Map generation method and device, equipment and storage medium

Abstract

The invention relates to the technical field of vehicle mapping, and discloses a map generation method and device, equipment and a storage medium, and the method comprises the steps: obtaining a target sub-map library; the target sub-graph library comprises at least two sub-graph sets; each sub-graph set comprises each sub-graph; each sub-graph is used for representing the feature of a designated area where each track point in the corresponding driving track is located; in the target sub-image library, performing feature fusion on the sub-images with the aligned features between the at least two sub-image sets to obtain an updated target sub-image library; and splicing the position-associated sub-graphs in the updated target sub-graph library to obtain a target map. According to the scheme, the elements in the map are continuously complemented through fusion of the sub-maps with the aligned features, so that the accuracy of the map generated through splicing is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle mapping, and in particular to a map generation method, device, equipment and storage medium. Background Art

[0002] A crowdsourced map is a map constructed from map data collected in a crowdsourcing mode. By distributing the data collection work to each ordinary vehicle, each ordinary vehicle collects the map data, avoiding the process of special personnel on-site collection and survey. In the existing process of constructing a crowdsourced map, generally, the result of a single trip of a trajectory is used for mapping. However, there are often missing or incorrect elements in single-time mapping, and there is a lack of map information in the unvisited areas; at the same time, since the real environment changes over time, there is a deviation between the map and the actual environment.

[0003] Therefore, there is an urgent need for a technical solution that can accurately generate a map according to the map data collected by a vehicle. Summary of the Invention

[0004] In view of this, the present application provides a map generation method, device, equipment and storage medium, which improves the accuracy of generating a map. The technical solution is as follows.

[0005] In a first aspect, a map generation method is provided, and the method includes:

[0006] Obtain a target sub-map library; the target sub-map library includes at least two sub-map sets; each sub-map set includes each sub-map; each of the sub-maps is used to characterize the features of the specified area where each trajectory point in the corresponding driving trajectory is located;

[0007] In the target sub-map library, perform feature fusion on the sub-maps with aligned features between the at least two sub-map sets to obtain an updated target sub-map library;

[0008] Stitch the sub-maps with associated positions in the updated target sub-map library to obtain a target map.

[0009] In a possible implementation manner, the method further includes:

[0010] Obtain at least one driving trajectory;

[0011] Sample the at least one driving trajectory at intervals of a first distance to obtain each trajectory point in the at least one driving trajectory;

[0012] Generate a sub-map set corresponding to the at least one driving trajectory according to the features in the specified area centered on each of the trajectory points;

[0013] Update the target sub-map library according to the sub-map set corresponding to the at least one driving trajectory.

[0014] In a possible implementation, the distance between the center points of the respective subgraphs associated with the position is less than a second distance; and the second distance is greater than a first distance.

[0015] In a possible implementation, in the target subgraph library, performing feature fusion on the subgraphs with feature alignment between the at least two subgraph sets to obtain an updated target subgraph library, including:

[0016] Determining whether the subgraphs with position association between the at least two subgraph sets are feature-aligned;

[0017] If they are feature-aligned, performing feature fusion on the feature-aligned subgraphs and saving the fused subgraphs in the set with an earlier time of the driving trajectory between the at least two subgraph sets.

[0018] In a possible implementation, the method further includes:

[0019] If the subgraphs with position association between the at least two subgraph sets are not feature-aligned, determining whether the subgraphs with position association between the at least two subgraph sets are messy subgraphs;

[0020] If the subgraphs with position association between the at least two subgraph sets are all messy subgraphs, retaining the subgraphs with position association between the at least two subgraph sets.

[0021] In a possible implementation, the method further includes:

[0022] If there are non-messy subgraphs among the subgraphs with position association between the at least two subgraph sets, using the non-messy subgraphs as candidate subgraphs;

[0023] When the number of candidate subgraphs with a first association relationship is greater than a quantity threshold, performing pairwise feature alignment on the candidate subgraphs with a first association relationship;

[0024] If the candidate subgraphs with a first association relationship are successfully aligned, fusing the successfully aligned candidate subgraphs to obtain a target subgraph and deleting the candidate subgraphs with a first association relationship.

[0025] In a possible implementation, determining whether the subgraphs with position association between the at least two subgraph sets are feature-aligned includes:

[0026] Selecting a first subgraph and a second subgraph from the at least two subgraph sets respectively; the first subgraph and the second subgraph are located in different subgraph sets and the first subgraph and the second subgraph are position-associated;

[0027] Determine the transformation matrix between the first sub - graph and the second sub - graph, and convert the first sub - graph into a third sub - graph according to the transformation matrix;

[0028] Sample the lane lines in the second sub - graph and the third sub - graph respectively to obtain first sampling points;

[0029] Perform clustering processing on the first sampling points to obtain each first point cluster;

[0030] Determine whether the first sub - graph and the second sub - graph are aligned according to the variances of each first point cluster.

[0031] In a possible implementation manner, the determining whether the first sub - graph and the second sub - graph are aligned according to the variances of each first point cluster includes:

[0032] Obtain a first quantity that is greater than a variance threshold among the variances of each first point cluster;

[0033] If the first quantity is less than or equal to a quantity threshold, determine that the first sub - graph and the second sub - graph are aligned.

[0034] In a possible implementation manner, the determining whether the sub - graph for judging the position association between at least two sub - graph sets is a messy sub - graph includes:

[0035] For a fourth sub - graph that is not feature - aligned, sample the lane lines in the fourth sub - graph to obtain second sampling points;

[0036] Perform clustering processing on the second sampling points to obtain each second point cluster;

[0037] Determine whether the fourth sub - graph is a messy sub - graph according to the variances of each second point cluster.

[0038] In a possible implementation manner, the determining whether the fourth sub - graph is a messy sub - graph according to the variances of each second point cluster includes:

[0039] Obtain a second quantity that is greater than a variance threshold among the variances of each second point cluster;

[0040] If the second quantity is less than or equal to a quantity threshold, determine whether the fourth sub - graph is a messy sub - graph.

[0041] In a possible implementation manner, the performing feature fusion on the sub - graphs with feature alignment between at least two sub - graph sets includes:

[0042] Determine the fusion weights corresponding to each of the feature - aligned sub - graphs according to the fusion times corresponding to each of the feature - aligned sub - graphs;

[0043] Fuse the subgraphs with feature alignment according to the respective fusion weights corresponding to the subgraphs with feature alignment.

[0044] In a second aspect, a map generation device is provided. The device includes:

[0045] A map gallery acquisition module, configured to acquire a target sub-map gallery; the target sub-map gallery includes at least two sub-map sets; each sub-map set includes respective sub-maps; the respective sub-maps are respectively used to characterize the features of the designated areas where the respective trajectory points in the corresponding driving trajectory are located;

[0046] A sub-map update module, configured to fuse the sub-maps with feature alignment between the at least two sub-map sets in the target sub-map gallery to obtain an updated target sub-map gallery;

[0047] A map fusion module, configured to splice the respective sub-maps with associated positions in the updated target sub-map gallery to obtain a target map.

[0048] In a third aspect, a computer device is provided. The computer device includes a memory and a processor, which are communicatively connected to each other. The memory stores computer instructions, and the processor executes the computer instructions to execute the above map generation method.

[0049] In a fourth aspect, a computer-readable storage medium is provided. Computer instructions are stored on the computer-readable storage medium, and the computer instructions are used to cause a computer to execute the above map generation method.

[0050] In a fifth aspect, a computer program product or a computer program is provided, including computer instructions, and the computer instructions are used to cause a computer to execute the above map generation method.

[0051] The technical solution provided by this application may include the following beneficial effects:

[0052] When generating a map based on crowdsourcing trajectory fusion, first obtain a target sub-map library; the target sub-map library includes at least two sub-map sets; each sub-map set includes various sub-maps; and each sub-map is respectively used to represent the characteristics of the specified area where each trajectory point in the corresponding driving trajectory is located; at this time, in the target sub-map library, determine the corresponding sub-maps between different sub-map sets, and perform feature fusion on the sub-maps with aligned features between different sub-map sets to obtain an updated target sub-map library; splice the sub-maps with associated positions in the updated target sub-map library to obtain a target map. In the above solution, first generate sub-maps of the characteristics of the specified area where each trajectory point of the driving trajectory is located, fuse the sub-maps with aligned features between different trajectories, and finally splice the sub-maps with associated positions in the target sub-map library, so as to continuously complete the elements in the map through the fusion of sub-maps with aligned features, and improve the accuracy of the spliced generated map. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for use in the description of the specific embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0054] Figure 1 FIG. shows a schematic architecture diagram of a vehicle control system according to an embodiment of the present application.

[0055] Figure 2 is a flowchart of a method for generating a map according to an exemplary embodiment.

[0056] Figure 3 is a flowchart of a method for generating a map according to an exemplary embodiment.

[0057] Figure 4 FIG. shows a schematic diagram of map fusion according to an embodiment of the present application.

[0058] Figure 5 FIG. shows a flowchart of map fusion and splicing according to an embodiment of the present application.

[0059] Figure 6 is a schematic structural diagram of a map generation device provided by an embodiment of the present application.

[0060] Figure 7 is a schematic structural diagram of a computer device provided by an optional embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0061] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative efforts fall within the protection scope of the present invention.

[0062] In the description of the embodiments of the present application, the term "corresponding" may indicate a direct or indirect corresponding relationship between two parties, may also indicate an association relationship between two parties, or may be a relationship such as indication and being indicated, configuration and being configured, etc.

[0063] Figure 1 The architecture diagram of a vehicle control system related to the embodiments of the present application is shown. As Figure 1 shown, the vehicle control system includes a target vehicle 110 and a cloud server 120.

[0064] A plurality of sensors are provided in the target vehicle 110, such as a position sensor, an image sensor, etc. During the driving process of the target vehicle 110, the position sensor of the target vehicle 110 will acquire the real-time position of the target vehicle 110 at a specified frequency, thereby forming a driving trajectory composed of a plurality of trajectory points during the driving process of the target vehicle 110.

[0065] Optionally, during the driving process of the target vehicle 110, the image sensor in the target vehicle 110 will also collect the surrounding environment information in real time and bind it to each trajectory point in the driving trajectory according to the time stamp.

[0066] Furthermore, an in-vehicle controller is also provided on the target vehicle 110. The in-vehicle controller can be a computer device with high computing power. The in-vehicle controller can receive the data of each sensor and process it to generate a feature map corresponding to each trajectory point in the driving trajectory.

[0067] Furthermore, the target vehicle 110 also includes a data transmission device. The data transmission device can transmit the data of each sensor on the target vehicle 110 to the cloud server 120 through a wireless network. After receiving the sensor data, the cloud server 120 can generate a feature map corresponding to each trajectory point in the driving trajectory of the target vehicle. At this time, the cloud server 120 can fuse the feature maps corresponding to the trajectory points of the driving trajectories of each vehicle, and then output a fused map.

[0068] Optionally, the cloud server 120 can also directly receive the set of feature maps uploaded by each vehicle, and then directly fuse the set of feature maps to output a fused map.

[0069] Optionally, the above cloud server may be a cloud server that provides technical computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN (Content Delivery Network), and big data and artificial intelligence platforms.

[0070] Optionally, the system may further include a management device, which is used to manage the system (such as managing the connection status between each module and the server), and the management device is connected to the server through a communication network. Optionally, the communication network is a wired network or a wireless network.

[0071] Optionally, the above wireless network uses standard communication technologies and / or protocols. The network is usually the Internet, but it can also be any other network, including but not limited to any combination of local area networks, metropolitan area networks, wide area networks, mobile, limited or wireless networks, private networks, or virtual private networks. In some embodiments, technologies and / or formats including Hypertext Markup Language, Extensible Markup Language, etc. are used to represent the data exchanged through the network. In addition, conventional encryption technologies such as Secure Sockets Layer, Transport Layer Security, Virtual Private Network, Internet Protocol Security, etc. can be used to encrypt all or some of the links. In other embodiments, customized and / or dedicated data communication technologies can also be used to replace or supplement the above data communication technologies.

[0072] Figure 2 is a flowchart of a method for generating a map shown according to an exemplary embodiment. This method is executed by a computer device, which can be set in a server, and the server can be a cloud server 120 as shown in Figure 1 As shown in Figure 2 As shown, the method for generating a map may include the following steps:

[0073] Step 201, obtain a target sub-gallery; the target sub-gallery includes at least two sub-gallery sets; each sub-gallery set includes respective sub-galleries.

[0074] The respective sub-galleries are respectively used to characterize the features of the specified areas where the respective trajectory points in the corresponding driving trajectory are located. Optionally, the respective trajectory points are sampled at a first distance in the driving trajectory.

[0075] During the driving process of the vehicle, the vehicle will perform real-time feature acquisition on the surrounding environment through sensors in the vehicle. The respective sub-galleries are also the feature maps obtained by the vehicle through sensors in the vehicle to collect the specified areas where the trajectory points are located when the vehicle travels to the respective trajectory points during the driving process. At this time, there are various elements in the sub-gallery, such as buildings, lane lines, etc. as the features in the sub-gallery.

[0076] Optionally, in the embodiments of the present application, there is a positional association between the subgraphs of each subgraph set. In the embodiments of the present application, the distance between the central points of the subgraphs with the positional association is less than a second distance; and the second distance is greater than a first distance. The subgraphs in the same subgraph set are actually the feature graphs around the respective trajectory points in the same driving trajectory. That is to say, in the embodiments of the present application, the distances between adjacent trajectory points in the driving trajectory are close, so that adjacent subgraphs in the same subgraph set are positionally associated.

[0077] Furthermore, there is an element overlap between adjacent subgraphs in the same subgraph set to avoid feature loss in the area passed by the driving trajectory when the subgraphs of the same subgraph set are subsequently fused.

[0078] Step 202, in the target subgraph library, perform feature fusion on the subgraphs with aligned features between the at least two subgraph sets to obtain an updated target subgraph library.

[0079] In the embodiments of the present application, the target subgraph library includes subgraph sets corresponding to each crowdsourcing trajectory. For example, for two subgraph sets corresponding to two crowdsourcing trajectories respectively, at this time, subgraphs that can be feature-aligned between the two subgraph sets can be obtained, and the subgraphs that can be feature-aligned are subjected to feature fusion.

[0080] In the embodiments of the present application, only the subgraphs with a positional association between two subgraph sets can be feature-aligned. Since feature alignment is used to determine whether subgraphs can be fused, the subgraphs to be fused need to meet the requirement that the positions of the central points of the subgraphs are close (that is, positionally associated).

[0081] Through the above process, the similar subgraphs in the target subgraph library are continuously fused, and the richness of the features of the subgraphs is continuously improved, so that the fused subgraphs can better represent the real environment.

[0082] Step 203, splice the subgraphs with a positional association in the updated target subgraph library to obtain a target map.

[0083] In the updated target subgraph library, the subgraphs with aligned features between different trajectories have been fused, that is, for different trajectories, the overlapping subgraphs have been fused; for the non-overlapping parts of different trajectories, the subgraphs with a positional association are spliced in sequence, and thus a fused map of the areas where each trajectory is located in the target subgraph library can be obtained.

[0084] In summary, when generating a map based on crowdsourcing trajectory fusion, first obtain a target sub-map library; the target sub-map library includes at least two sub-map sets; each sub-map set includes each sub-map; and each sub-map is respectively used to represent the characteristics of the specified area where each trajectory point in the corresponding driving trajectory is located; at this time, in the target sub-map library, determine the sub-maps corresponding to different sub-map sets, and perform feature fusion on the sub-maps with aligned features between different sub-map sets to obtain an updated target sub-map library; splice each sub-map with associated positions in the updated target sub-map library to obtain a target map. In the above solution, first generate sub-maps for the characteristics of the specified areas where each trajectory point of the driving trajectory is located, fuse the sub-maps with aligned features between different trajectories, and finally splice the sub-maps with associated positions in the target sub-map library, so as to continuously complete the elements in the map through the fusion of sub-maps with aligned features, and improve the accuracy of the spliced generated map.

[0085] Figure 3 It is a flowchart of a map generation method shown according to an exemplary embodiment. This method is executed by a computer device, and the computer device can be set in a target vehicle or a cloud server. As Figure 3 shown, this map generation method may include the following steps:

[0086] Step 301, obtain a target sub-map library.

[0087] The target sub-map library includes at least two sub-map sets; each sub-map set includes each sub-map; each sub-map is respectively used to represent the characteristics of the specified area where each trajectory point in the corresponding driving trajectory is located.

[0088] The key to sub-map generation is to extract the features (such as lane lines, buildings, etc.) around each trajectory point according to the data collected by the sensors during the vehicle driving process. In the embodiments of the present application, each sub-map set corresponds to a driving trajectory, and by sampling (the first distance), the interval between trajectory points is ensured to be reasonable, and there is no blind area in the sub-map coverage range.

[0089] Step 302, determine whether the sub-maps with associated positions between the at least two sub-map sets are feature-aligned.

[0090] In the embodiments of the present application, it is possible to determine whether the sub-maps that are between different sub-map sets but have associated positions are feature-aligned.

[0091] Optionally, in the embodiments of the present application, if there are multiple sub-graph sets, the multiple sub-graph sets can be sorted according to the generation time of the corresponding driving trajectories, and two sub-graph sets can be sequentially selected in chronological order to perform feature alignment and fusion operations; for example, the first sub-graph set and the second sub-graph set can be selected first, and then the second sub-graph set and the third sub-graph set can be selected, and so on, until a position association relationship of sub-graphs is established between all sub-graph sets and the fusion of sub-graphs with feature alignment is completed.

[0092] Optionally, if the feature alignment fails for the sub-graphs with position association between two candidate sub-graph sets, the feature alignment process between the two candidate sub-graph sets is skipped, and the next sub-graph set is sequentially selected.

[0093] If two sub-graphs are in different sub-graph sets, it means that the two sub-graphs respectively represent the features of the specified regions where the trajectory points of different driving trajectories are located; optionally, if there is a position association between the two sub-graphs, it means that the distance between the centers of the two sub-graphs is less than the second distance, that is, they are close in position; or if there is a position association between the two sub-graphs, it means that there is an overlapping part between the two sub-graphs.

[0094] Therefore, the sub-graphs with position association between at least two sub-graph sets represent the features near the overlapping or adjacent trajectory points in different driving trajectories, that is, the features of the same area. If the sub-graphs with position association between different sub-graph sets can be aligned, the features of the sub-graphs with position association but in different sub-graph sets can be fused according to the transformation matrix, thereby improving the reliability of the features of the same area.

[0095] Further, the computer device can determine whether the sub-graphs are aligned through clustering. Specifically, a first sub-graph and a second sub-graph are respectively selected from the at least two sub-graph sets; the first sub-graph and the second sub-graph are in different sub-graph sets, and the first sub-graph and the second sub-graph are in position association; the transformation matrix between the first sub-graph and the second sub-graph is determined, and the first sub-graph is converted into a third sub-graph according to the transformation matrix; the lane lines in the second sub-graph and the third sub-graph are respectively sampled to obtain first sampling points; the first sampling points are subjected to clustering processing to obtain respective first point clusters; and whether the first sub-graph and the second sub-graph are aligned is determined according to the variances of the respective first point clusters.

[0096] That is to say, in the embodiments of the present application, all lane line elements in two subgraphs to be aligned can be selected for statistics. First, the lane lines are sampled into point elements, and all the points are clustered. Points with close distances are clustered into one category. Then, the variance of the points in the principal component of the point cluster is calculated by the principal component analysis method. If the variance of the secondary component is greater than a threshold (determined according to the actual situation), the point cluster is considered a messy point cluster. Calculate the messy state of all point clusters. If the number of messy point clusters is greater than half of the total number of point clusters, it is considered that the current area alignment fails; otherwise, the alignment is successful.

[0097] Step 303A, if the features are aligned, the subgraphs with aligned features are subjected to feature fusion, and the fused subgraphs are saved in the set with an earlier time in the driving trajectory between at least two subgraph sets.

[0098] If the features are aligned, the subgraphs with aligned features can be subjected to feature fusion to obtain a fused subgraph. For example, when the first subgraph in the first subgraph set is feature-aligned with the second subgraph in the second subgraph set, at this time, the first subgraph and the second subgraph can be fused to obtain a first fused subgraph; and at this time, the first fused subgraph has both the position association relationship between the first subgraph and other subgraphs and the position association relationship between the second subgraph and other subgraphs.

[0099] Further, the computer device obtains the first quantity that is greater than the variance threshold among the variances of each first point cluster; if the first quantity is less than or equal to the quantity threshold, it is determined that the first subgraph is aligned with the second subgraph.

[0100] In the embodiments of the present application, the fusion weights corresponding to each subgraph with aligned features can also be determined according to the fusion times corresponding to each subgraph with aligned features; the greater the fusion times of the subgraph, the greater its proportion in the fusion; then, according to the fusion weights corresponding to each subgraph with aligned features, the subgraphs with aligned features are subjected to feature fusion.

[0101] For example, when the first fused subgraph obtained after fusing the first subgraph and the second subgraph is saved in the first subgraph set, at this time, the fusion times of the first fused subgraph is 1. If a new third subgraph set is added to the target subgraph library at this time, and the fifth subgraph in the third subgraph set is position-associated and feature-aligned with the first fused subgraph, then the fifth subgraph and the first fused subgraph can be subjected to feature fusion to obtain a second fused subgraph, and the fusion times of the second fused subgraph are set to 2.

[0102] Step 303B, if the subgraphs with position association between at least two subgraph sets are not feature-aligned, determine whether the subgraphs with position association between at least two subgraph sets are messy subgraphs.

[0103] Optionally, the computer device can also determine whether a sub-graph is a messy sub-graph through clustering processing. Specifically, for the fourth sub-graph that is not feature-aligned, sample the lane lines in the fourth sub-graph to obtain second sampling points; perform clustering processing on the second sampling points to obtain each second point cluster; determine whether the fourth sub-graph is a messy sub-graph according to the variances of each second point cluster.

[0104] Further, the computer device obtains the second quantity greater than the variance threshold among the variances of each second point cluster; if the second quantity is less than or equal to the quantity threshold, determine whether the fourth sub-graph is a messy sub-graph.

[0105] Step 304, if there is a non-messy sub-graph among the sub-graphs with position association between at least two sub-graph sets, use the non-messy sub-graph as a candidate sub-graph.

[0106] That is, in the embodiments of the present application, if the alignment of the sub-graphs with position association between different sub-graph sets fails, add the sub-graphs that meet the conditions of non-messy sub-graphs to the candidate pool for subsequent further analysis and processing. Although the candidate sub-graphs fail to align between the two sub-graph sets, their internal features are clear and stable and have high alignment potential, and can be used for alignment with the sub-graphs generated by other driving trajectories subsequently.

[0107] Step 305, when the number of candidate sub-graphs with the first association relationship is greater than the quantity threshold, perform pairwise feature alignment on the candidate sub-graphs with the first association relationship.

[0108] When the number of sub-graphs in different sets but with mutually associated positions (that is, there is the first association relationship, and at this time the spatial positions between the candidate sub-graphs are close, that is, the center point distance is less than the second distance) is greater than the quantity threshold (at least greater than or equal to 3), then at this time, pairwise feature alignment can be performed on the candidate sub-graphs with the first association relationship. For the specific implementation method of feature alignment, refer to Step 303A, which will not be elaborated here.

[0109] Step 306, if the candidate sub-graphs with the first association relationship are successfully aligned, fuse the successfully aligned candidate sub-graphs to obtain a target sub-graph, and delete the candidate sub-graphs with the first association relationship.

[0110] If there are successfully aligned candidate sub-graphs among the candidate sub-graphs with the first association relationship, fuse the successfully aligned candidate sub-graphs to obtain a target sub-graph, and delete all candidate sub-graphs with the first association relationship in the target sub-graph library, so as to further screen out sub-graphs with spatial consistency and structural similarity through the feature alignment between candidate sub-graphs, ensuring a reliable basis for fusion.

[0111] Optionally, if all the sub-graphs with position association between at least two sub-graph sets are messy sub-graphs, retain the sub-graphs with position association between the at least two sub-graph sets.

[0112] If the subgraphs with positional association between at least two subgraph sets are all messy subgraphs, then at this time, in order to avoid losing potentially useful information, the messy subgraphs can be retained in the subgraph set first.

[0113] Step 307: Stitch together the subgraphs with positional association in the updated target subgraph library to obtain the target map.

[0114] Please refer to Figure 4 , which shows a schematic diagram of map fusion involved in an embodiment of the present application. As Figure 4 shown, for example, sample a trajectory point every 10 meters; then take all the elements within 20 meters near the trajectory point as the center and form a subgraph with the current trajectory point. In fact, all the subgraphs in each trajectory have overlapping parts and should be associated. And for trajectory 1, if the subgraphs in it are fused with the subgraphs of driving trajectory 2, the fused subgraphs are still the associated subgraphs of trajectory 1;

[0115] At this time, the stitching and fusion process is to first select trajectory 1 in chronological order, and then fuse all the subgraphs associated with driving trajectory 1. And because the fused subgraphs are also positionally associated with some subgraphs of trajectory 2, the subgraphs of trajectory 1 and trajectory 2 can be associated. Therefore, in this way, the regional maps of the trajectories with overlap can be fused together. For non-overlapping trajectories, such as trajectory 3 that does not intersect with trajectory 1 or 2, it can be used as a new starting point for stitching and fusion to generate a map of another area.

[0116] Optionally, in the embodiment of the present application, when there is a new driving trajectory, the cloud server can obtain at least one driving trajectory; sample the at least one driving trajectory at intervals of a first distance to obtain each trajectory point in the at least one driving trajectory; generate a subgraph set corresponding to the at least one driving trajectory according to the features within a specified area centered on each of the trajectory points; and update the target subgraph library according to the subgraph set corresponding to the at least one driving trajectory.

[0117] Please refer to Figure 5 , which shows a flowchart of map fusion stitching involved in an embodiment of the present application. As Figure 5 shown, whenever an original trajectory data is obtained, a single-trajectory map is directly generated through the single-trajectory mapping logic, and the single-trajectory map is subgraph-segmented to obtain a new subgraph set. At this time, the new subgraph set and all the subgraph libraries (that is, the target subgraph library) are re-associated, aligned, fused, and updated through the solution shown in steps 301 to 307, so as to update the target subgraph library according to the fused subgraphs. Finally, the subgraphs with positional association in the updated target subgraph library are stitched together to generate the latest driving map.

[0118] In summary, when generating a map based on crowdsourcing trajectory fusion, first obtain a target sub-map library; the target sub-map library includes at least two sub-map sets; each sub-map set includes each sub-map; and each sub-map is respectively used to characterize the features of the designated area where each trajectory point in the corresponding driving trajectory is located; at this time, in the target sub-map library, determine the sub-maps corresponding to different sub-map sets, and fuse the sub-maps with aligned features between different sub-map sets to obtain an updated target sub-map library; splice the sub-maps with associated positions in the updated target sub-map library to obtain a target map. In the above solution, first generate sub-maps of the features of the designated areas where each trajectory point of the driving trajectory is located, fuse the sub-maps with aligned features between different trajectories, and finally splice the sub-maps with associated positions in the target sub-map library, so as to continuously complete the elements in the map through the fusion of the sub-maps with aligned features, and improve the accuracy of the spliced-generated map.

[0119] In an embodiment of the present application, a map generation device is further provided. This device is used to implement the above embodiments and preferred implementation manners, and those that have been described will not be repeated. As used below, the term "module" can be a combination of software and / or hardware that can achieve a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, implementation in hardware, or a combination of software and hardware is also possible and contemplated.

[0120] An embodiment of the present application provides a map generation device, Figure 6 which is a structural schematic diagram of a map generation device provided by an embodiment of the present application. The device includes:

[0121] A map library acquisition module 601, configured to acquire a target sub-map library; the target sub-map library includes at least two sub-map sets; each sub-map set includes each sub-map; and each sub-map is respectively used to characterize the features of the designated area where each trajectory point in the corresponding driving trajectory is located;

[0122] A sub-map update module 602, configured to fuse the sub-maps with aligned features between the at least two sub-map sets in the target sub-map library to obtain an updated target sub-map library;

[0123] A map fusion module 603, configured to splice the sub-maps with associated positions in the updated target sub-map library to obtain a target map.

[0124] The further function descriptions of the above various modules and units are the same as those in the corresponding above embodiments, and will not be repeated here.

[0125] In summary, when generating a map based on crowdsourcing trajectory fusion, first obtain a target sub-map library; the target sub-map library includes at least two sub-map sets; each sub-map set includes respective sub-maps; and each sub-map is respectively used to represent the characteristics of the specified area where each trajectory point in the corresponding driving trajectory is located; at this time, in the target sub-map library, determine the corresponding sub-maps between different sub-map sets, and perform feature fusion on the sub-maps with aligned features between different sub-map sets to obtain an updated target sub-map library; splice the sub-maps with associated positions in the updated target sub-map library to obtain a target map. In the above solution, first generate sub-maps for the characteristics of the specified area where each trajectory point of the driving trajectory is located, fuse the sub-maps with aligned features between different trajectories, and finally splice the sub-maps with associated positions in the target sub-map library, so as to continuously complete the elements in the map through the fusion of sub-maps with aligned features, and improve the accuracy of the spliced generated map.

[0126] The map generation device in this embodiment is presented in the form of functional units. Here, the unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and a memory that execute one or more software or fixed programs, and / or other devices that can provide the above functions.

[0127] The embodiment of the present invention further provides a computer device. The computer device is set in a target vehicle, or the computer device is implemented as a cloud server corresponding to the target vehicle. The computer device has the above Figure 6 shown map generation device.

[0128] Please refer to Figure 7 , Figure 7 which is a schematic structural diagram of a computer device provided by an optional embodiment of the present invention. As Figure 7 shown, the computer device includes: one or more processors 10, a memory 20, and interfaces for connecting various components, including a high-speed interface and a low-speed interface. Each component communicates with each other using different buses and can be installed on a common main board or installed in other ways as needed. The processor can process instructions executed within the computer device, including instructions stored in the memory or on the memory to display graphic information in a graphical user interface on an external input / output device (such as a display device coupled to the interface). In some optional embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Similarly, multiple computer devices can be connected, and each device provides some necessary operations (such as an array of servers, a set of blade servers, or a multi-processor system). Figure 7 In

[0129] The processor 10 may be a central processing unit, a network processor, or a combination thereof. Among them, the processor 10 may further include a hardware chip. The above-mentioned hardware chip may be an application-specific integrated circuit, a programmable logic device, or a combination thereof. The above-mentioned programmable logic device may be a complex programmable logic device, a field programmable gate array, a generic array logic, or any combination thereof.

[0130] Among them, the memory 20 stores instructions executable by at least one processor 10, so that the at least one processor 10 executes the method shown in the above embodiments.

[0131] The memory 20 may include a program storage area and a data storage area. Among them, the program storage area may store an operating system and application programs required for at least one function; the data storage area may store data created according to the use of the computer device, etc. In addition, the memory 20 may include a high-speed random access memory, and may further include a non-transitory memory, such as at least one magnetic disk storage device, a flash memory device, or other non-transitory solid-state storage devices. In some alternative embodiments, the memory 20 may optionally include a memory remotely provided with respect to the processor 10, and these remote memories may be connected to the computer device through a network. Examples of the above-mentioned network include but are not limited to the Internet, an enterprise intranet, a local area network, a mobile communication network, and combinations thereof.

[0132] The memory 20 may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a flash memory, a hard disk, or a solid-state drive; the memory 20 may further include a combination of the above types of memories.

[0133] The computer device further includes a communication interface 30 for the computer device to communicate with other devices or communication networks.

[0134] Embodiments of the present invention also provide a computer-readable storage medium. The method according to the embodiments of the present invention can be implemented in hardware, firmware, or be implemented as computer code that can be recorded on a storage medium, or be implemented as computer code that is originally stored in a remote storage medium or a non-transitory machine-readable storage medium and downloaded through a network and will be stored in a local storage medium, so that the method described herein can be stored as such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only memory, a random access memory, a flash memory, a hard disk, or a solid-state drive, etc.; further, the storage medium can also include a combination of the above-mentioned types of memories. It can be understood that a computer, a processor, a microprocessor controller, or programmable hardware includes a storage component that can store or receive software or computer code. When the software or computer code is accessed and executed by the computer, the processor, or the hardware, the method shown in the above embodiments is implemented.

[0135] A part of the present invention can be applied as a computer program product, for example, computer program instructions. When executed by a computer, through the operation of the computer, the methods and / or technical solutions according to the present invention can be called or provided. Those skilled in the art should understand that the forms in which computer program instructions exist in a computer-readable medium include, but are not limited to, source files, executable files, installation package files, etc. Correspondingly, the ways in which computer program instructions are executed by a computer include, but are not limited to: the computer directly executes the instructions, or the computer compiles the instructions and then executes the corresponding compiled program, or the computer reads and executes the instructions, or the computer reads and installs the instructions and then executes the corresponding installed program. Herein, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to the computer.

[0136] Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A map generation method, characterized in that: The method comprises: Acquire a target sub-graph library; the target sub-graph library includes at least two sub-graph sets; each sub-graph set includes individual sub-graphs; the individual sub-graphs are used to represent the characteristics of the designated area where each track point in the corresponding driving track is located; In the target sub-graph library, feature fusion is performed on the sub-graphs whose features are aligned between the at least two sub-graph sets to obtain an updated target sub-graph library; The sub-maps associated with the locations in the updated target sub-map library are stitched together to obtain the target map.

2. The method according to claim 1, characterized in that: The method further comprises: Obtain at least one driving trajectory; Sampling the at least one driving trajectory at intervals of a first distance to obtain each trajectory point in the at least one driving trajectory; Generate at least one subgraph set corresponding to the driving trajectory according to the features in the specified area centered at each trajectory point; The target sub-graph library is updated according to the sub-graph set corresponding to the at least one driving trajectory.

3. The method according to claim 2, characterized in that The distance between the center points of the position-associated subgraphs is less than the second distance; and the second distance is greater than the first distance.

4. The method according to any one of claims 1 to 3, characterized in that: The step of fusing the subgraphs whose features are aligned between the at least two subgraph sets in the target subgraph library to obtain an updated target subgraph library includes: Determining whether the subgraphs of the at least two subgraph sets that are positionally associated are feature aligned; If the features are aligned, the subgraphs with the aligned features are subjected to feature fusion, and the fused subgraphs are stored in a set with an earlier time of the driving trajectory between at least two subgraph sets.

5. The method according to claim 4, characterized in that The method further comprises: If the subgraphs associated with the positions of at least two subgraph sets are not feature aligned, determining whether the subgraphs associated with the positions of at least two subgraph sets are messy subgraphs; If the subgraphs associated with each other in positions between at least two subgraph sets are all messy subgraphs, the subgraphs associated with each other in positions between the at least two subgraph sets are retained.

6. The method according to claim 5, characterized in that The method further comprises: If there is a non-messy subgraph among the subgraphs that are positionally associated between at least two subgraph sets, the non-messy subgraph is used as a candidate subgraph; When the number of candidate subgraphs with the first association relationship is greater than the number threshold, pairwise feature alignment is performed on the candidate subgraphs with the first association relationship; If the candidate subgraphs with the first association relationship are aligned successfully, the successfully aligned candidate subgraphs are fused to obtain a target subgraph, and the candidate subgraphs with the first association relationship are deleted.

7. The method according to claim 5, characterized in that The determining whether the subgraphs of the at least two subgraph sets that are positionally associated are feature aligned includes: A first sub-graph and a second sub-graph are respectively selected from the at least two sub-graph sets; the first sub-graph and the second sub-graph are located in different sub-graph sets, and the first sub-graph and the second sub-graph are positionally associated; Determine a transformation matrix between the first sub-image and the second sub-image, and transform the first sub-image into a third sub-image according to the transformation matrix; Sampling the lane lines in the second sub-image and the third sub-image respectively to obtain first sampling points; Performing clustering processing on the first sampling points to obtain first point clusters; Whether the first sub-image is aligned with the second sub-image is determined according to the variances of the first point clusters.

8. The method according to claim 7, characterized in that The step of determining whether the first sub-image is aligned with the second sub-image according to the variances of the point clusters includes: Obtaining a first number of variances of each of the first point clusters that is greater than a variance threshold; If the first number is less than or equal to the number threshold, it is determined that the first sub-image is aligned with the second sub-image.

9. The method according to claim 5, characterized in that The determining whether the subgraphs with position association between at least two subgraph sets are messy subgraphs includes: For the fourth sub-image without feature alignment, sampling the lane lines in the fourth sub-image to obtain second sampling points; Performing clustering processing on the second sampling points to obtain second point clusters; According to the variances of the second point clusters, it is determined whether the fourth sub-graph is a messy sub-graph.

10. The method according to claim 9, characterized in that The step of determining whether the fourth subgraph is a messy subgraph according to the variances of the second point clusters includes: Obtaining a second number of variances of each of the second point clusters that is greater than a variance threshold; If the second number is less than or equal to the number threshold, it is determined whether the fourth sub-graph is a messy sub-graph.

11. The method according to any one of claims 1 to 3, characterized in that: The step of fusing the subgraphs whose features are aligned between the at least two subgraph sets includes: Determining the fusion weights corresponding to the sub-graphs of the feature alignment according to the fusion times corresponding to the sub-graphs of the feature alignment; According to the fusion weights corresponding to the feature-aligned sub-graphs, the features of the feature-aligned sub-graphs are fused.

12. A map generating device, characterized in that: The device comprises: A library acquisition module is used to acquire a target sub-image library; the target sub-image library includes at least two sub-image sets; each sub-image set includes individual sub-images; and each sub-image is used to represent the characteristics of a designated area where each track point in the corresponding driving track is located; A sub-graph updating module, configured to perform feature fusion on sub-graphs with aligned features between the at least two sub-graph sets in the target sub-graph library, so as to obtain an updated target sub-graph library; The map fusion module is used to stitch together the sub-maps associated with the locations in the updated target sub-map library to obtain the target map.

13. A computer device, characterized in that: include: A memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the map generation method according to any one of claims 1 to 7 by executing the computer instructions.

14. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a computer to execute the map generation method according to any one of claims 1 to 7.

Images