Map generation method and related products

Abstract

The embodiment of the present application discloses the field of smart cars, and in particular, relates to a high-precision map generation method and related products in the field of high-precision map generation. The method includes: obtaining a traffic expression graph, the traffic expression graph is used to express static traffic scenes, the traffic expression graph includes at least two layers of graph structures, the at least two layers of graph structures include a first layer of graph structure and a second layer of graph structure, the type of nodes in the first layer of graph structure is different from the type of nodes in the second layer of graph structure; according to the traffic expression graph, a traffic expression graph sequence is obtained, and the traffic expression graph sequence is a serialized expression of the traffic expression graph; according to the traffic expression graph sequence, a map sequence is obtained; according to the map sequence, an electronic map is obtained; compared with expressing static traffic scenes in the form of isomorphic graphs, various elements contained in static traffic scenes can be more accurately expressed in a unified and global manner, thereby providing more realistic road network surrounding facilities information.

Description

Technical Field

[0001] The present application relates to the field of computers, and in particular to a map generation method and related products. Background Art

[0002] High definition map (HD map): An electronic map with high precision, high freshness and high richness, with absolute and relative accuracy within 1 meter. High-precision map is called HD map in English, which is defined from the perspective of data accuracy and feature richness. From the perspective of the classification standard of autonomous driving functions, high-precision map can be called HAD map (highly automated driving map) in English. High-precision maps contain rich information, including road information such as road type, curvature, lane line position, as well as environmental object information such as roadside infrastructure, obstacles, traffic signs, etc., as well as real-time dynamic information such as traffic flow and traffic light status information.

[0003] As the level of autonomous driving intelligence continues to improve, the road conditions that autonomous driving systems need to face and process will become increasingly complex, which requires the core technical capabilities of "perception, decision-making and execution" to be stronger, more robust and safer. Simulation evaluation of autonomous driving systems is one of the powerful ways to effectively reduce the cost of algorithm testing and improve verification efficiency. Among them, the breadth and richness of the road network coverage of high-precision maps are important guarantees for the adequacy of algorithm verification. Currently, the open source high-precision map data sets are all collected through actual vehicles and sensors and completed through manual annotation methods. Although these data sets are authentic and reliable, they have high production costs and long cycles, and the data volume and data richness are difficult to meet the needs of autonomous driving evaluation. Therefore, it is necessary to study how to efficiently generate high-precision maps that can be used for autonomous driving simulation evaluation. Summary of the invention

[0004] The embodiments of the present application disclose a map generation method and related products, which can efficiently generate high-precision maps that can be used for autonomous driving simulation evaluation.

[0005] In a first aspect, an embodiment of the present application provides a map generation method, the method comprising: obtaining a traffic expression map, the traffic expression map being used to express a static traffic scene, the traffic expression map comprising at least two layers of graph structures, the at least two layers of graph structures comprising a first layer of graph structure and a second layer of graph structure, the type of nodes in the first layer of graph structure being different from the type of nodes in the second layer of graph structure; obtaining a traffic expression map sequence according to the traffic expression map, the traffic expression map sequence being a serialized expression of the traffic expression map; obtaining a map sequence according to the traffic expression map sequence, the map sequence being a serialized expression of an electronic map; obtaining the electronic map according to the map sequence.

[0006] In an embodiment of the present application, a traffic expression graph includes a first-layer graph structure and a second-layer graph structure, and the type of nodes in the first-layer graph structure is different from the type of nodes in the second-layer graph structure. The traffic expression graph (which can be understood as a heterogeneous graph) includes a first-layer graph structure and a second-layer graph structure, and the type of nodes in the first-layer graph structure is different from the type of nodes in the second-layer graph structure. Therefore, multiple types of elements in a static traffic scene can be comprehensively expressed, and more complex static traffic scenes can be effectively represented. The map generation method provided in an embodiment of the present application expresses a static traffic scene through a traffic expression graph. Compared with expressing a static traffic scene in the form of an isomorphic graph (there is only one type of node in the isomorphic graph and the attribute type, number of parameters, etc. of the node description are the same), it can more accurately and effectively express the various elements contained in the static traffic scene in a unified and global manner, thereby providing more realistic road network surrounding facility information and having better scalability.

[0007] In one possible implementation, the traffic expression graph sequence includes an element sequence and a constraint sequence, the element sequence represents the attribute information of the nodes in the traffic expression graph, and the constraint sequence represents the connection relationship between any two nodes in the first-layer graph structure, and the connection relationship between any node in the first-layer graph structure and any node in the second-layer graph structure.

[0008] In this implementation, the traffic expression graph sequence includes an element sequence and a constraint sequence. The element sequence represents the attribute information of the nodes in the traffic expression graph, and the constraint sequence represents the connection relationship between any two nodes in the first-layer graph structure, and the connection relationship between any node in the first-layer graph structure and any node in the second-layer graph structure; it can more accurately and effectively express the various elements contained in the static traffic scene and the relationship between the various elements in a unified and global manner.

[0009] In a possible implementation, obtaining a traffic expression graph sequence according to the traffic expression graph includes: serializing the attribute information of one or more nodes in the traffic expression graph to obtain an element sequence; serializing the intra-layer connection relationship and the inter-layer connection relationship in the traffic expression graph to obtain a constraint sequence; the intra-layer connection relationship includes the connection relationship between any two nodes in the first-layer graph structure, and the inter-layer connection relationship includes the connection relationship between any node in the first-layer graph structure and any node in the second-layer graph structure; using the combination of the element sequence and the constraint sequence as the serialized expression of the traffic expression graph to obtain the traffic expression graph sequence. In the present application, the inter-layer connection relationship may be referred to as an inter-layer pointing relationship.

[0010] In this implementation, the attribute information of each node in the traffic expression graph is serialized to obtain an element sequence; the intra-layer connection relationship and inter-layer connection relationship in the traffic expression graph are serialized to obtain a constraint sequence; and by encoding and describing each node, intra-layer connection relationship, and inter-layer connection relationship in the traffic expression graph respectively, a sequence representation obtained by encoding the traffic expression graph, i.e., a traffic expression graph sequence, is obtained.

[0011] In a possible implementation, the traffic expression graph is a stacked hierarchical graph, and the first-layer graph structure and the second-layer graph structure are located in different layers (or hierarchies).

[0012] In this implementation, the traffic expression diagram is a stacked hierarchical diagram, which can more accurately and effectively express the relationship between various elements contained in the static traffic scene in a unified and global manner.

[0013] In a possible implementation, the first layer graph structure and the second layer graph structure are constructed based on urban planning information; the urban planning information is used to obtain an urban planning map; and the connection relationship between the nodes in the first layer graph structure and the nodes in the second layer graph structure is determined to obtain the traffic expression map.

[0014] In this implementation, the connection relationship between the nodes in the first-layer graph structure and the nodes in the second-layer graph structure is determined, and the traffic expression graph obtained can be regarded as a "stacked" form, integrating multiple layers of graphs (such as the first-layer graph structure and the second-layer graph structure) into one traffic expression graph; a traffic expression graph that expresses static traffic scenes can be accurately formed.

[0015] In a possible implementation, constructing the first-layer graph structure based on the urban planning information includes: extracting corner points in the urban planning map corresponding to the urban planning information as vertices to obtain multiple vertices; determining the connection relationship between the multiple vertices to obtain the first-layer graph structure.

[0016] In this implementation, the connection relationship between multiple vertices is determined to obtain a first-level graph structure; the first-level graph structure can be quickly constructed.

[0017] In a possible implementation, constructing the second-layer graph structure based on the urban planning information includes: extracting traffic elements belonging to the first type from the urban planning map corresponding to the urban planning information, the first type of traffic elements including: any one of functional areas, intersections, and lane areas; determining the connection relationship between the first type of traffic elements to obtain the second-layer graph structure.

[0018] In this implementation, a second-layer graph structure is obtained based on the connection relationship between the first type of traffic elements, so that the second-layer graph structure and other layer graph structures can be used to construct a traffic expression graph that expresses a static traffic scene.

[0019] In a possible implementation, each node in the first-layer graph structure is of a first type, each node in the second-layer graph structure is of a second type, and the first type is different from the second type. Both the first-layer graph structure and the second-layer graph structure are isomorphic graphs.

[0020] In this implementation, the first-layer graph structure expresses the relationship between nodes of the first type and nodes of the first type, and the second-layer graph structure expresses the relationship between nodes of the second type and nodes of the second type. The traffic expression graph includes the first-layer graph structure and the second-layer graph structure, which can more accurately and effectively express the various elements contained in the static traffic scene in a unified and global manner.

[0021] In a possible implementation, the map sequence includes a first element sequence and a second element sequence, the length of the first element sequence is different from the length of the second element sequence, the first element sequence represents a first traffic element, the second element sequence represents a second traffic element, the type of the first traffic element is different from the type of the second traffic element, and the first traffic element and the second traffic element correspond to traffic elements in the static traffic scene.

[0022] In this implementation, the lengths of the first element sequence and the second element sequence are different, and different types of traffic elements are represented by sequences of indefinite length, which can accurately represent different types of traffic elements.

[0023] In a possible implementation, the map sequence also includes a first constraint sequence and a second constraint sequence, the length of the first constraint sequence is different from the length of the second constraint sequence, the first constraint sequence represents the connection relationship between the third traffic element and the fourth traffic element, the second constraint sequence represents the connection relationship between the fifth traffic element and the sixth traffic element, and the third traffic element, the fourth traffic element, the fifth traffic element, and the sixth traffic element correspond to the traffic elements in the static traffic scene.

[0024] In this implementation, the first constraint sequence and the second constraint sequence have different lengths, and different types of connection relationships are represented by sequences of indefinite length, which can accurately represent different types of connection relationships.

[0025] In a possible implementation, the traffic expression map sequence represents the functional areas (land types), roads, and intersection information in the city planning map, and the map sequence represents one or more of crosswalks, sidewalks, traffic lights, stop lines, lanes, and road attachments. The map sequence represents more detailed road network information. In other words, the city planning map includes land types, roads, etc., but does not contain high-precision map information such as specific lanes and road attachments; the map sequence represents high-precision map information such as lanes and road attachments.

[0026] In this implementation, the traffic expression map sequence represents the functional areas (land use types), roads, and intersection information in the city planning map. Users can generate the corresponding electronic map by inputting the city planning map, and can efficiently and conveniently generate high-precision map data that can be used for autonomous driving simulation evaluation.

[0027] In a possible implementation, the electronic map is a high-precision map (or high-accuracy map).

[0028] In this implementation, a high-precision map can be generated by inputting a city planning map, which can efficiently and conveniently generate high-precision map data that can be used for autonomous driving simulation evaluation.

[0029] In a possible implementation, the traffic expression graph also includes a third-layer graph structure and a fourth-layer graph structure, and the types of nodes in any two layers of the first-layer graph structure, the second-layer graph structure, the third-layer graph structure, and the fourth-layer graph structure are different.

[0030] In a second aspect, an embodiment of the present application provides a map generation device, which has the function of implementing the operations in the above-mentioned first aspect method embodiment. The function can be implemented by hardware, or can be implemented by hardware executing corresponding software. The hardware or software includes one or more modules corresponding to the above-mentioned functions. In a possible implementation, the device includes: an acquisition unit, which is used to acquire a traffic expression map, the traffic expression map is used to express a static traffic scene, the traffic expression map includes at least two layers of graph structure, the at least two layers of graph structure include a first layer of graph structure and a second layer of graph structure, and the type of nodes in the first layer of graph structure is different from the type of nodes in the second layer of graph structure; an encoding unit, which is used to obtain a traffic expression map sequence according to the traffic expression map, and the traffic expression map sequence is a serialized expression of the traffic expression map; a processing unit, which is used to obtain a map sequence according to the traffic expression map sequence, and the map sequence is a serialized expression of an electronic map; and the electronic map is obtained according to the map sequence.

[0031] In one possible implementation, the traffic expression graph sequence includes an element sequence and a constraint sequence, the element sequence represents the attribute information of the nodes in the traffic expression graph, and the constraint sequence represents the connection relationship between any two nodes in the first-layer graph structure, and the connection relationship between any node in the first-layer graph structure and any node in the second-layer graph structure.

[0032] In a possible implementation, the encoding unit is specifically used to serialize the attribute information of one or more nodes in the traffic expression graph to obtain an element sequence; serialize the intra-layer connection relationship and the inter-layer connection relationship in the traffic expression graph to obtain a constraint sequence; the intra-layer connection relationship includes the connection relationship between any two nodes in the first-layer graph structure, and the inter-layer connection relationship includes the connection relationship between any node in the first-layer graph structure and any node in the second-layer graph structure; the combination of the element sequence and the constraint sequence is used as a serialized expression of the traffic expression graph to obtain the traffic expression graph sequence.

[0033] In a possible implementation, the acquisition unit is specifically used to construct the first layer graph structure and the second layer graph structure based on urban planning information; the urban planning information is used to obtain the urban planning map; the connection relationship between the nodes in the first layer graph structure and the nodes in the second layer graph structure is determined to obtain the traffic expression map.

[0034] In a possible implementation, the acquisition unit is specifically used to extract corner points in the urban planning map corresponding to the urban planning information as vertices to obtain multiple vertices; determine the connection relationship between the multiple vertices to obtain the first layer graph structure.

[0035] In a possible implementation, the acquisition unit is specifically used to extract traffic elements belonging to the first type from the urban planning map corresponding to the urban planning information, and the first type of traffic elements includes: any one of functional areas, intersections, and lane areas; determine the connection relationship between the first type of traffic elements to obtain the second layer graph structure.

[0036] In a possible implementation, each node in the first-layer graph structure belongs to a first type, each node in the second-layer graph structure belongs to a second type, and the first type is different from the second type.

[0037] In a possible implementation, the map sequence includes a first element sequence and a second element sequence, the length of the first element sequence is different from the length of the second element sequence, the first element sequence represents a first traffic element, the second element sequence represents a second traffic element, the type of the first traffic element is different from the type of the second traffic element, and the first traffic element and the second traffic element correspond to traffic elements in the static traffic scene.

[0038] In a possible implementation, the second element sequence also includes a first constraint sequence and a second constraint sequence, the length of the first constraint sequence is different from the length of the second constraint sequence, the first constraint sequence represents the connection relationship between the third traffic element and the fourth traffic element, the second constraint sequence represents the connection relationship between the fifth traffic element and the sixth traffic element, and the third traffic element, the fourth traffic element, the fifth traffic element, and the sixth traffic element correspond to the traffic elements in the static traffic scene.

[0039] In a possible implementation, the traffic expression map sequence represents the functional areas (land types), roads, and intersection information in the city planning map, and the map sequence represents one or more of crosswalks, sidewalks, traffic lights, stop lines, lanes, and road attachments. The map sequence represents more detailed road network information. In other words, the city planning map includes land types, roads, etc., but does not contain high-precision map information such as specific lanes and road attachments; the map sequence represents high-precision map information such as lanes and road attachments.

[0040] In a possible implementation, the electronic map is a high-precision map (or high-accuracy map).

[0041] In a possible implementation, the map generating device further includes: an input unit, configured to input the city planning map.

[0042] In a possible implementation, the map generating device further includes: an output unit for displaying a global city planning map; and an input unit for inputting a portion of the global city planning map selected by a user as the city planning map.

[0043] In a possible implementation, the map generating apparatus further includes: a communication unit, configured to receive the city planning map sent by a user via a terminal device, and send the electronic map to the terminal device.

[0044] In a possible implementation, the traffic expression graph also includes a third-layer graph structure and a fourth-layer graph structure, and the types of nodes in any two layers of the first-layer graph structure, the second-layer graph structure, the third-layer graph structure, and the fourth-layer graph structure are different.

[0045] Regarding the technical effects brought about by the second aspect or various possible implementations of the second aspect, reference may be made to the introduction to the technical effects of the first aspect or various possible implementations of the first aspect.

[0046] In a third aspect, an embodiment of the present application provides another map generating device, which includes a processor, which can be used to execute computer-executable instructions stored in a memory to execute the method shown in the above-mentioned first aspect or any possible implementation of the first aspect.

[0047] In a possible implementation, the memory is located outside the map generating device.

[0048] In a possible implementation, the memory is located within the map generating device.

[0049] In the embodiment of the present application, the processor and the memory may also be integrated into one device, that is, the processor and the memory may also be integrated together.

[0050] In a possible implementation, the map generating apparatus further includes an input / output device, and the input / output device is used to input a city planning map and output an electronic map.

[0051] In a fourth aspect, an embodiment of the present application provides another map generating device, the map generating device comprising a processing circuit and an interface circuit, the interface circuit being used to obtain data or output data, such as inputting a city planning map and outputting an electronic map; the processing circuit being used to execute the corresponding method as shown in the first aspect or any possible implementation of the first aspect.

[0052] In a fifth aspect, the present application provides a computer-readable storage medium for storing a computer program, which, when executed on a computer, enables the method shown in the above-mentioned first aspect or any possible implementation of the first aspect to be executed.

[0053] In a sixth aspect, the present application provides a computer program product, which includes a computer program or a computer code, and when the computer program product is run on a computer, the method shown in the above-mentioned first aspect or any possible implementation of the first aspect is executed. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the background technology, the drawings required for use in the embodiments of the present application or the background technology will be described below.

[0055] Figure 1 An example of a city planning map provided in an embodiment of the present application;

[0056] Figure 2 An example of a heterogeneous graph provided in an embodiment of the present application;

[0057] Figure 3 This is an example of a high-precision map visualization screenshot corresponding to the city planning map;

[0058] Figure 4 A flow chart of a map generation method provided in an embodiment of the present application;

[0059] Figure 5A An example of a vertex layer graph structure provided in an embodiment of the present application;

[0060] Figure 5B An example of a lane area layer map structure provided in an embodiment of the present application;

[0061] Figure 5C An example of an intersection layer graph structure provided in an embodiment of the present application;

[0062] Figure 5D An example of a functional layer diagram structure provided in an embodiment of the present application;

[0063] Figure 6 A flowchart of another map generation method provided in an embodiment of the present application;

[0064] Figure 7 An example of a traffic expression diagram provided in an embodiment of the present application;

[0065] Figure 8 An example of a visualization result of an electronic map provided in an embodiment of the present application;

[0066] Fig. 9 A schematic diagram of the structure of a map generating device provided in an embodiment of the present application;

[0067] Fig.10 A schematic diagram of the structure of a terminal device provided in an embodiment of the present application;

[0068] Fig.11 It is a structural diagram of a server provided in an embodiment of the present application. DETAILED DESCRIPTION

[0069] The terms "first" and "second" in the specification, claims and drawings of this application are only used to distinguish different objects, rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units that are not listed, or may optionally include other steps or units that are inherent to these processes, methods, products or devices.

[0070] The "embodiment" mentioned in this article means that the specific features, structures or characteristics described in conjunction with the embodiment can be included in at least one embodiment of the present application. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It can be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0071] The terms used in the following embodiments of the present application are only for the purpose of describing specific embodiments, and are not intended to be used as limitations to the present application. As used in the specification and the appended claims of the present application, the singular expressions "one", "a kind of", "said", "above", "the" and "this" are intended to also include plural expressions, unless there is a clear contrary indication in the context. It should also be understood that the term "and / or" used in the present application refers to and includes any or all possible combinations of one or more listed items. For example, "A and / or B" can represent: only A exists, only B exists, and A and B exist at the same time, wherein A, B can be singular or plural. The term "multiple" used in the present application refers to two or more.

[0072] As described in the background technology section, it is currently necessary to study how to efficiently generate a method for high-precision maps that can be used for autonomous driving simulation evaluation. The present application provides a method for efficiently generating high-precision maps that can be used for autonomous driving simulation evaluation. In the map generation method provided in the present application, the static traffic scene in the city planning map is expressed in a heterogeneous graph (i.e., a heterogeneous graph is used to express the static traffic scene in the city planning map), which can achieve a relatively unified and global expression of the static traffic scene. In the present application, the traffic expression graph is a heterogeneous graph. The form of the heterogeneous graph allows the graph to contain multiple types of nodes of different types, and the attribute types and parameter numbers describing the nodes can be different, so multiple types of elements in the static traffic scene can be comprehensively expressed, that is, more complex static traffic scenes can be effectively represented. Since heterogeneous graphs can represent more complex static traffic scenes than isomorphic graphs, the map generation scheme provided in the present application that uses heterogeneous graphs to represent static traffic scenes can provide real road network surrounding facilities information (or road network surrounding facilities information) compared with the map generation scheme that uses isomorphic graphs to represent static traffic scenes, and has better scalability. In addition, the rich information about road network surrounding facilities provided by heterogeneous graphs can provide a reference for generating more realistic traffic flow scenarios.

[0073] The following is a brief introduction to the scenarios to which the map generation method provided in the application embodiment is applicable.

[0074] Map generation scenario 1: The user inputs a city planning map (including the corresponding city planning legend analysis) to the map generation device through an input device (such as a mouse, keyboard, etc.); the map generation device generates a high-precision map file based on the city planning map. For example, the user inputs a city planning map in JPEG (joint photographic experts group) format or portable network graphics (PNG) format to the map generation device through an input device. JPEG is a product of the JPEG standard, which was developed by the International Organization for Standardization and is a compression standard for continuous-tone still images. The JPEG format is a commonly used image file format with the suffix .jpg or .jpeg. PNG is a bitmap format that uses a lossless compression algorithm with the suffix .png. The map generation device can be a terminal device with certain data processing capabilities, such as a desktop computer, a laptop computer, etc.

[0075] Map generation scenario 2: The user selects a part of the global city planning map provided by the map generation device as a local city planning map through an input device (such as a mouse, keyboard, etc.); the map generation device generates a high-precision map file based on the local city planning map. In this scenario, the map generation device provides a global city planning map for the user to select, and the user can select any continuous area in the global city planning map as a local city planning map; the map generation device generates a high-precision map file (corresponding to the continuous area) based on the local city planning map. The map generation device can be a terminal device with certain data processing capabilities, such as a desktop computer, a laptop computer, etc.

[0076] Map generation scenario 3: A user sends a city planning map (including corresponding city planning legend analysis) to a map generation device (such as a server) through a terminal device (such as a mobile phone, laptop, etc.); the map generation device generates a high-precision map file based on the city planning map and sends the high-precision map file to the terminal device.

[0077] The following first introduces the technical features related to the map generation method provided by this application.

[0078] Isomorphic Graph

[0079] There is only one type of node in the isomorphic graph, and the attribute type and number of parameters describing the node are the same. Currently, the commonly used methods are to express static traffic scenes in the form of isomorphic graphs. Since such methods are limited to describing static traffic scenes or focus on a specific aspect, such as the description of the road network, it is difficult to express the various elements contained in the static traffic scene in a unified and global manner.

[0080] Heterogeneous Graph

[0081] The form of heterogeneous graph allows the graph to contain multiple types of nodes, and the attribute types and parameter numbers describing the nodes can be different. Therefore, multiple types of elements in static traffic scenes can be comprehensively expressed, that is, more complex static traffic scenes can be effectively characterized. The heterogeneous graph (i.e., traffic expression graph) provided in the present application may include a multi-layer graph structure. Exemplarily, the heterogeneous graph provided in the present application is divided into four layers of graph structure, namely: vertex layer, intersection layer, lane area layer, and functional area layer. The graph structure of each layer is in the form of an isomorphic graph, that is, each layer of the graph structure contains only one type of node. Each node in the intersection layer (i.e., intersection layer node) represents an intersection. Each node in the vertex layer represents a vertex layer node. Each node in the lane area layer represents a lane area layer node, such as a lane area. Each node in the functional area layer represents a functional area layer node, such as a functional area. Vertex layer nodes, intersection layer nodes, lane area layer nodes, and functional area layer nodes are four different types of nodes, which will be described later. In a heterogeneous graph, the connection relationship between nodes at different layers can be defined (or expressed), for example, the road connection relationship between nodes at the intersection layer and nodes at the lane area is defined. In a possible implementation, the map generation device fuses the multi-layer graph structure into a heterogeneous graph (i.e., a traffic expression graph) in a "stacked" form to form a heterogeneous graph expression model for static traffic scenes. Figure 1 An example of a city planning map provided in an embodiment of the present application. Figure 2 An example of a heterogeneous graph provided in an embodiment of the present application. Figure 2 The heterogeneous graph in is used to characterize Figure 1 A static traffic scene in the area in the black matrix box in . Exemplarily, the heterogeneous graph provided in the present application is divided into three-layer graph structures, namely: vertex layer, intersection layer, and functional area layer. Exemplarily, the heterogeneous graph provided in the present application is divided into three-layer graph structures, namely: vertex layer, lane area layer, and functional area layer. Exemplarily, the heterogeneous graph provided in the present application is divided into two-layer graph structures, namely: vertex layer, lane area layer. It should be understood that the heterogeneous graph provided in the present application is divided into two or more layers of graph structures, that is, two or more layers of graph structures are fused into a heterogeneous graph, and each layer of graph structure is a isomorphic graph.

[0082] Serialized representation of heterogeneous graphs and electronic maps

[0083] The present application defines a sequence representation of a heterogeneous graph and a sequence representation of an electronic map (e.g., a high-precision map). The present application defines (or provides) encoding rules for obtaining a heterogeneous graph sequence based on a heterogeneous graph and an encoding rule (or encoding method) for obtaining a map sequence based on an electronic map. The following describes the definition of a heterogeneous graph sequence, the process of generating a heterogeneous graph sequence based on a heterogeneous graph, and the definition of a map sequence, the process of generating a map sequence based on an electronic map.

[0084] The heterogeneous graph sequence (i.e., the traffic map expression sequence) is divided into two parts: elements and constraints (i.e., element sequence and constraint sequence), and the definitions and coding rules are described separately. The element sequence part of the heterogeneous graph sequence is the coding sequence of the attribute information (or the node data of each layer) of the nodes in each layer of the heterogeneous graph, and the constraint sequence part is the coding sequence of the intra-layer connection and inter-layer connection relationship in the heterogeneous graph. Exemplarily, the heterogeneous graph contains four layers of graph structure: vertex layer, lane area layer, functional area layer, and intersection layer. Therefore, the element sequence part of the heterogeneous graph contains four different types of nodes: vertex layer nodes, lane area layer nodes, functional area layer nodes, and intersection layer nodes. Since different types of nodes contain different attributes, different types of element sequences are composed of a sequence of indefinite length:

[0085] Vertex: P node =(t node ,ID,type,x,y,z). node Represents a sequence of elements of a vertex layer node. node Indicates that the node is a vertex layer node, ID indicates the index value of the node, type indicates the properties of the node, which are divided into two categories: native vertices and secondary vertices, and x, y, z indicate the three-dimensional coordinates of the node. Native vertices refer to corner points extracted from the city planning map, which must be included in the heterogeneous graph; secondary vertices refer to nodes manually added by users or generated by programs, which indicate the entrances and exits of functional areas, etc. They are guiding information for generating models and may not be included in the heterogeneous graph. It should be understood that any vertex layer node (or vertex) in the heterogeneous graph can be represented by P node A vertex layer node represents a vertex.

[0086] Lane Area: P road Represents a sequence of elements of a lane area node. road Indicates that the node is a lane area node, ID indicates the index value of the node, type indicates the attributes of the node, including exit and entrance, etc., x, y, z indicate the three-dimensional coordinates of the center point of the node, represents the index of the vertex layer node constituting the lane area, direction represents the position of the entrance and exit of the lane area relative to the center point of the node, and area represents the area of ​​the lane area. It should be understood that any lane area node (or lane area) in the heterogeneous graph can be represented by P road A lane area node represents a lane area.

[0087] Functional area: P function Represents a sequence of elements of a function area node. functionIndicates that the node is a functional area node, ID indicates the index value of the node, type indicates the attributes of the node, including multiple types such as including exit and entrance, and x, y, z indicate the three-dimensional coordinates of the center point of the node. P represents the index of the vertex layer node that constitutes the functional area, direction represents the position of the entrance and exit of the node relative to the center point of the node, and area represents the area of ​​the functional area. It should be understood that any functional area node (or functional area) in the heterogeneous graph can be represented by P function A functional area node represents a functional area.

[0088] Intersection: P junction Represents a sequence of elements of an intersection node. Among them, t junction Indicates that the node is an intersection node, ID indicates the index value of the node, type indicates the attributes of the node, including three-way intersection, four-way intersection, etc., x, y, z indicate the three-dimensional coordinates of the center point of the node, represents the index of the vertex layer node constituting the intersection, and area represents the area of ​​the intersection. It should be understood that any intersection node (or intersection) in the heterogeneous graph can be represented by P junction An intersection node represents an intersection.

[0089] In the constraint sequence of a heterogeneous graph, since the heterogeneous graph contains two types of connections, namely, intra-layer connection relationship and inter-layer connection relationship, the constraint sequence of a heterogeneous graph can contain two types. Different types of constraint sequences contain different types of attributes, so the constraint sequence of a heterogeneous graph sequence consists of indefinite length sequences:

[0090] Intra-layer connection relationship: C inside =(t inside ,ID1,ID2,type). C inside Indicates the connection relationship between two nodes belonging to the same layer of graph structure in a heterogeneous graph, such as the connection relationship between two vertex layer nodes or the connection relationship between two lane area nodes. inside Indicates that the type of the connection relationship is an intra-layer connection relationship, ID1 and ID2 indicate the indexes of the two intra-layer nodes connected by the connection relationship, and type indicates the connection relationship type of the connection relationship, such as adjacent and connected relationships in the vertex layer. Two intra-layer nodes refer to two nodes belonging to the same layer of the graph structure. In other words, two intra-layer nodes refer to two nodes of the same type.

[0091] Inter-layer connection relationship: C pointer =(t pointer ,start,end,type). C pointerIndicates the connection relationship between two nodes belonging to different layers of graph structure in a heterogeneous graph, such as the connection relationship between the vertex layer node and the lane area node. pointer Indicates that the connection relationship is an inter-layer pointing relationship, start indicates the index of the starting node of the inter-layer pointing relationship, end indicates the index of the ending node of the inter-layer pointing relationship, and type indicates the connection relationship type of the connection relationship. For example, there is an inter-layer pointing relationship between the lane area and the functional area, where the lane area is connected to the functional area through an entrance and exit, wherein the starting node of the inter-layer pointing relationship is the lane area node, and the ending node of the inter-layer pointing relationship is the functional area node.

[0092] In one possible implementation, the layers (graph structure) of the heterogeneous graph are divided into a vertex layer, an intersection layer, a lane area layer, and a functional area layer. The nodes are assigned corresponding types, attributes, or numbers of parameters according to the urban planning map, and the intra-layer connection relationship and the inter-layer connection relationship are constructed, so as to realize the comprehensive expression of the traffic elements and surrounding facilities in the urban planning map, that is, to generate a heterogeneous graph that comprehensively expresses the traffic elements and surrounding facilities in the urban planning map.

[0093] In a possible implementation, the map generation device generates a heterogeneous graph sequence (i.e., a traffic expression graph sequence) by traversing the nodes of each layer of the stacked hierarchical graph (i.e., a heterogeneous graph), the connection relationships within each layer, and the pointing relationships between each layer. Since the number of attributes of the nodes in each layer is different and the attribute lengths of the nodes in different layers are different, the sequence finally obtained by encoding the stacked hierarchical graph is a variable-length sequence, that is, the heterogeneous graph sequence includes sequences of different lengths. In this application, the symbol Λ is defined as the starting symbol of the sequence and each element, and Ω is the ending symbol of the sequence, thereby representing a variable-length sequence. Exemplarily, the map generation device traverses each node in the order of vertex layer, lane area layer, functional area layer, and intersection layer; then, traverses the connection relationships within each layer in the same order; finally, traverses the pointing relationships between each layer in the order of intersection-lane area, intersection-functional area, and lane area-functional area.

[0094] By encoding and describing the nodes of each layer, the connection relationship within the layer, and the pointing relationship between the layers, the stacked hierarchical graph is traversed, and finally an example of a sequence representation (i.e., a heterogeneous graph sequence) obtained by encoding the stacked hierarchical graph is obtained as follows:

[0095]

[0096] Among them, before the first Ω is the sequence obtained by encoding the nodes (primitive) of each layer (i.e., the element sequence), and between the two Ω is the sequence obtained by encoding the constraints (constraint sequence). The whole constitutes a variable-length sequence representation obtained by encoding the stacked hierarchical graph. In other words, before the first Ω is the element sequence of each node in the heterogeneous graph, and between the two Ω is the constraint sequence obtained by encoding each connection relationship in the heterogeneous graph (including intra-layer connection relationship and inter-layer pointing relationship). It should be understood that the element attributes of each layer of the heterogeneous graph (i.e., the attribute information of each node) are serialized, as well as the intra-layer constraints (i.e., intra-layer connection relationship) and inter-layer constraint relationship (i.e., inter-layer connection relationship) are serialized, and the sequence combination is used as a serialized expression of the heterogeneous graph, i.e., the heterogeneous graph sequence.

[0097] The definition of a high-precision map sequence can also be divided into two parts: an element sequence and a constraint sequence. This application proposes a highly compatible high-precision map sequence encoding rule by referring to various high-precision map formats such as OpenDrive, LaneLet, and NuScenes. OpenDrive is a descriptive file of the road network structure. The NuScenes dataset is a large-scale autonomous driving dataset established by the autonomous driving company nuTonomy. In one possible implementation, the element sequence of the high-precision map includes 9 categories: roads, lanes, tunnels, bridges, intersections, traffic signs, traffic markings, traffic lights, and road accessories, where each element is represented by a set of indefinite-length sequences according to the type:

[0098] N=(t,p); (2)

[0099] t represents the element (primitive) type, and p represents the parameter set of this type of primitive. According to the parameter description (or attribute information) of different types of primitives, the primitives of the high-precision map can be encoded as a sequence: each value represents the parameter value in a primitive, and can also contain the start symbols of different traffic elements: r (road), l (lane), j (intersection), a (traffic light), s (traffic sign), b (bridge), t (tunnel), m (traffic marking), o (road attachment); the stop symbol e of the primitive and the stop symbol d of the entire sequence. Primitives and parameters are arranged in a fixed order. An example encoding sequence for a road is as follows: id1, type1, speed1, lanesCount1, controlPoints1, e, d.

[0100] The constraint sequences of high-precision maps are divided into two categories: lane connection relationships between roads and lane connection relationships within intersections. Similarly, the constraint sequences corresponding to different types of connection relationships contain different attributes, which are represented by a set of indefinite length sequences:

[0101] R=(N i ,L a ,N j ,L b ); (3)

[0102] The sequence of formula (3) is represented as the lane with ID a in the road with ID (index value) i, connected to the lane with ID b in the road with ID j. Traffic elements are encoded as a sequence according to their adjacency. Each value in the constraint sequence represents the index of a primitive, and may also contain the start symbol of the connection relationship between different elements: r (lane connection relationship between roads), j (lane connection relationship in intersections), the connection relationship stop symbol e and the stop symbol d of the entire sequence. Lane connection relationships (connection) can be sorted from large to small according to primitive index. Taking two adjacent two-way single-lane roads as an example, the constraint sequence is represented as: r,n1,1,n2,1,e,n1,-1,n2,-1,e,d. Among them, r,n1,1,n2,1,e means that the lane with ID 1 in the road with ID (index value) 1 is connected to the lane with ID 1 in the road with ID 2, and n1,-1,n2,-1 means that the lane with ID -1 in the road with ID (index value) 1 is connected to the lane with ID -1 in the road with ID 2.

[0103] In one possible implementation, a high-precision map sequence is generated by combining an element sequence and a constraint sequence, wherein the element sequence is arranged in a fixed order of roads, lanes, tunnels, bridges, intersections, traffic signs, traffic markings, traffic lights, and road accessories, and the constraint sequence is arranged in a fixed order of lane connection relationships between roads and lane connection relationships in intersections. It should be understood that both the element sequence and the constraint sequence can be sorted in other fixed orders, and this application does not limit this. In other words, attribute extraction and serialization encoding are performed on different elements in the high-precision map, and serialization encoding of constraints between elements is constructed, and the sequence combination is used as a serialized expression of the high-precision map, so as to obtain a high-precision map sequence.

[0104] Based on the above definition and description of the sequence representation of heterogeneous graphs and the sequence representation of electronic maps (such as high-precision maps), the present application proposes a deep learning method architecture (i.e., a map generation model) for sequence-to-sequence generation of high-precision maps based on a data-driven approach. In other words, the deep learning method architecture (i.e., the map generation model) provided by the present application takes a heterogeneous graph sequence as input and a high-precision map sequence as output. The map generation device can ultimately process the high-precision map sequence into high-precision map data according to the encoding rules of the high-precision map.

[0105] The following first describes the training sample dataset used to train the map generation model. The input of the map generation model proposed in this application is a heterogeneous graph sequence (i.e., a traffic map expression sequence), such as functional areas (land use types), roads, and intersection information in a city planning map, while the high-precision map sequence contains more detailed road network information such as roads, lanes, and road accessories. Currently, no single open source dataset can be directly used to train the map generation model proposed in this application. Therefore, this application proposes a method for constructing a fused dataset based on the NuScenes dataset and OpenStreetMap data.

[0106] NuScenes dataset is a large autonomous driving dataset, which released a high-precision map expansion package in 2019. The map contains 11 semantic layers, including crosswalks, sidewalks, traffic lights, stop lines, lanes, etc., which meet the high-precision map data content required by this solution. The map contains semantic vector maps (json format) and corresponding PNG formats of 4 maps: Boston Harbor, Queenstown Singapore, North Singapore, and Holland Village Singapore. The OpenStreetMap (OSM) project is a well-known global roadmap production example of a volunteered geographic information (VGI) project with a large number of voluntary participants. The project provides urban planning data covering the world, including land use types, roads, etc. Although it does not contain high-precision map information such as specific lanes and road attachments, it contains the data required to construct heterogeneous graphs in this application.

[0107] This application can use the NuScenes dataset as the main data set, and based on the description, longitude, and contour information of its four maps (Boston Harbor, Queenstown Singapore, Singapore North, and Singapore Holland Village), find the corresponding data in the OpenStreetMap data, and align and fuse them based on the roads to form a fused data set in json format.

[0108] Registration and fusion process: The nuScenes dataset contains high-precision map (road) data required for heterogeneous maps, and the OpenStreetMap dataset contains functional area data required for heterogeneous maps, so registration is required to obtain a fused dataset for training. A possible registration and fusion process is as follows:

[0109] (1) Download data: traverse the longitude and latitude of all nodes in nuScenes, take the boundary value, and download the data in the corresponding area of ​​OpenStreetMap;

[0110] (2) Traversing functional areas: Traverse and download all functional area data in OpenStreetMap data, and the longitude and latitude coordinates of all its nodes;

[0111] (3) Constructing a functional area semantic layer in nuScenes data: Construct a functional area semantic layer and assign a unique token value to each functional area. The functional area type is the type in OpenStreetMap. Traverse the lane topology and longitude and latitude coordinates in nuScenes. For each closed / semi-closed area consisting of lanes that contains a functional area, each lane will reference the functional area token.

[0112] (4) Post-processing: Perform visualization and manually resolve data conflicts.

[0113] The fused data set contains the land use type, coordinates and other information required for the heterogeneous graph sequence (i.e., traffic map expression sequence), as well as detailed road network information such as roads, lanes, and road accessories required for high-precision maps. For example, the model training device is based on the NuScenes data set, and downloads the corresponding data from the global OpenStreetMap data through the description, road contour, and longitude and latitude information of the NuScenes data; obtains the functional area and land use type data through the OpenStreetMap data, adds the "function" (functional area) field to the NuScenes data, and stores the data by referencing all the roads that enclose the closed area, thereby obtaining a fused data set. Figure 3 This is an example of a high-precision map visualization screenshot corresponding to the city planning map. The fused dataset is obtained by registering and fusion of the OpenStreetMap dataset and the NuScenes dataset. The fused dataset contains both the land use type and coordinate information required for the heterogeneous graph sequence and the detailed road network information such as roads, lanes, and road accessories required for the high-precision map.

[0114] After the training sample dataset is constructed, the map generation model can be trained. A possible training process is as follows:

[0115] (1) Model parameter initialization, that is, initializing the parameters of the map generation model;

[0116] (2) Input the heterogeneous graph sequence in the dataset into the generative model and forward calculate the output map sequence;

[0117] (3) Obtain the loss function of the map sequence calculation value and the label value (ground truth) and calculate the gradient vector;

[0118] (4) Adjust the parameters of the map generation model through the gradient vector so that the loss function is adjusted to a decreasing trend;

[0119] (5) Repeat the above process until the loss function reaches the set value or no longer decreases.

[0120] The entire training process of the map generation model can adopt the teacher forcing mechanism, that is, in the process of training the network, the label value of the training data corresponding to the previous item is directly used as the input of the next state.

[0121] In the present application, the map generation model may adopt the Transformer architecture. The Transformer architecture is a sequence-to-sequence deep learning generation model architecture. In the present application, the map generation model may adopt the encoder-decoder architecture, in which the encoder and decoder are spliced ​​using the same six-layer architecture. In the input embedding step, the present application adopts input embedding in units of nodes and constraints, that is, each node of each layer, each constraint in the intra-layer constraint, and each constraint in the inter-layer constraint are respectively regarded as a unified unit, mapped to a vector of equal length, and then position encoded, and then used as the input of the encoder. The above describes the input (heterogeneous graph sequence) and output (map sequence) of the map generation model, as well as how to construct a training sample data set for training the map generation model. For those skilled in the art, any sequence-to-sequence model can be used to train the map generation model, which will not be described in detail here. In other words, the map generation model can be finally obtained by end-to-end training of the input and output using a sequence-to-sequence model.

[0122] The map generation method provided by this application is introduced below in conjunction with the accompanying drawings.

[0123] Figure 4 A flowchart of a map generation method provided in an embodiment of the present application. Figure 4 As shown, the method includes:

[0124] 401. A map generating device obtains a traffic expression map.

[0125] The map generating device may be a terminal device with data processing capabilities, such as a tablet computer, a laptop computer, a desktop computer, or the like, or may be a cloud server, a network server, an application server, or the like.

[0126] The above traffic expression graph is a heterogeneous graph. The above traffic expression graph includes at least two layers of graph structures, and the at least two layers of graph structures include a first layer of graph structure and a second layer of graph structure. The type of nodes in the above first layer of graph structure is different from the type of nodes in the above second layer of graph structure. For example, the nodes in the first layer of graph structure are the above vertex layer nodes, and the vertices in the second layer of graph structure are intersection nodes.

[0127] The above traffic expression diagram can be used to express static traffic scenes. Here, static traffic scenes can be understood as traffic scenes in a static state, that is, the states of vehicles, pedestrians, traffic lights, etc. do not change. Static traffic scenes can include road scenes, tunnel scenes, bridge scenes, intersection scenes, and any other scenes involving traffic conditions.

[0128] The first-layer graph structure may be any one of the above-mentioned vertex layer, lane area layer, functional area layer, and intersection layer, and the second-layer graph structure may be any one of the above-mentioned vertex layer, lane area layer, functional area layer, and intersection layer that is different from the first-layer graph structure. For example, the first-layer graph structure is a vertex layer, and the nodes in the first-layer graph structure are all vertex layer nodes, and the second-layer graph structure is a lane area layer, and the nodes in the second-layer graph structure are all lane area nodes. For another example, the first-layer graph structure is a vertex layer, and the nodes in the first-layer graph structure are all vertex layer nodes, and the second-layer graph structure is an intersection layer, and the nodes in the second-layer graph structure are all intersection nodes. The above-mentioned traffic expression graph may also include other layers of graph structures. In other words, the traffic expression graph obtained by the map generation device includes two or more layers of graph structures, that is, two or more layers of graph structures are fused into a heterogeneous graph. Exemplarily, the traffic expression graph includes a vertex layer, an intersection layer, a lane area layer, and a functional area layer. Exemplarily, the traffic expression graph includes three layers of graph structures, namely: a vertex layer, an intersection layer, and a functional area layer. Exemplarily, the traffic expression graph includes three layers of graph structure, namely: vertex layer, lane area layer, and functional area layer. Exemplarily, the traffic expression graph includes two layers of graph structure, namely: vertex layer and lane area layer. It should be noted that there is only one type of node in each layer of the graph structure of the traffic expression graph. Exemplarily, the nodes in the vertex layer are all vertex layer nodes, the nodes in the lane area layer are all lane area nodes, the nodes in the intersection layer are all intersection nodes, and the nodes in the functional area layer are all functional area nodes.

[0129] In a possible implementation, the traffic expression graph also includes a third-layer graph structure and a fourth-layer graph structure, and the types of nodes in any two layers of the first-layer graph structure, the second-layer graph structure, the third-layer graph structure, and the fourth-layer graph structure are different. In other words, the traffic expression graph includes at least two layers of graph structures mentioned above, and also includes the third-layer graph structure and the fourth-layer graph structure. Exemplarily, the first-layer graph structure is a vertex layer, the second-layer graph structure is a lane area layer, the third-layer graph structure is an intersection layer, and the fourth-layer graph structure is a functional area layer.

[0130] In a possible implementation, each node in the first-layer graph structure belongs to a first type (e.g., vertex layer nodes), each node in the second-layer graph structure belongs to a second type (e.g., function area nodes), and the first type is different from the second type. Both the first-layer graph structure and the second-layer graph structure are isomorphic graphs.

[0131] A possible implementation of step 401 is as follows: the map generation device constructs a first-layer graph structure and a second-layer graph structure according to the urban planning information; determines (defines) the connection relationship between the nodes in the first-layer graph structure and the nodes in the second-layer graph structure to obtain a traffic expression map. Exemplarily, the urban planning information is an urban planning map (including the corresponding urban planning legend analysis). That is, the map generation device can construct a traffic expression map according to the urban planning information.

[0132] An example of a map generation device constructing a traffic expression map based on urban planning information is as follows: the map generation device extracts multiple vertices in the urban planning map corresponding to the urban planning information, and constructs a vertex layer graph structure by defining (determining) the connection relationship between each vertex; based on the vertex layer graph structure, extracts intersections, lane areas, functional areas and other elements (or nodes, such as intersection nodes, lane area nodes, functional area nodes) from the urban planning map, and constructs an intersection layer graph structure, a lane area layer graph structure, and a functional area layer graph structure by respectively defining (determining) the connection relationship between each element of the same type; by defining the connection relationship between nodes of different layer graph structures, each layer graph structure is integrated in the form of "stacking" to form a "heterogeneous graph" expression of the urban planning map. The vertices in the urban planning map can be understood as the points where two or more line segments intersect. Figure 5A This is an example of a vertex layer graph structure provided in an embodiment of the present application. Figure 5A As shown, n1 to n 31 They represent a primary vertex or a secondary vertex respectively. The solid line between two vertices represents a connected relationship, and the dotted line between two vertices represents an adjacent relationship. Figure 5B This is an example of a lane area layer map structure provided in an embodiment of the present application. Figure 5B As shown, r1 to r7 represent a lane area node respectively, and the line between any two lane area nodes represents the connection relationship between the two lane area nodes. Different lines can represent different types of connection relationships. Figure 5C This is an example of an intersection zone layer diagram structure provided in an embodiment of the present application. Figure 5C As shown, J1, J2, and J3 represent an intersection node respectively, and the line between two intersection nodes represents the connection relationship between the two intersections. The intersection layer graph structure can contain multiple different types of intersection nodes. Figure 5D This is an example of a functional area layer diagram structure provided in an embodiment of the present application. Figure 5DAs shown in the figure, f1, f2, f3, f4, f5, and f6 represent a functional area node respectively, and the line between two functional area nodes represents the connection relationship between the two functional areas. The functional area layer graph structure can contain functional area nodes of various land use types, such as public green space, square land, commercial and financial land, cultural and entertainment land, educational land, residential land, scientific and technological research and development land, commercial and office comprehensive land, etc. The functional area layer graph structure can contain various different types of connection relationships, such as connection through lanes, connection through intersections, etc.

[0133] 402. The map generating device obtains a traffic expression map sequence according to the traffic expression map.

[0134] The above traffic expression sequence is a serialized expression of the traffic expression map. The map generation device can obtain the traffic expression map sequence from the traffic expression map according to the above defined coding rules.

[0135] A possible implementation of step 402 is as follows: serialize the attribute information of one or more nodes (e.g., all nodes) in the above-mentioned traffic expression graph to obtain an element sequence; serialize the intra-layer connection relationship and the inter-layer connection relationship in the above-mentioned traffic expression graph to obtain a constraint sequence; the above-mentioned intra-layer connection relationship includes the connection relationship between any two nodes in the above-mentioned first-layer graph structure, and the above-mentioned inter-layer connection relationship includes the connection relationship between any node in the above-mentioned first-layer graph structure and any node in the above-mentioned second-layer graph structure; use the combination of the above-mentioned element sequence and the above-mentioned constraint sequence as the serialized expression of the above-mentioned traffic expression graph to obtain the above-mentioned traffic expression graph sequence. Exemplarily, the attribute information of each node in the above-mentioned traffic expression graph (e.g., including vertex layer nodes, intersection nodes, lane area nodes, and functional area nodes) is serialized to obtain an element sequence.

[0136] Since the definition of heterogeneous graph sequence (i.e., traffic expression graph) and the process of generating heterogeneous graph sequence (i.e., traffic expression sequence) based on heterogeneous graph have been described above (serialized representation of heterogeneous graph and electronic map), they will not be repeated here.

[0137] In a possible implementation, the traffic expression graph sequence includes an element sequence and a constraint sequence, wherein the element sequence represents the attribute information of the nodes in the traffic expression graph, and the constraint sequence represents the connection relationship between two nodes in the first-layer graph structure, and the connection relationship between the nodes in the first-layer graph structure and the nodes in the second-layer graph structure. It should be understood that the traffic expression graph sequence may include an element sequence corresponding to each node in the traffic expression graph, and a constraint sequence corresponding to each connection relationship.

[0138] 403. The map generating device obtains a map sequence according to the traffic expression graph sequence.

[0139] A possible implementation of step 403 is as follows: input the traffic expression graph sequence into the map generation model for processing to obtain a map sequence. The map generation model can be a sequence-to-sequence model obtained by training. The method of training the map generation model has been described above and will not be described here.

[0140] 404. The map generating device obtains an electronic map according to the map sequence.

[0141] The electronic map may be a high-precision map file. A possible implementation of step 404 is as follows: the map generation device converts the output map sequence into a map file in json format, ie, the electronic map, according to the map encoding rule (ie, the high-precision map sequence encoding rule described above).

[0142] In an embodiment of the present application, a traffic expression graph is obtained; based on the traffic expression graph, a traffic expression graph sequence is obtained; based on the traffic expression graph sequence, a map sequence is obtained. The form of a heterogeneous graph allows the graph to contain multiple types of nodes, and the attribute types and parameter numbers of the node descriptions may be different, so multiple types of elements in the static traffic scene can be comprehensively expressed, and more complex static traffic scenes can be effectively represented. In the map generation method provided in an embodiment of the present application, the static traffic scene is expressed through a traffic expression graph (a heterogeneous graph). Compared with expressing the static traffic scene in the form of an isomorphic graph (there is only one type of node in the isomorphic graph and the attribute types, parameter numbers, etc. of the node descriptions are the same), the various elements contained in the static traffic scene can be more accurately and effectively expressed in a unified and global manner, thereby providing more realistic information about the surrounding facilities of the road network, and having better scalability.

[0143] Figure 6 A flowchart of another map generation method provided in an embodiment of the present application. Figure 6 The method flow is Figure 4 A possible implementation of the method described in the embodiment of the present invention. In this implementation, a method for constructing a traffic expression map by a map generation device and a method for obtaining a sequence of traffic expression maps based on the traffic expression map are described. Figure 6 As shown, the method includes:

[0144] 601. The map generating device constructs a first-layer graph structure according to the city planning information, and constructs a second-layer graph structure according to the city planning information.

[0145] Each node in the above-mentioned first-layer graph structure belongs to the first type, each node in the above-mentioned second-layer graph structure belongs to the second type, and the above-mentioned first type is different from the above-mentioned second type. The map generation device can also construct other layers of graph structures according to urban planning information. Urban planning information can be an urban planning map. Exemplarily, the traffic expression map includes three layers of graph structures, namely: a first-layer graph structure, a second-layer graph structure, and a third-layer graph structure. For example, the first-layer graph structure, the second-layer graph structure, and the third-layer graph structure are respectively a vertex layer, an intersection layer, and a functional area layer. Exemplarily, the traffic expression map includes a vertex layer, an intersection layer, a lane area layer, and a functional area layer; the first-layer graph structure and the second-layer graph structure are any two layers of these four layers. The map generation device can construct a vertex layer, an intersection layer, a lane area layer, and a functional area layer according to urban planning information.

[0146] In a possible implementation, the first-layer graph structure is a vertex layer; the map generation device constructs the first-layer graph structure according to the urban planning information as follows: extracting corner points in the urban planning map corresponding to the urban planning information as vertices to obtain multiple vertices; determining the connection relationship between the multiple vertices to obtain the first-layer graph structure. The corner points in the urban planning map refer to the intersections of two or more straight lines in the urban planning map. Figure 5A , Figure 5A Each primary vertex in is a corner point in the city planning map. Based on the primary vertices, the map generation device can automatically add secondary vertices according to pre-configured rules, and can also support users to manually add secondary vertices.

[0147] The map generation device may construct the second-layer graph structure according to the urban planning information in the following manner: extract multiple traffic elements of the first type from the urban planning map corresponding to the urban planning information, wherein the traffic elements of the first type include any one of functional areas, intersections, and lane areas; determine the connection relationship between the multiple traffic elements to obtain the second-layer graph structure. For example, the map generation device traverses each traffic element of the second type (e.g., lane areas, functional areas, intersections) in the urban planning map; the map generation device constructs the second-layer graph structure by defining (determining) the connection relationship between each node of the second type.

[0148] 602. The map generating device determines the connection relationship between the nodes in the first layer graph structure and the nodes in the second layer graph structure to obtain a traffic expression graph.

[0149] A possible implementation of step 602 is as follows: multiple layers of graphs (e.g., the first layer graph structure and the second layer graph structure) are fused into a heterogeneous graph, i.e., a traffic expression graph, in a "stacked" form; a heterogeneous graph that expresses a static traffic scene can be accurately formed. Exemplarily, the map generation device can determine the connection relationship between nodes in different layers of graph structures, for example, the road connection relationship between the intersection layer nodes and the lane area nodes is defined, and the multiple layers of graphs are fused into a heterogeneous graph in a "stacked" form to form a heterogeneous graph expression model for static traffic scenes, see Figure 2 .

[0150] An example of a map generating device constructing a traffic expression map based on a city planning map corresponding to urban planning information is as follows: the city planning map input by the user is a partial area of ​​the city planning map of Singapore's One Tech City, which includes 8 lane areas, 2 intersections, 2 functional areas, and 20 native nodes (or called native vertices); the map generating device constructs a vertex layer, an intersection layer, a lane area layer, and a functional area layer according to the city planning map; by determining the connection relationship between nodes in different layers of the graph structure, the vertex layer, the intersection layer, the lane area layer, and the functional area layer are merged into a heterogeneous graph, namely, the traffic expression map. Figure 7 An example of a traffic expression diagram provided in an embodiment of the present application. Figure 7 The traffic expression map in the figure is a traffic expression map constructed by the map generating device based on part of the urban planning map of Singapore's One Tech City. Figure 7 In the figure, n represents the vertex layer node, j represents the intersection layer node, r represents the lane area node, and f represents the functional area layer node.

[0151] 603. The map generating device serializes the attribute information of each node in the traffic expression graph to obtain an element sequence.

[0152] The element sequence may include a coding sequence of the attribute information of each layer of nodes in the traffic expression graph. Serializing the attribute information of each node in the traffic expression graph can be understood as encoding the attribute information of each node in the traffic expression graph to obtain an element sequence representing the attribute information of the node. For example, P node Represents a sequence of elements of a vertex layer node, P road Represents a sequence of elements of a lane area node, P function Represents a sequence of elements of a functional area node, P junction represents an element sequence of an intersection node. It should be understood that the map generation device can node Represents any vertex-level node in the traffic expression graph, through P road Represents any lane node in the traffic expression graph, through P function Represents any functional area node in the traffic expression diagram, through P junctionrepresents any intersection node in the traffic expression graph. In other words, P node is the serialization of the attribute information of a vertex layer node, P function is the serialization of a function area node, P function is the serialization of a function area node, P junction is a serialization of an intersection node. Table 1 is an example of an element sequence provided in an embodiment of the present application.

[0153] Table 1

[0154]

[0155]

[0156] Each row in Table 1 is a sequence of elements of a node, where n represents a vertex layer node, j represents an intersection layer node, r represents a lane area node, f represents a functional area layer node, and Λ represents the starting symbol of the sequence and each element. The meaning of each parameter in Table 1 can be found in P node , P road , P function , P junction The meaning of each parameter in .

[0157] 604. The map generating device serializes the intra-layer connection relationship and the inter-layer connection relationship in the traffic expression graph to obtain a constraint sequence.

[0158] The intra-layer connection relationship in the above-mentioned traffic expression graph includes the connection relationship between two nodes in the first-layer graph structure. The inter-layer connection relationship in the above-mentioned traffic expression graph includes the connection relationship between the nodes in the above-mentioned first-layer graph structure and the nodes in the above-mentioned second-layer graph structure. The constraint sequence may include the encoding sequence of each layer connection relationship and each inter-layer connection relationship in the traffic expression graph. The map generation device serializes the intra-layer connection relationship and the inter-layer connection relationship in the traffic expression graph, which can be understood as encoding each intra-layer connection relationship and each inter-layer connection relationship in the traffic expression graph respectively to obtain the encoding sequence of each connection relationship. Exemplarily, the map generation device generates a constraint sequence through C inside It represents the connection relationship between two nodes in the same layer of the graph structure in the traffic expression graph, that is, the intra-layer connection relationship; through C pointer Indicates the connection relationship between two nodes belonging to different layers of graph structure in the traffic expression graph, that is, the inter-layer connection relationship. In one possible implementation, the map generation device traverses the connection relationship within each layer in the order of vertex layer, lane area layer, functional area layer, and intersection layer to obtain the coding sequence of the connection relationship within each layer; traverses the pointing relationship between each layer in the order of intersection-lane area, intersection-functional area, and lane area-functional area to obtain the coding sequence of the connection relationship between each layer. Table 2 is an example of the constraint sequence provided in the embodiment of the present application.

[0159] Table 2

[0160]

[0161]

[0162] Each line in Table 2 represents a constraint sequence of a connection relationship, where n represents a vertex layer node, j represents an intersection layer node, r represents a lane area node, and f represents a functional area layer node. Λ is the starting symbol of the sequence and each connection relationship. The meaning of each parameter in Table 2 can be found in C inside and C pointer The meaning of each parameter in .

[0163] 605. The map generating device uses the combination of the element sequence and the constraint sequence as a serialized expression of a traffic expression graph to obtain a traffic expression graph sequence.

[0164] The element sequence is the element part of the traffic expression graph sequence, and the constraint sequence is the constraint part of the traffic expression graph sequence. An example of a traffic expression graph sequence is shown in formula (1).

[0165] 606. The map generation device inputs the traffic expression diagram sequence into the map generation model for processing to obtain a map sequence.

[0166] In a possible implementation, the map sequence includes a first element sequence and a second element sequence, the length of the first element sequence is different from the length of the second element sequence, the first element sequence represents a first traffic element, the second element sequence represents a second traffic element, the type of the first traffic element is different from the type of the second traffic element, and the first traffic element and the second traffic element correspond to traffic elements in the static traffic scene. The first traffic element and the second traffic element can be any two of roads, lanes, tunnels, bridges, intersections, traffic signs, traffic markings, traffic lights, and road accessories. For example, the first traffic element is an intersection, and the second traffic element is a road, and the intersection and the road belong to different types of traffic elements. In this example, the first traffic element can be represented by a sequence of length M, and the second traffic element can be represented by a sequence of length N, where M and N are integers greater than 0, and M is different from N. The first traffic element can be represented by a sequence of length M, which can be understood as the first traffic element can be represented by M groups of parameters. For example, the first traffic element (i.e., intersection) can be represented by four groups of parameters, each group of parameters includes one or more parameters, and the four groups of parameters are: ID (identification), type (type), speed (speed), Red_road (internal road of the intersection); wherein, ID is the identification of the intersection, type represents the type of the intersection, speed represents the speed limit of the intersection, and Red_road represents the identification (or index) of the road belonging to the intersection. In the intersection definition, the ID of the internal road of the intersection can be referenced in Ref_road to indicate that it belongs to the intersection road. Table 3 shows an example of an intersection in a high-precision map, and each of the second to fifth columns in Table 3 represents a group of parameters. Referring to Table 3, the first traffic element (i.e., intersection) can be represented by a sequence containing four groups of parameters, that is, it can be represented by a sequence of length four.

[0167] Table 3

[0168] Intersection id type speed Ref_road <![CDATA[j1]]> 1 1 60 [5,6,7,8,9,10,11,12] <![CDATA[j2]]> 2 1 60 [13,14,15,16,17,18,19,20]

[0169] For example, the second traffic element (i.e., road) can be represented by sixteen groups of parameters or more, each group of parameters includes one or more parameters, and the sixteen groups of parameters are: Λ (starting symbol), ID, type (type), speed (speed), leftlanes (left lane), rightlanes (right lane), width (width), controlpoints (control points), road marking starting symbol, type_line (marking type), loc1 (position), road attachment starting symbol, type_obj1 (road attachment type), loc2 (position), type_obj2 (road attachment type), loc3 (position); wherein Λ represents the road The start symbol of , ID is the identifier of the road, type indicates the type of the road, speed indicates the speed limit of the road, leftlanes indicates the number of left lanes in the road, rightlanes indicates the number of right lanes in the road, width indicates the width of each lane in the road, controlpoints is a set of coordinates indicating the center point of the road, type_line refers to the position of the road marking in the road, loc1 indicates the position of the road marking, type_obj1 indicates the type of road appendage 1, loc2 indicates the position of road appendage 1, type_obj2 indicates the type of road appendage 2, and loc3 indicates the position of road appendage 2. Control points, that is, the way to indicate the position and shape of the road in the OpenDrive data format, are a set of coordinates of the center point of the road. The number of coordinates of the center point is at least 2. Connecting the center points in sequence can obtain the specific shape description of the road. type_line and loc correspond to each other one by one, representing the type and position of the road marking respectively. Road markings mainly include yellow solid lines, yellow dashed lines, white solid lines and other types. loc indicates the lane to which the marking belongs. The default marking position is the right position of the lane to which it belongs, the marking shape is the same as the lane shape, and the marking width is the preset default value. The reference of loc to the lane to which it belongs is defined according to the OpenDrive standard, where 0 represents the center lane, 1, 2, ... represents the left lane from the center to the edge, and -1, -2, ... represents the right lane from the center to the edge. type_obj corresponds to loc one by one, representing the type and position of the road attachment respectively. Road attachments mainly include traffic lights, signs and other types. The definition of loc can be described using the st coordinate system in OpenDrive to indicate the relative position of the road attachment on the road. In this example, the second traffic element (i.e., the road) can be represented by a sequence containing sixteen groups of parameters, that is, a sequence of length sixteen. It should be understood that the second traffic element (i.e., the road) can be represented by more or fewer groups of parameters. In other words, different roads can be represented by sequences of different lengths. Table 4 shows an example of a road in a high-precision map.In Table 4, r1, r2, r3, and r4 represent four different roads. Each of the second to seventeenth columns in Table 4 represents a set of parameters. Referring to Table 4, the second traffic element (i.e., the road) can be represented by a sequence containing sixteen sets of parameters, that is, a sequence with a length of sixteen.

[0170] Table 4

[0171]

[0172] The first element sequence and the second element sequence can be N=(t,p), where t represents the element (primitive) type and p represents the parameter set of the primitive type. A map sequence may include multiple element sequences, and different types of elements can be represented by a set of indefinite length sequences. In other words, N=(t,p) can represent different types of traffic elements. In this implementation, the lengths of the first element sequence and the second element sequence are different, and different types of traffic elements are represented by indefinite length sequences, which can accurately represent different types of traffic elements.

[0173] In a possible implementation, the map sequence further includes a first constraint sequence and a second constraint sequence. The length of the first constraint sequence is different from the length of the second constraint sequence. The first constraint sequence represents the connection relationship between the third traffic element and the fourth traffic element. The second constraint sequence represents the connection relationship between the fifth traffic element and the sixth traffic element. The third traffic element, the fourth traffic element, the fifth traffic element, and the sixth traffic element correspond to the traffic elements in the static traffic scene. The first constraint sequence and the second constraint sequence can be represented by R=(N i ,L a ,N j ,L b ) means, R=(N i ,L a ,N j ,L b ) indicates that the lane with ID a in the road with ID (index value) i is connected to the lane with ID b in the road with ID j. A map sequence can include multiple constraint sequences, and different types of connection relationships can be represented by a set of indefinite length sequences. In other words, R = (N i ,L a ,N j ,L b ) can represent different types of connection relationships. In this implementation, the lengths of the first constraint sequence and the second constraint sequence are different, and different types of connection relationships are represented by sequences of indefinite length, which can accurately represent different types of connection relationships.

[0174] 607. The map generating device obtains an electronic map according to the map sequence.

[0175] Step 607 may refer to step 404. The map generating device may display the processed electronic map through visualization software. Figure 8 An example of a visualization result of an electronic map provided in an embodiment of the present application.

[0176] The map generation method provided in the embodiment of the present application expresses the static traffic scene through a traffic expression graph (i.e., a heterogeneous graph). Compared with expressing the static traffic scene in the form of a homogeneous graph (there is only one type of node in the homogeneous graph and the attribute type, parameter number, etc. of the node description are the same), it can more accurately and effectively express the various elements contained in the static traffic scene in a unified and global manner, thereby providing more realistic road network surrounding facility information and having better scalability. In other words, the heterogeneous graph expression of the urban planning map, in addition to supporting the basic elements, semantic attributes and complex constraint relationships in the traffic scene, also provides richer and clearer traffic-related road network scene information. The map generation device can generate high-precision map data for autonomous driving simulation conveniently and efficiently based on the input of the real urban planning map. The map generation method provided in the embodiment of the present application provides a rich information on surrounding facilities outside the road network, so it can provide a reference for generating more realistic traffic flow scenes. In addition, the functional area description information in the heterogeneous graph can be used as a basis for the 3D rendering simulator to generate real city scenes, thereby constructing a high-quality end-to-end simulation.

[0177] The structure of a map generating device that can implement the map generating method provided in the embodiment of the present application is described below with reference to the accompanying drawings.

[0178] Fig. 9 This is a schematic diagram of the structure of a map generation device provided in an embodiment of the present application. Fig. 9 As shown, the map generating device includes:

[0179] An acquisition unit 901 is used to acquire a traffic expression graph, where the traffic expression graph is used to express a static traffic scene, and the traffic expression graph includes at least two layers of graph structures, where the at least two layers of graph structures include a first layer of graph structure and a second layer of graph structure, and the type of nodes in the first layer of graph structure is different from the type of nodes in the second layer of graph structure;

[0180] The encoding unit 902 is used to obtain a traffic expression graph sequence according to the traffic expression graph, where the traffic expression graph sequence is a serialized expression of the traffic expression graph;

[0181] The processing unit 903 is used to obtain a map sequence according to the traffic expression diagram sequence; and obtain an electronic map according to the map sequence.

[0182] In a possible implementation, the encoding unit 902 is specifically used to serialize the attribute information of each node in the above-mentioned traffic expression graph to obtain an element sequence; serialize the intra-layer connection relationship and the inter-layer connection relationship in the above-mentioned traffic expression graph to obtain a constraint sequence; the intra-layer connection relationship in the above-mentioned traffic expression graph includes the connection relationship between two nodes in the above-mentioned first-layer graph structure, and the inter-layer connection relationship in the above-mentioned traffic expression graph includes the connection relationship between the nodes in the above-mentioned first-layer graph structure and the nodes in the above-mentioned second-layer graph structure; the combination of the above-mentioned element sequence and the above-mentioned constraint sequence is used as the serialized expression of the above-mentioned traffic expression graph to obtain the above-mentioned traffic expression graph sequence.

[0183] In a possible implementation, the acquisition unit 901 is specifically used to construct the above-mentioned first-layer graph structure and the above-mentioned second-layer graph structure according to urban planning information; the above-mentioned urban planning information is used to obtain the urban planning map; determine the connection relationship between the nodes in the above-mentioned first-layer graph structure and the nodes in the above-mentioned second-layer graph structure to obtain the above-mentioned traffic expression map.

[0184] In a possible implementation, the acquisition unit 901 is specifically used to extract multiple vertices in the urban planning graph corresponding to the above urban planning information; determine the connection relationship between the above multiple vertices, and obtain the above first-layer graph structure.

[0185] In a possible implementation, the acquisition unit 901 is specifically used to extract traffic elements belonging to the first type from the urban planning map corresponding to the above-mentioned urban planning information, and the above-mentioned first type of traffic elements includes: any one of functional areas, intersections, and lane areas; determine the connection relationship between the above-mentioned first type of traffic elements to obtain the above-mentioned second-layer graph structure.

[0186] In a possible implementation, each node in the first-layer graph structure belongs to a first type, each node in the second-layer graph structure belongs to a second type, and the first type is different from the second type.

[0187] In a possible implementation, the map generation device further includes: an input unit 904 for inputting the above-mentioned city planning map. For example, the user can input the city planning map to the map generation device through the input unit 904 (eg, including a keyboard, a mouse, a touch screen, etc.).

[0188] In a possible implementation, the map generation device further includes: an output unit 905, which is used to display a global city planning map; an input unit 904, which is used to input a part of the global city planning map selected by a user as the city planning map. For example, the user selects a part of the global city planning map provided by the map generation device as the city planning map through an input unit (such as a mouse, keyboard, etc.); the map generation device generates a high-precision map file based on the city planning map. Alternatively, the output unit 905 is used to output an electronic map, such as a high-precision map file.

[0189] In a possible implementation, the map generating apparatus further includes: a communication unit 906, configured to receive a city planning map sent by a user via a terminal device, and to send the electronic map to the terminal device.

[0190] Fig.10 This is a schematic diagram of the structure of a terminal device provided in an embodiment of the present application. Fig.10 As shown, the terminal device 100 includes a processor 1001, a memory 1002, and an input / output device 1003. The processor 1001, the memory 1002, and the input / output device 1003 are interconnected via a bus. Fig.10 The terminal device in may be the map generating device in the aforementioned embodiment.

[0191] The memory 1002 includes, but is not limited to, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), or portable read-only memory (CDROM), and is used for related instructions and data. The input and output device 1003 is used to input and output data.

[0192] The processor 1001 may be one or more central processing units (CPUs). When the processor 1001 is a CPU, the CPU may be a single-core CPU or a multi-core CPU. Fig.10The structure of the terminal device shown. In some embodiments, the processor 1001 can implement the functions of the acquisition unit 901, the encoding unit 902, and the processing unit 903; the input-output device 1003 can implement the functions of the input unit 904, the output unit 905, and the communication unit 906. For example, the input-output device 1003 includes a display, and the display can display a map and / or a global city planning map. For another example, the input-output device 1003 includes an input device such as a keyboard, a mouse, and a touch screen, and the input device is used to input a part of the global city planning map selected by the user as a city planning map. For another example, the input-output device 1003 includes a communication interface, which is used to receive city planning maps sent by other devices and send electronic maps to other devices.

[0193] Fig.11 1 is a schematic diagram of the structure of a server provided in an embodiment of the present application. The server 1100 may have relatively large differences due to different configurations or performances, and may include one or more central processing units (CPU) 1122 (for example, one or more processors) and memory 1132, and one or more storage media 1130 (for example, one or more mass storage devices) storing application programs 1142 or data 1144. Among them, the memory 1132 and the storage medium 1130 may be short-term storage or permanent storage. The program stored in the storage medium 1130 may include one or more modules (not shown in the figure), and each module may include a series of instruction operations on the server. Furthermore, the central processing unit 1122 may be configured to communicate with the storage medium 1130 to execute a series of instruction operations in the storage medium 1130 on the server 1100. The server 1100 may be the above-mentioned map generation device.

[0194] The server 1100 may also include one or more power supplies 1126, one or more wired or wireless network interfaces 1150, one or more input and output interfaces 1158, and / or one or more operating systems 1141, such as Windows ServerTM, Mac OS XTM, UnixTM, LinuxTM, FreeBSDTM, etc.

[0195] The steps performed by the map generation device in the above embodiment can be based on the Fig.11In some embodiments, the central processor 1122 can implement the functions of the acquisition unit 901, the encoding unit 902, and the processing unit 903; the input and output interface 1158 can implement the functions of the communication unit 906. For example, the input and output device 1003 includes a communication interface, which is used to receive a city planning map sent by other devices and send an electronic map to other devices.

[0196] In an embodiment of the present application, a computer-readable storage medium is provided. The computer-readable storage medium stores a computer program. When the computer program is executed by a processor, the map generation method provided in the above embodiment is implemented.

[0197] An embodiment of the present application provides a computer program product including instructions, which, when executed on a computer, enables the computer to execute the map generation method provided by the aforementioned embodiment.

[0198] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed in the present application, and these modifications or replacements should be included in the protection scope of the present application. Therefore, the protection scope of the present application shall be based on the protection scope of the claims.

Claims

1. A map generation method, characterized in that: include: Acquire a traffic expression graph, where the traffic expression graph is used to express a static traffic scene, where the traffic expression graph includes at least two layers of graph structures, where the at least two layers of graph structures include a first layer of graph structures and a second layer of graph structures, where a type of node in the first layer of graph structures is different from a type of node in the second layer of graph structures; According to the traffic expression graph, a traffic expression graph sequence is obtained, wherein the traffic expression graph sequence is a serialized expression of the traffic expression graph, and the traffic expression graph sequence includes an element sequence and a constraint sequence, wherein the element sequence represents attribute information of nodes in the traffic expression graph, and the constraint sequence represents a connection relationship between any two nodes in the first-layer graph structure, and a connection relationship between any node in the first-layer graph structure and any node in the second-layer graph structure; According to the traffic expression graph sequence, a map sequence is obtained, wherein the map sequence is a serialized expression of an electronic map; The electronic map is obtained according to the map sequence.

2. The method according to claim 1, characterized in that The step of obtaining a traffic expression graph sequence according to the traffic expression graph comprises: Serializing the attribute information of one or more nodes in the traffic expression graph to obtain an element sequence; The intra-layer connection relationship and the inter-layer connection relationship in the traffic expression graph are serialized to obtain a constraint sequence; the intra-layer connection relationship includes a connection relationship between any two nodes in the first-layer graph structure, and the inter-layer connection relationship includes a connection relationship between any node in the first-layer graph structure and any node in the second-layer graph structure; The combination of the element sequence and the constraint sequence is used as a serialized expression of the traffic expression graph to obtain the traffic expression graph sequence.

3. The method according to claim 1 or 2, characterized in that: The obtaining of the traffic expression diagram comprises: Constructing the first-layer graph structure and the second-layer graph structure according to the city planning information; the city planning information is used to obtain a city planning map; Determine the connection relationship between the nodes in the first layer graph structure and the nodes in the second layer graph structure to obtain the traffic expression graph.

4. The method according to claim 3, characterized in that The constructing the first-layer graph structure according to the urban planning information includes: Extracting corner points in the urban planning map corresponding to the urban planning information as vertices to obtain a plurality of vertices; Determine the connection relationship between the multiple vertices to obtain the first layer graph structure.

5. The method according to claim 3, characterized in that: The constructing the second-layer graph structure according to the urban planning information includes: Extracting a first type of traffic element from the urban planning map corresponding to the urban planning information, where the first type of traffic element includes: any one of a functional area, an intersection, and a lane area; Determine the connection relationship between the first type of traffic elements to obtain the second layer graph structure.

6. The method according to claim 1 or 2, characterized in that: The traffic expression graph also includes a third-layer graph structure and a fourth-layer graph structure, and the types of nodes in any two layers of the first-layer graph structure, the second-layer graph structure, the third-layer graph structure and the fourth-layer graph structure are different.

7. The method according to claim 1 or 2, characterized in that: The map sequence includes a first element sequence and a second element sequence, the length of the first element sequence is different from the length of the second element sequence, the first element sequence represents a first traffic element, the second element sequence represents a second traffic element, the type of the first traffic element is different from the type of the second traffic element, and the first traffic element and the second traffic element correspond to traffic elements in the static traffic scene.

8. The method according to claim 7, characterized in that The map sequence also includes a first constraint sequence and a second constraint sequence, the length of the first constraint sequence is different from the length of the second constraint sequence, the first constraint sequence represents a connection relationship between a third traffic element and a fourth traffic element, the second constraint sequence represents a connection relationship between a fifth traffic element and a sixth traffic element, and the third traffic element, the fourth traffic element, the fifth traffic element, and the sixth traffic element correspond to traffic elements in the static traffic scene.

9. A map generating device, characterized in that: include: An acquisition unit, used for acquiring a traffic expression graph, wherein the traffic expression graph is used for expressing a static traffic scene, wherein the traffic expression graph has at least two layers of graph structures, wherein the at least two layers of graph structures include a first layer of graph structure and a second layer of graph structure, and the type of nodes in the first layer of graph structure is different from the type of nodes in the second layer of graph structure; An encoding unit, configured to obtain a traffic expression graph sequence according to the traffic expression graph, wherein the traffic expression graph sequence is a serialized expression of the traffic expression graph, and the traffic expression graph sequence includes an element sequence and a constraint sequence, wherein the element sequence represents attribute information of nodes in the traffic expression graph, and the constraint sequence represents a connection relationship between any two nodes in the first-layer graph structure, and a connection relationship between any node in the first-layer graph structure and any node in the second-layer graph structure; The processing unit is used to obtain a map sequence according to the traffic expression map sequence, wherein the map sequence is a serialized expression of an electronic map; and obtain the electronic map according to the map sequence.

10. The device according to claim 9, characterized in that The encoding unit is specifically used to serialize the attribute information of one or more nodes in the traffic expression graph to obtain an element sequence; The intra-layer connection relationship and the inter-layer connection relationship in the traffic expression graph are serialized to obtain a constraint sequence; the intra-layer connection relationship includes a connection relationship between any two nodes in the first-layer graph structure, and the inter-layer connection relationship includes a connection relationship between any node in the first-layer graph structure and any node in the second-layer graph structure; The combination of the element sequence and the constraint sequence is used as a serialized expression of the traffic expression graph to obtain the traffic expression graph sequence.

11. The device according to claim 9 or 10, characterized in that The acquisition unit is specifically used to construct the first layer graph structure and the second layer graph structure according to the city planning information; the city planning information is used to obtain the city planning map; Determine the connection relationship between the nodes in the first layer graph structure and the nodes in the second layer graph structure to obtain the traffic expression graph.

12. The device according to claim 11, characterized in that The acquisition unit is specifically used to extract corner points in the urban planning map corresponding to the urban planning information as vertices to obtain a plurality of vertices; Determine the connection relationship between the multiple vertices to obtain the first layer graph structure.

13. The device according to claim 11, characterized in that The acquisition unit is specifically used to extract traffic elements belonging to the first type from the urban planning map corresponding to the urban planning information, where the traffic elements of the first type include: any one of a functional area, an intersection, and a lane area; Determine the connection relationship between the first type of traffic elements to obtain the second layer graph structure.

14. The device according to claim 9 or 10, characterized in that The traffic expression graph also includes a third-layer graph structure and a fourth-layer graph structure, and the types of nodes in any two layers of the first-layer graph structure, the second-layer graph structure, the third-layer graph structure and the fourth-layer graph structure are different.

15. The device according to claim 9 or 10, characterized in that The map sequence includes a first element sequence and a second element sequence, the length of the first element sequence is different from the length of the second element sequence, the first element sequence represents a first traffic element, the second element sequence represents a second traffic element, the type of the first traffic element is different from the type of the second traffic element, and the first traffic element and the second traffic element correspond to traffic elements in the static traffic scene.

16. The device according to claim 15, characterized in that The second element sequence also includes a first constraint sequence and a second constraint sequence, the length of the first constraint sequence is different from the length of the second constraint sequence, the first constraint sequence represents the connection relationship between the third traffic element and the fourth traffic element, the second constraint sequence represents the connection relationship between the fifth traffic element and the sixth traffic element, and the third traffic element, the fourth traffic element, the fifth traffic element, and the sixth traffic element correspond to traffic elements in the static traffic scene.

17. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, wherein the computer program includes program instructions. When the program instructions are executed by a processor, the processor executes the method according to any one of claims 1 to 8.

Images