted if all determined distances are greater than the specified additional maximum value.If one, in particular exactly one, of the specified distances lies between the maximum value and the further maximum value, the corresponding landmark node of the base factor graph, and in particular also the respective further landmark node, is considered shifted. This results in the advantage that all landmark nodes and thus all recorded mapping objects are incorporated into the factor graph, in particular because the change detection algorithm is executed for each additional landmark node if necessary. This allows an improved digital map to be created or updated.

[0025] In one embodiment, the change detection algorithm can be implemented as follows: For the SLAM method itself, the insertion or deletion of landmark nodes, which can also be referred to as observation nodes, is particularly unproblematic over time. Even if observations of a mapping object are no longer present in future maplets from a certain point in time, this may not disrupt the SLAM method. In particular, it is also unproblematic if additional observations occur from a certain point in time. This can occur, for example, even without structural changes to the parking space, simply by obscuring mapping objects, for example, by dynamic or static vehicles.The modification of a mapping object is particularly problematic: If, for example, a post in a parking space is moved a short distance, the SLAM method might produce a suboptimal global map without additional analysis, as it would attempt to determine the best compromise between the eigenposes and observations. The detection of modified mapping objects can preferably be implemented by systematically searching for similar mapping objects in the vicinity of a mapping object within the factor graph using a neighborhood search. A temporal analysis of found modification candidates can be used to identify moved mapping objects. The information generated in this way is taken into account, in particular, within the data association during the execution of the SLAM method to ensure that modified observation nodes are not associated with one another.

[0026] In one embodiment, identified nodes are marked as inserted, deleted, or moved.

[0027] This allows for a traceable history of the factor graph. In particular, the landmark nodes and the additional landmark nodes are marked according to their evaluation. For example, the extended factor graph becomes the base factor graph when the procedure is repeated, especially for additional individual trip data. This makes it possible to determine, based on the landmark nodes and the additional landmark nodes, whether they have been inserted, deleted, moved, or, in particular, unchanged, after each execution of the procedure.

[0028] Preferably, landmark nodes marked as deleted are not actually deleted from the factor graph. This allows the individual trip data from the previous trip to be used to improve the digital map. For example, mapping objects corresponding to landmark nodes marked as deleted are not displayed in the digital map when the digital map is created or updated.

[0029] In one embodiment, each node and each subsequent node has a timestamp. This allows for improved tracking of the history of the factor graph.

[0030] For example, nodes marked as changed, which in particular represent the same mapping object, are displayed as the landmark node with the more recent timestamp in the created or updated digital map.

[0031] In one embodiment, a route network is generated based on the additional individual trip data and inserted into the generated or updated map. This makes the route network available to subsequent vehicles.

[0032] In particular, a trajectory of the vehicle is recorded during the subsequent journey. Preferably, the recorded trajectory is inserted into the digital map, in particular as a road network. The inserted road network is used, for example, to determine the trajectories of further journeys, in particular at least partially autonomous journeys.

[0033] In one embodiment, a digital predecessor map is provided based on the provided base factor graph. The digital predecessor map is updated based on the extended factor graph, taking into account the identified nodes.

[0034] For example, an updated global map is created each time additional individual trip data is made available. This has the advantage that the global digital map is based on the most recent available individual trip data.

[0035] It is also possible that the global digital map is only created or updated after several additional individual trip data items have been provided and the base factor graph has been expanded to include several additional nodes and factors based on the several additional individual trip data items. For example, a certain number of additional individual trip data items is specified as the starting point for the digital map update. It is also possible that the digital map is updated or created when a predefined number of marked nodes, particularly nodes marked as inserted, moved, or deleted, is exceeded. This can, for example, prevent unnecessary map updates.

[0036] In one embodiment, the created or updated map is provided to a server device external to the vehicle. For example, other vehicles receive the created or updated map from the server device.

[0037] It is also possible for the created or updated digital map to be verified based on predefined criteria before it is made available to the server device or before it can be received by other vehicles. These criteria relate, for example, to heuristic completeness rules. If, for example, it is determined during verification that at least one mapping object has a course, position, or orientation that is physically impossible or illogical, the created or updated map will not be made available to the server device or sent to the other vehicles, especially if the verification is performed, for example, in the server device.

[0038] In one embodiment, the updated digital map is compared with the previous digital map. Based on this comparison, the digital map is modified. For example, mapping objects that have a course, position, or orientation that is physically impossible or illogical can be adjusted. This removes, in particular, inconsistencies in the generated or updated map, and the map is made available to the server device and / or the other vehicles.

[0039] In one embodiment, particularly after the additional individual trip data have been provided, a check is carried out to determine whether individual trip data from the previous trip has already been provided, particularly for the relevant surroundings of the vehicle. If this is not the case, the base factor graph is generated, and the method is terminated. If corresponding individual trip data has already been provided, the further method steps are carried out.

[0040] Another aspect of the invention relates to an electronic mapping system for creating a digital map. The mapping system comprises a recording system, a storage unit, and an evaluation unit, wherein: - the evaluation unit is configured to obtain a basic factor graph comprising nodes and factors generated on the basis of individual trip data from a previous trip; - the recording system is designed to record further individual journey data relating to a further journey of a vehicle which is different from the previous journey and to make this data available to the evaluation unit; - the evaluation unit is configured to extend the basic factor graph by generating further nodes and further factors and inserting the further nodes and the further factors into the basic factor graph on the basis of the further individual trip data provided, so that an extended factor graph is generated by extending the basic factor graph; - the evaluation unit is configured to identify nodes of the further nodes that have changed by more than a predetermined deviation compared to the nodes of the provided basic factor graph; - the evaluation unit is designed to create a digital map based on the extended factor graph, taking into account the identified nodes.

[0041] For example, the vehicle has the detection system. Optionally, the detection system has a first detection unit for detecting mapping objects and a second detection unit for detecting the vehicle's own movement.

[0042] The first detection unit comprises, for example, one or more sensors for detecting the surroundings of the vehicle. The first detection unit can be understood, for example, as a sensor system capable of generating sensor data or sensor signals that map, display, or reproduce the surroundings of the vehicle. For example, the first detection unit can comprise cameras, radar systems, lidar systems, or ultrasonic sensor systems.

[0043] The second detection unit has, in particular, one or more sensors for detecting the position and / or orientation of the vehicle. A GNSS system, for example, is used to detect the position of the vehicle. To detect the orientation of the vehicle, the pitch angle, yaw angle, and roll angle of the vehicle are detected, for example, using an attitude sensor. Alternatively or additionally, the second detection unit detects the position and / or orientation of the vehicle, for example, using odometry.

[0044] The evaluation unit, in particular, comprises one or more computing units. For example, an external server unit comprises the evaluation unit. Alternatively or additionally, it is possible for the vehicle to comprise the evaluation unit. In particular, the external server unit comprises the storage unit.

[0045] In one embodiment, the ego-motion data in the fleet vehicles is based on a high-precision global navigation satellite system (GNSS system) in the vehicles when driving through areas without GNSS reception. This can be referred to as "dead reckoning" via vehicle odometry. If necessary, the ego-motion data thus inherently share the same reference system and can be directly transferred to nodes of the factor graph.

[0046] In one embodiment, an additional step for adapting the self-motion data takes place, especially if the vehicle does not have the option of a GNSS system. For example, a maplet can be brought into a global reference system using a radar- or camera-based odometry system, such as radar odometry, visual odometry, radar localization, and / or visual localization. Alternatively or additionally, the maplet can be consistently aligned accordingly through backend adaptation, for example, using ICP (iterative closest point) or NDT (normal distribution transform).

[0047] A further aspect of the invention relates to a mapping system. The mapping system is configured to implement the method according to the invention or an advantageous embodiment thereof. In particular, the mapping system executes the method.

[0048] A further aspect of the invention relates to a vehicle having a mapping system according to the invention.

[0049] Another aspect of the invention relates to a method for driving a vehicle. The vehicle is driven based on a digital map created using the method according to the invention.

[0050] Features and advantages of the method according to the invention for creating a digital map or an embodiment thereof can further develop the further aspects of the invention and vice versa.

[0051] The invention also includes combinations of the features of the described embodiments.

[0052] Exemplary embodiments of the invention are described below. Shown are: Fig. 1 an embodiment of a mapping system according to the invention; and Fig. 2 an embodiment of a factor graph.

[0053] The exemplary embodiments explained below are preferred exemplary embodiments of the invention. In the exemplary embodiments, the described components each represent individual, independently considered features of the invention, which also further develop the invention independently of one another and are thus also to be considered as components of the invention, either individually or in a combination other than that shown. Furthermore, the described exemplary embodiments can also be supplemented by further features of the invention already described.

[0054] In the figures, functionally identical elements are provided with the same reference numerals.

[0055] In Fig. Figure 1 schematically illustrates an embodiment of a mapping system 1 for creating a digital map. The mapping system 1 comprises a recording system 2, an evaluation unit 3, and optionally a storage unit 4. In particular, the mapping system 1 is configured to perform the following steps: - Providing a basic factor graph 5 ( Fig. 2), having node 6 ( Fig. 2) and factors 7 ( Fig. 2) which were generated based on individual trip data from a previous trip; - Providing further individual journey data relating to a further journey of a vehicle which is different from the previous journey; - Extending the base factor graph 5 by creating additional nodes 8 ( Fig. 2) and other factors 9 ( Fig. 2) and inserting the further nodes 8 and the further factors 9 into the basic factor graph 5 on the basis of the further individual trip data provided, so that by extending the basic factor graph 5 an extended factor graph 10 ( Fig. 2) is generated; - Identifying nodes of the further nodes 8 that have changed by more than a predetermined deviation compared to the nodes 6 of the provided basic factor graph 5; - Creating a digital map based on the extended factor graph 10 taking into account the identified nodes.

[0056] For example, a first vehicle performs a first trip, for example, in a parking garage. For example, the first vehicle has a recording system 2 for recording the individual trip data. The individual trip data includes, in particular, the vehicle's own motion data and mapping objects, which can also be referred to as landmarks. The vehicle's own motion data includes, in particular, the vehicle's positions and orientations. If necessary, the recording system 2 repeats the recording between 0.5 milliseconds and 1 second, in particular every millisecond.

[0057] For example, the individual trip data is provided to the evaluation unit 3. In particular, the evaluation unit 3 checks whether individual trip data already exists for the location in question, in particular the parking garage. If this is not the case, a basic factor graph 5 is generated from the provided individual trip data. For example, in this case the digital map is created based on the basic factor graph 5. If individual trip data already exists, the provided individual trip data is in particular further individual trip data and the basic factor graph 5 is expanded accordingly to form the factor graph 10. If a digital map has already been created, it is preferably updated taking into account the expanded factor graph 10. If no digital map has yet been created, it is created taking into account the expanded factor graph 10.

[0058] Fig. 2 shows a schematic embodiment of an extended factor graph 10, in particular after implementation of the method for creating the digital map. The extended factor graph 10 comprises, for example, the base factor graph 5. The base factor graph 5 comprises, in particular, the nodes 6 and the factors 7. The nodes 6 contain, for example, self-position nodes 11 and landmark nodes 12. The factors 7 contain, for example, self-position factors 13, landmark factors 14, and in particular, map adaptation factors 15. In particular, the self-position factors 13 connect the self-position nodes 11 to one another. In particular, the landmark factors 14 connect self-position nodes 11 to landmark nodes 12. The map adaptation factors 15 are optionally connected to the self-position nodes 11, which represent a position of the vehicle that was acquired using a GNSS system of the acquisition system 2.

[0059] The extended factor graph 10 also includes the additional nodes 8 and the additional factors 9. The additional nodes 8 include, for example, additional self-position nodes 16 and additional landmark nodes 17, analogous to the basic factor graph 5. The additional factors 9 include, in particular, additional self-position factors 18, additional landmark factors 19, and optionally additional map adjustment factors 20, analogous to the basic factor graph 5.

[0060] In the Fig.2, the node 12a and the further node 17a can be identified if necessary. The landmark node 17 is, for example, more than the specified distance away from a neighboring node, such as the node 12b. Therefore, it is particularly recognized that the node 12b and the further node 17a do not match. Optionally, the distance between the node 12b and the further node 17a exceeds a further maximum distance, so that the further node 17a is identified as inserted and, in particular, not as shifted. If necessary, a check is carried out for all landmark nodes 12 of the basic factor graph 5 to determine whether a matching and / or a shifted further landmark node 17 can be found. For example, no matching or shifted further landmark node 17 can be found for the node 12a of the basic factor graph 5.Therefore, node 12a is also identified if necessary. Specifically, node 12a is marked as deleted, and the other node 17a is marked as inserted.

[0061] A global map is preferably generated based on the extended factor graph 10. For example, mapping objects represented by nodes marked as deleted, such as node 12a, are not displayed in the digital map. If, upon repeating the method, the node 12a detected as deleted is identified as visible again, it is possible to reuse information about the node 12a of the original base factor graph 5.

[0062] Mapping objects represented by nodes marked as inserted are specifically inserted into the digital map. Optionally, a digital predecessor map based on the base factor graph 5 is provided. If necessary, the predecessor map is updated based on the extended factor graph 10.

[0063] In one embodiment, the following steps can be carried out, in particular as soon as individual trip data, which can also be referred to as a maplet, are provided, in particular to an evaluation unit: - Check whether there is already a maplet for an environment, in particular a parking area, for example, which can be determined from an available NDS map using GNSS coordinates. - If this is not the case, a base factor graph is generated from the self-motion data and mapping object observations. - If this is the case, new nodes are inserted into the factor graph existing for the location from the self-motion data and mapping object observations. - The change detection algorithm is implemented based on the factor graph. Clustering / neighborhood search of observation nodes in close proximity is used to identify those that have changed over time, particularly those that have been added, deleted, or moved. - The graph-based SLAM method is performed on the basis of the annotated factor graph. The temporal markers are taken into account during the data association step. - The result is saved as a global map. Mapping objects marked as deleted are removed from the map. The entire set of self-motion data modified during the SLAM process is used as the basis for generating the path network, such as waypoint clustering, links, lanes, and trajectories. - The new global map is compared with the previous version, if necessary, and any deviations are analyzed. Heuristic completeness rules are applied. If both deviations and completeness rules allow it, the map is approved by the backend mapping algorithm. Alternatively or additionally, a human user can perform a visual review and then approve the map for download to vehicles. Optionally, incomplete or not yet sufficiently converged global maps can also be approved to gain time with regard to the availability of the customer function being used, which may be possible, for example, depending on the security level of the customer function. List of reference symbols 1 mapping system 2 Recording system 3 Evaluation unit 4 storage unit 5 Basic factor graph 6 knots 7 factors 8 more nodes 9 other factors 10 factor graph 11 Self-position nodes 12 landmark nodes 12a knots 12b knots 13 Own Position Factors 14 landmark factors 15 map adjustment factors 16 additional self-position nodes 17 additional landmark nodes 17a knots 18 additional own position factors 19 additional landmark factors 20 map adjustment factors QUOTES CONTAINED IN THE DESCRIPTION

[0000] This list of documents submitted by the applicant was generated automatically and is included solely for the convenience of the reader. This list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions. Cited patent literature

[0000] WO 2020 / 016385 A1

[0002]

Claims

[1] Method for creating a digital map of the surroundings of a vehicle by means of a mapping system (1), comprising the steps: - Providing a basic factor graph (5) comprising nodes (6) and factors (7) generated on the basis of individual trip data from a previous trip; - Providing further individual journey data relating to a further journey of a vehicle which is different from the previous journey; - extending the basic factor graph (5) by generating further nodes (8) and further factors (9) and inserting the further nodes (8) and the further factors (9) into the basic factor graph (5) on the basis of the further individual trip data provided, so that an extended factor graph (10) is generated by expanding the basic factor graph (5); - identifying nodes (8) of the further nodes (8) that have changed by more than a predetermined deviation compared to the nodes (6) of the provided basic factor graph (5); - Creating a digital map based on the extended factor graph (10) taking into account the identified nodes (8). [2] Method according to claim 1, wherein the provided basic factor graph (5) has as nodes (6) own position nodes (11) and landmark nodes (12) of detected landmarks in the surroundings of the vehicle and the provided basic factor graph (5) has as factors (7) landmark factors (14) and own position factors (13), wherein own position nodes (11) represent traveled positions of the vehicle and landmark nodes (12) represent detected landmarks, wherein own position factors (13) indicate a local relationship between own position nodes (11) and landmark factors (14) indicate a local relationship between own position nodes (11) and landmark nodes (12). [3] Method according to claim 2, wherein, for identifying the nodes (8), landmark nodes (17) of the further nodes (8) are compared with landmark nodes (12) of the nodes (6) of the provided basic factor graph (5). [4] Method according to one of the preceding claims, wherein a change detection algorithm is carried out to identify the nodes (8), which determines distances of the nodes (6) to the further nodes (8) by means of the factors (7) and the further factors (9). [5] Method according to one of the preceding claims, wherein identified nodes (8) are marked as inserted, deleted or moved. [6] Method according to one of the preceding claims, wherein each node (6) and each further node (8) has a time stamp. [7] Method according to one of the preceding claims, wherein a route network is generated depending on the further individual trip data and is inserted into the generated map. [8] Method according to one of the preceding claims, wherein a digital predecessor map is provided based on the provided basic factor graph (5) and the digital predecessor map is updated based on the extended factor graph (10) taking into account the identified nodes (6). [9] Method according to claim 8, wherein the updated digital map is compared with the previous digital map and the digital map is changed on the basis of this comparison. [10] Mapping system (1) for creating a digital map, comprising a recording system (2), a storage unit (4) and an evaluation unit (3), wherein: - the evaluation unit (3) is configured to obtain a basic factor graph (5) comprising nodes (6) and factors (7) which were generated on the basis of individual trip data from a previous trip; - the recording system (2) is designed to record further individual journey data relating to a further journey of a vehicle which is different from the previous journey and to provide it to the evaluation unit (3); - the evaluation unit (3) is configured to expand the basic factor graph (5) by generating further nodes (8) and further factors (9) and inserting the further nodes (8) and the further factors (9) into the basic factor graph (5) on the basis of the further individual trip data provided, so that an expanded factor graph (10) is generated by expanding the basic factor graph (5); - the evaluation unit (3) is configured to identify nodes (8) of the further nodes (8) which have changed by more than a predetermined deviation compared to the nodes (6) of the provided basic factor graph (5); - the evaluation unit (3) is configured to create a digital map based on the extended factor graph (10) taking into account the identified nodes (8).

Images