High-precision map construction method and system based on laser point cloud and real scene three-dimensional

Abstract

The present invention discloses a method and system for constructing a high-precision map based on laser point cloud and real-scene 3D. The method includes: obtaining laser point cloud data and real-scene 3D image data; obtaining full-element terrain data based on the laser point cloud data and the real-scene 3D image data; reading corresponding element data from the full-element terrain data according to preset layer logic to construct a high-precision map, or reading corresponding real-scene semantic data to construct a road component-level real-scene 3D model based on the real-scene semantic data. The method of the present invention can uniformly manage data from different sources by separately constructing a first database and a second database. When processing data, the association logic between the first database and the second database can accurately match and load data, so that a high-precision map can be quickly constructed or a road component-level real-scene 3D model can be independently constructed according to specific needs.

Description

Technical Field

[0001] The present invention belongs to the field of map data processing, and in particular relates to a method and system for constructing a high-precision map based on laser point cloud and real-scene three-dimensional. Background Art

[0002] In the field of autonomous driving, where vehicle, road, and cloud systems are integrated, high-precision maps are undoubtedly a crucial component. Due to the massive volume of HD map data and the difficulty in integrating data collected by different devices and methods, processing data using a single source can result in low accuracy and precision. Fusion of multiple data sources requires separate loading and fusion through different software, resulting in high processing complexity. Furthermore, for scenarios requiring high real-time updates, existing technologies require separate loading and updating of each data source, resulting in slow update efficiency. Summary of the Invention

[0003] In order to solve the above problems existing in the prior art, the present invention provides a method and system for constructing high-precision maps based on laser point cloud and real-world 3D. The technical problems to be solved by the present invention are achieved through the following technical solutions:

[0004] A method for constructing a high-precision map based on laser point cloud and real-scene three-dimensional data, comprising:

[0005] Acquire laser point cloud data and real-scene 3D image data;

[0006] Obtaining full-factor terrain data based on the laser point cloud data and the real-scene three-dimensional image data;

[0007] Corresponding element data is read from the full-element terrain data according to preset layer logic to construct a high-precision map and a road component-level real-scene three-dimensional model.

[0008] In one specific embodiment, obtaining laser point cloud data and real-scene three-dimensional image data includes:

[0009] constructing a first database and a second database;

[0010] In response to receiving raw laser point cloud data sent by a plurality of point cloud data sources, extracting point cloud core data of the raw laser point cloud data in each point cloud data source, classifying the core data of each point cloud data source and storing them in a first database as laser point cloud data, wherein the first database includes a laser point cloud data index;

[0011] In response to receiving original real-scene 3D image data sent by multiple real-scene data sources, extracting real-scene core data from the original real-scene 3D image data in each real-scene data source, classifying the real-scene core data of each real-scene data source, and storing the data in a second database as real-scene 3D image data, wherein the second database includes an index of the real-scene 3D image data;

[0012] The first database and the second database have a preset association relationship, and the preset association relationship is determined according to the laser point cloud data and the real-scene three-dimensional image data.

[0013] In a specific embodiment, the point cloud core data includes point cloud coordinate data, point cloud RGB data, and point cloud classification data; the real scene core data includes real scene coordinate data, real scene texture data, and real scene semantic data; the real scene semantic data has a real scene coding field, a real scene attribute field, a first real scene association field, and a second real scene association field;

[0014] Accordingly, the real scene core data of each real scene data source is classified and stored in the second database as real scene three-dimensional image data, including:

[0015] Read real scene coordinate data;

[0016] When it is determined that the real-scene coordinate data and the coordinate data in the other real-scene data source are in an inclusion relationship, the real-scene code field and the inclusion status of the other real-scene data source are stored in the first real-scene association field; or, when it is determined that the real-scene coordinate data and the coordinate data in the other real-scene data source are in an intersection relationship, it is determined whether the intersection area of the real-scene coordinate data and the coordinate data in the other real-scene data source is greater than a first threshold, and if so, the real-scene code field and the intersection status of the other real-scene data source are stored in the first real-scene association field;

[0017] When it is determined that the real-scene coordinate data and the point cloud coordinate data are in an inclusion relationship, the point cloud classification data and the inclusion status are stored in the second real-scene associated field; or, when it is determined that the real-scene coordinate data and the point cloud coordinate data are in an intersection relationship, it is determined whether the intersection area of the real-scene coordinate data and the point cloud coordinate data is greater than the first threshold. If so, the real-scene coding field and the intersection status of the other real-scene data source are stored in the second real-scene associated field.

[0018] In a specific embodiment, obtaining full-factor terrain data based on the laser point cloud data and the real-scene three-dimensional image data includes:

[0019] Read the real scene core data from the second database;

[0020] Classifying the real scene core data to obtain full-factor grid data, wherein the full-factor data includes point factor data, line factor data, surface factor data, and volume factor data;

[0021] Reading point cloud core data from the first database to generate full-factor elevation data corresponding to the full-factor grid data through a preset association relationship between the first database and the second database;

[0022] The full-factor grid data and the full-factor elevation data are superimposed to obtain full-factor terrain data.

[0023] In a specific embodiment, reading corresponding element data from the full-element terrain data according to a preset layer logic to construct a high-precision map includes:

[0024] Obtain the topological relationships between layers, within layers, and elements of a high-precision map based on the preset layer logic;

[0025] According to the topological relationship between layers, the topological relationship within layers and the topological relationship of elements, the corresponding full-factor data are extracted from the first database and the second database in sequence to build a high-precision map, or the corresponding real-scene semantic data is read to build a road component-level real-scene three-dimensional model based on the real-scene semantic data.

[0026] The present invention also provides a high-precision map construction system based on laser point cloud and real-scene three-dimensional, including:

[0027] Data acquisition module, used to acquire laser point cloud data and real-scene 3D image data;

[0028] A full-factor terrain data generation module, configured to obtain full-factor terrain data based on the laser point cloud data and the real-scene three-dimensional image data;

[0029] The map model construction module is used to read the corresponding element data from the full-element terrain data according to the preset layer logic to construct a high-precision map and a road component-level real-scene three-dimensional model.

[0030] In a specific embodiment, the data acquisition module includes:

[0031] A database construction unit, configured to construct a first database and a second database;

[0032] a laser point cloud data processing unit configured to, in response to receiving raw laser point cloud data sent by a plurality of point cloud data sources, extract point cloud core data from the raw laser point cloud data of each point cloud data source, classify the core data of each point cloud data source, and store the data in a first database as laser point cloud data, wherein the first database includes a laser point cloud data index;

[0033] a real-scene 3D image data processing unit configured to, in response to receiving raw real-scene 3D image data sent by a plurality of real-scene data sources, extract real-scene core data from the raw real-scene 3D image data from each real-scene data source, classify the real-scene core data from each real-scene data source, and store the data in a second database as real-scene 3D image data, wherein the second database includes a real-scene 3D image data index;

[0034] The first database and the second database have a preset association relationship, and the preset association relationship is determined according to the laser point cloud data and the real-scene three-dimensional image data.

[0035] In a specific embodiment, the point cloud core data includes point cloud coordinate data, point cloud RGB data, and point cloud classification data; the real scene core data includes real scene coordinate data, real scene texture data, and real scene semantic data; the real scene semantic data has a real scene coding field, a real scene attribute field, a first real scene association field, and a second real scene association field;

[0036] The real scene 3D image data processing unit specifically includes:

[0037] A real scene coordinate reading subunit, used for reading real scene coordinate data;

[0038] A first real-scene associated field acquisition subunit is configured to, when determining that the real-scene coordinate data and the coordinate data in another real-scene data source are in an inclusion relationship, store the real-scene code field and the inclusion status of the other real-scene data source in the first real-scene associated field; or, when determining that the real-scene coordinate data and the coordinate data in another real-scene data source are in an intersection relationship, determine whether an intersection area between the real-scene coordinate data and the coordinate data in the other real-scene data source is greater than a first threshold, and if so, store the real-scene code field and the intersection status of the other real-scene data source in the first real-scene associated field;

[0039] The second real scene associated field acquisition subunit is used to store the point cloud classification data and the inclusion status in the second real scene associated field when it is determined that the real scene coordinate data and the point cloud coordinate data are in an inclusion relationship; or, when it is determined that the real scene coordinate data and the point cloud coordinate data are in an intersection relationship, determine whether the intersection area of the real scene coordinate data and the point cloud coordinate data is greater than a first threshold. If so, store the real scene coding field and the intersection status of the other real scene data source in the second real scene associated field.

[0040] In a specific embodiment, the full-factor terrain data generation module includes:

[0041] A real scene core data reading unit, configured to read the real scene core data from the second database;

[0042] A full-factor grid data processing unit, configured to classify the real scene core data to obtain full-factor grid data, wherein the full-factor data includes point factor data, line factor data, surface factor data, and volume factor data;

[0043] a full-factor elevation data processing unit, configured to read point cloud core data from the first database, so as to generate full-factor elevation data corresponding to the full-factor grid data through a preset association relationship between the first database and the second database;

[0044] The full-factor terrain data generating unit is configured to superimpose the full-factor grid data and the full-factor elevation data to obtain full-factor terrain data.

[0045] In a specific embodiment, the map model construction module includes:

[0046] A topological relationship determination unit is used to obtain the inter-layer topological relationship, intra-layer topological relationship and element topological relationship of the high-precision map according to the preset layer logic;

[0047] The map model construction unit is used to extract the corresponding full-factor data from the first database and the second database in sequence according to the topological relationship between layers, the topological relationship within layers and the topological relationship of elements to construct a high-precision map, or to read the corresponding real-scene semantic data to construct a road component-level real-scene three-dimensional model based on the real-scene semantic data.

[0048] Beneficial effects of the present invention:

[0049] The high-precision map construction method based on laser point cloud and real-scene three-dimensional of the present invention can uniformly manage data from different sources by separately constructing a first database and a second database. When processing data, the association logic between the first database and the second database can accurately match and load data, so that high-precision maps can be quickly constructed or real-scene three-dimensional models of road components can be independently constructed according to specific needs.

[0050] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 This is a flow chart of a method for constructing a high-precision map based on laser point cloud and real-scene three-dimensional images, provided by an embodiment of the present invention;

[0052] Figure 2 This is a module block diagram of a high-precision map construction system based on laser point cloud and real-scene three-dimensional provided by an embodiment of the present invention;

[0053] Figure 3This is a schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0054] The present invention will be further described in detail below with reference to specific examples, but the embodiments of the present invention are not limited thereto. Example

[0055] See Figure 1 , Figure 1 The following is a flow chart of a method for constructing a high-precision map based on laser point cloud and real-scene 3D provided by an embodiment of the present invention, including:

[0056] S1. Acquire laser point cloud data and real-scene 3D image data. Specifically, laser point cloud data is mainly collected by vehicle-mounted laser radar and airborne laser radar, and real-scene 3D image data is mainly collected by multi-angle cameras.

[0057] Due to the large amount of laser point cloud data and real-scene 3D image data, acquiring all of this data through real-time acquisition would require significant manpower, material resources, and time. Therefore, the laser point cloud data and real-scene 3D image data of this embodiment can be acquired from a wide range of sources, including, for example, real-time acquisition data and data of varying specifications collected from existing databases. To facilitate the storage of laser point cloud data and real-scene 3D image data from different sources for subsequent processing, this step can be specifically implemented through the following steps:

[0058] S11, constructing a first database and a second database;

[0059] S12. In response to receiving original laser point cloud data sent by several point cloud data sources, extracting the point cloud core data of the original laser point cloud data in each point cloud data source, classifying the core data of each point cloud data source and storing them in a first database as laser point cloud data, wherein the first database includes a laser point cloud data index. It should be noted that since the point cloud data sources come from a wide range of sources, the data content sent by each point cloud data source may be different. Therefore, in this embodiment, only the point cloud core data is extracted, classified, and stored uniformly, and a laser point cloud data index is established to facilitate database access.

[0060] S13. In response to receiving raw real-scene 3D image data sent from several real-scene data sources, extract the real-scene core data of the raw real-scene 3D image data from each real-scene data source, classify the real-scene core data of each real-scene data source, and store it in a second database as real-scene 3D image data, wherein the second database includes a real-scene 3D image data index; wherein the first database and the second database have a preset association relationship, and the preset association relationship is determined based on the laser point cloud data and the real-scene 3D image data. When constructing a high-precision map, it is necessary to combine the laser point cloud data and the real-scene 3D image data. Therefore, when storing the first and second databases of this embodiment, the association relationship needs to be confirmed. By binding the data at the database stage, data loss and display confusion caused by data overlay errors during the subsequent model overlay process can be avoided, and processing time can be significantly reduced.

[0061] The high-precision map construction method based on laser point cloud and real-scene three-dimensional in this embodiment can uniformly manage data from different sources by separately constructing a first database and a second database. When processing data, the association logic between the first database and the second database can accurately match and load data, so that high-precision maps can be quickly constructed or real-scene three-dimensional models at the road component level can be independently constructed according to specific needs.

[0062] In a preferred embodiment, the point cloud core data includes point cloud coordinate data, point cloud RGB data, and point cloud classification data; the real scene core data includes real scene coordinate data, real scene texture data, and real scene semantic data; the real scene semantic data has a real scene coding field, a real scene attribute field, a first real scene association field, and a second real scene association field;

[0063] The corresponding step S13 may specifically include the following steps:

[0064] S131, read the real scene coordinate data;

[0065] S132: When it is determined that the real-scene coordinate data and the coordinate data in another real-scene data source are in an inclusion relationship, the real-scene code field and the inclusion status of the other real-scene data source are stored in the first real-scene association field; or, when it is determined that the real-scene coordinate data and the coordinate data in another real-scene data source are in an intersection relationship, it is determined whether the intersection area between the real-scene coordinate data and the coordinate data in the other real-scene data source is greater than a first threshold; if so, the real-scene code field and the intersection status of the other real-scene data source are stored in the first real-scene association field;

[0066] The processing of real-life three-dimensional image data is the most complex in the process of building a high-precision map, and the processing effect has a great impact on the map display details. Therefore, in order to ensure the accuracy of data extraction, it is necessary to compare data from different sources to ensure that the final selected data has high accuracy. The real-life coordinate data in this step is the reference coordinate data, which is compared and judged with the coordinate data in other real-life data sources to correct the reference coordinate data. Specifically, if it is determined that the real-life coordinate data and the coordinate data in other real-life data sources are in an inclusion relationship, it means that the two are in the same grid. After recording the inclusion relationship, if the grid data needs to be read later, all the data associated with it can be obtained directly through the first real-life association field without having to load a separate map. When it is determined that the real-scene coordinate data and the coordinate data in other real-scene data sources are in an intersecting relationship, it is considered that there is no direct logical association between the two. For example, it may be an element that was not associated before in different layers, so no need to proceed. However, due to the possibility of errors between different data sources, data that was originally in a containment relationship or the same relationship may be mistakenly considered to be in an intersecting relationship. Therefore, it is necessary to further determine whether the area of the intersection area is slightly different from the area of one of the two data, that is, whether the intersection area is greater than the first threshold. If it is greater than the first threshold, it means that the area of the intersection area is close to the area of one of the two data. Then, the real-scene coding field and the real-scene attribute field are combined to determine whether they are consistent, and a more accurate position can be further reconfirmed based on the position of the intersection area. The first threshold of this embodiment can be, for example, 95%.

[0067] S133. When it is determined that the real-scene coordinate data and the point cloud coordinate data are in an inclusion relationship, the point cloud classification data and the inclusion status are stored in the second real-scene associated field; or, when it is determined that the real-scene coordinate data and the point cloud coordinate data are in an intersection relationship, it is determined whether the intersection area of the real-scene coordinate data and the point cloud coordinate data is greater than a first threshold. If so, the real-scene coding field and the intersection status of the other real-scene data source are stored in the second real-scene associated field.

[0068] To establish an association between the first database and the second database, this embodiment uses a second real-scene association field to establish the association. For a specific implementation, see step S132. This second real-scene association field allows for direct access to the associated laser point cloud data from the database when loading real-scene 3D image data, facilitating local loading and display.

[0069] S2. Obtaining full-factor terrain data based on the laser point cloud data and the real-scene 3D image data;

[0070] S21. Read real scene core data from the second database; specifically, it is necessary to read real scene coordinate data, real scene texture data, real scene semantic data, and the real scene coding field, real scene attribute field, first real scene association field, and second real scene association field in the real scene semantic data.

[0071] S22. Classify the real-scene core data to obtain full-factor grid data, wherein the full-factor data includes point feature data, line feature data, surface feature data, and volume feature data; specifically, point features mainly include independent landforms, such as street lamps, manhole covers, etc., line features mainly include polyline data such as road boundaries and pipeline networks, surface features mainly include polygonal data such as building outlines, and volume features mainly include three-dimensional structures that require detailed drawing, such as bus stops, waiting rooms, underground pipeline corridors, and other data that require special processing.

[0072] S23. Reading the point cloud core data from the first database to generate full-factor elevation data corresponding to the full-factor grid data through a preset association relationship between the first database and the second database; that is, generating full-factor elevation data corresponding to the full-factor grid data through the second real-scene association field. It should be noted that since elevation data may be missing for point feature data, line feature data, surface feature data, and volume feature data in the full-factor grid data when corresponding to the full-factor elevation data, continuity processing of surface and internal data can be performed using methods such as kriging interpolation.

[0073] S24. Superimposing the full-factor grid data and the full-factor elevation data to obtain full-factor terrain data.

[0074] S3. Read the corresponding element data from the full-element terrain data according to the preset layer logic to construct a high-precision map and a road component-level real-scene 3D model. Specifically including:

[0075] S31. Obtaining the inter-layer topological relationship, intra-layer topological relationship, and element topological relationship of the high-precision map according to the preset layer logic;

[0076] After constructing the first database and the second database, this embodiment makes it possible to realize customized high-precision map construction according to different scene requirements. Therefore, it is not necessary to generate according to the standardized model in existing software tools. The organizational structure of layer logic is generally "map-region-map-layer group-layer-element". The topological relationship within the layer mainly includes the association relationship between layers, and the element topological relationship mainly includes the connection relationship and boundary relationship between point element data, line element data, surface element data and volume element data.

[0077] S32. Based on the topological relationships between layers, the topological relationships within layers, and the topological relationships of elements, the corresponding full-factor data is extracted from the first database and the second database in sequence to construct a high-precision map, or the corresponding real-scene semantic data is read to construct a road component-level real-scene three-dimensional model based on the real-scene semantic data. It should be noted that since the full-factor data of this implementation is extracted through the first database and the second database, there is no need to import and load it through a multi-terminal interface, and a high-precision map can be directly constructed. It should be noted that for scenarios with low global usage requirements but high local usage requirements, the solution of this implementation can directly construct a road component-level real-scene three-dimensional model or construct a road component-level real-scene three-dimensional model on an already loaded high-precision map without the need for global reloading. In this case, only the corresponding real-scene semantic data needs to be read separately, and the preset layer logic can be adjusted according to specific needs based on the logic of the real-scene semantic data to achieve separate loading, thereby saving processing resources and improving data processing efficiency. Specifically, the road component-level real-scene three-dimensional model includes a road network layer, a lane network layer, a road marking layer, and a road facility layer.

[0078] In a specific scenario, high-precision maps can be used for autonomous driving of intelligent connected vehicles, and real-life three-dimensional models of road components can be automatically produced based on the full-factor terrain map of the road. They can be used as the data base for the cloud control platform, as a visual three-dimensional map for the vehicle side, or for urban traffic simulation experiments, to assist in urban traffic planning and other application scenarios.

[0079] Please continue to see Figure 2 This embodiment also provides a high-precision map construction system based on laser point cloud and real-scene 3D, including:

[0080] Data acquisition module, used to acquire laser point cloud data and real-scene 3D image data;

[0081] A full-factor terrain data generation module, configured to obtain full-factor terrain data based on the laser point cloud data and the real-scene three-dimensional image data;

[0082] The map model construction module is used to read the corresponding element data from the full-element terrain data according to the preset layer logic to construct a high-precision map and a road component-level real-scene three-dimensional model.

[0083] In a specific embodiment, the data acquisition module includes:

[0084] A database construction unit, configured to construct a first database and a second database;

[0085] a laser point cloud data processing unit configured to, in response to receiving raw laser point cloud data sent by a plurality of point cloud data sources, extract point cloud core data from the raw laser point cloud data of each point cloud data source, classify the core data of each point cloud data source, and store the data in a first database as laser point cloud data, wherein the first database includes a laser point cloud data index;

[0086] a real-scene 3D image data processing unit configured to, in response to receiving raw real-scene 3D image data sent by a plurality of real-scene data sources, extract real-scene core data from the raw real-scene 3D image data from each real-scene data source, classify the real-scene core data from each real-scene data source, and store the data in a second database as real-scene 3D image data, wherein the second database includes a real-scene 3D image data index;

[0087] The first database and the second database have a preset association relationship, and the preset association relationship is determined according to the laser point cloud data and the real-scene three-dimensional image data.

[0088] In a specific embodiment, the point cloud core data includes point cloud coordinate data, point cloud RGB data, and point cloud classification data; the real scene core data includes real scene coordinate data, real scene texture data, and real scene semantic data; the real scene semantic data has a real scene coding field, a real scene attribute field, a first real scene association field, and a second real scene association field;

[0089] The real scene 3D image data processing unit specifically includes:

[0090] A real scene coordinate reading subunit, used for reading real scene coordinate data;

[0091] A first real-scene associated field acquisition subunit is configured to, when determining that the real-scene coordinate data and the coordinate data in another real-scene data source are in an inclusion relationship, store the real-scene code field and the inclusion status of the other real-scene data source in the first real-scene associated field; or, when determining that the real-scene coordinate data and the coordinate data in another real-scene data source are in an intersection relationship, determine whether an intersection area between the real-scene coordinate data and the coordinate data in the other real-scene data source is greater than a first threshold, and if so, store the real-scene code field and the intersection status of the other real-scene data source in the first real-scene associated field;

[0092] The second real scene associated field acquisition subunit is used to store the point cloud classification data and the inclusion status in the second real scene associated field when it is determined that the real scene coordinate data and the point cloud coordinate data are in an inclusion relationship; or, when it is determined that the real scene coordinate data and the point cloud coordinate data are in an intersection relationship, determine whether the intersection area of the real scene coordinate data and the point cloud coordinate data is greater than a first threshold. If so, store the real scene coding field and the intersection status of the other real scene data source in the second real scene associated field.

[0093] In a specific embodiment, the full-factor terrain data generation module includes:

[0094] A real scene core data reading unit, configured to read the real scene core data from the second database;

[0095] A full-factor grid data processing unit, configured to classify the real scene core data to obtain full-factor grid data, wherein the full-factor data includes point factor data, line factor data, surface factor data, and volume factor data;

[0096] a full-factor elevation data processing unit, configured to read point cloud core data from the first database, so as to generate full-factor elevation data corresponding to the full-factor grid data through a preset association relationship between the first database and the second database;

[0097] The full-factor terrain data generating unit is configured to superimpose the full-factor grid data and the full-factor elevation data to obtain full-factor terrain data.

[0098] In a specific embodiment, the map model construction module includes:

[0099] A topological relationship determination unit is used to obtain the inter-layer topological relationship, intra-layer topological relationship and element topological relationship of the high-precision map according to the preset layer logic;

[0100] The map model construction unit is used to extract the corresponding full-factor data from the first database and the second database in sequence according to the topological relationship between layers, the topological relationship within layers and the topological relationship of elements to construct a high-precision map, or to read the corresponding real-scene semantic data to construct a road component-level real-scene three-dimensional model based on the real-scene semantic data.

[0101] The embodiment of the present invention further provides an electronic device, such as Figure 3 As shown, it includes a processor 31, a communication interface 32, a memory 33 and a communication bus 34, wherein the processor 31, the communication interface 32, and the memory 33 communicate with each other through the communication bus 34.

[0102] Memory 33, for storing computer programs;

[0103] The processor 31 is configured to execute the program stored in the memory 33 by performing the following steps:

[0104] Acquire laser point cloud data and real-scene 3D image data;

[0105] Obtaining full-factor terrain data based on the laser point cloud data and the real-scene three-dimensional image data;

[0106] According to the preset layer logic, the corresponding element data is read from the full-element terrain data to construct a high-precision map, or the corresponding real-scene semantic data is read to construct a road component-level real-scene three-dimensional model according to the real-scene semantic data.

[0107] The communication bus mentioned in the electronic device mentioned above may be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus. This communication bus can be divided into an address bus, a data bus, a control bus, etc. For ease of illustration, only one thick line is used in the figure, but this does not mean that there is only one bus or only one type of bus.

[0108] The communication interface is used for communication between the above electronic device and other devices.

[0109] The memory may include random access memory (RAM) or non-volatile memory (NVM), such as at least one disk storage. Alternatively, the memory may be at least one storage device located away from the processor.

[0110] The above-mentioned processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, and discrete hardware components.

[0111] The method provided in the embodiments of the present invention can be applied to electronic devices. Specifically, the electronic devices can be desktop computers, portable computers, smart mobile terminals, servers, etc. This is not limited here; any electronic device that can implement the present invention falls within the scope of protection of the present invention.

[0112] The terminal device provided by the embodiments of the present invention may exist in various forms, including but not limited to:

[0113] (1) Mobile communication devices: These devices are characterized by their mobile communication capabilities and are primarily designed to provide voice and data communications. These terminals include smartphones (e.g., iPhones), multimedia phones, feature phones, and low-end phones.

[0114] (2) Ultra-mobile personal computer devices: These devices fall under the category of personal computers, have computing and processing capabilities, and generally also have mobile Internet access. These terminals include PDAs, MIDs, and UMPCs, such as the iPad.

[0115] (3) Portable entertainment devices: These devices can display and play multimedia content. These devices include audio and video players (such as iPods), handheld game consoles, e-books, smart toys, and portable car navigation devices.

[0116] (4) Other electronic devices with data interaction functions.

[0117] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of the technical features being referred to. Thus, a feature identified as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, "plurality" means two or more, unless otherwise specifically defined.

[0118] In the description of this specification, the reference terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" mean that the specific features, structures, materials, or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in any appropriate manner in any one or more embodiments or examples. In addition, those skilled in the art can combine and combine different embodiments or examples described in this specification.

[0119] Although the present application is described herein in conjunction with various embodiments, in the process of implementing the claimed application, those skilled in the art can understand and implement other changes to the disclosed embodiments by reviewing the drawings, the disclosure, and the appended claims. In the claims, the word "comprising" does not exclude other components or steps, and "a" or "an" does not exclude multiple situations. A single processor or other unit can implement several functions listed in the claims. Certain measures are recorded in different dependent claims, but this does not mean that these measures cannot be combined to produce good results.

[0120] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, devices (equipment), or computer program products. Therefore, the application can adopt the form of complete hardware embodiment, complete software embodiment, or the embodiment in combination with software and hardware, which are all collectively referred to as "module" or "system" herein. Moreover, the application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code. The computer program is stored / distributed in a suitable medium, provided together with other hardware or as a part of hardware, or other distribution forms can be adopted, such as by the Internet or other wired or wireless telecommunication systems.

[0121] The present application is described with reference to the flowcharts and / or block diagrams of the methods, apparatus (devices) and computer program products of the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0122] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0123] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0124] The above is a further detailed description of the present invention in conjunction with specific preferred embodiments, and the specific implementation of the present invention should not be considered to be limited to these descriptions. For those skilled in the art of the present invention, without departing from the concept of the present invention, several simple deductions or substitutions can be made, which should be considered to fall within the scope of protection of the present invention.

Claims

1. A method for constructing a high-precision map based on laser point cloud and real-scene three-dimensional, characterized in that: include: Acquire laser point cloud data and real-scene 3D image data; Obtaining full-factor terrain data based on the laser point cloud data and the real-scene three-dimensional image data; Reading corresponding element data from the full-element terrain data according to preset layer logic to construct a high-precision map and a road component-level real-scene three-dimensional model; Obtaining laser point cloud data and real-scene 3D image data includes: constructing a first database and a second database; In response to receiving raw laser point cloud data sent by a plurality of point cloud data sources, extracting point cloud core data of the raw laser point cloud data in each point cloud data source, classifying the core data of each point cloud data source and storing them in a first database as laser point cloud data, wherein the first database includes a laser point cloud data index; In response to receiving original real-scene 3D image data sent by multiple real-scene data sources, extracting real-scene core data from the original real-scene 3D image data in each real-scene data source, classifying the real-scene core data of each real-scene data source, and storing the data in a second database as real-scene 3D image data, wherein the second database includes an index of the real-scene 3D image data; The first database and the second database have a preset association relationship, and the preset association relationship is determined according to the laser point cloud data and the real-scene three-dimensional image data; The point cloud core data includes point cloud coordinate data, point cloud RGB data and point cloud classification data; the real scene core data includes real scene coordinate data, real scene texture data and real scene semantic data; the real scene semantic data includes a real scene coding field, a real scene attribute field, a first real scene association field and a second real scene association field; Accordingly, the real scene core data of each real scene data source is classified and stored in the second database as real scene three-dimensional image data, including: Read real scene coordinate data; When it is determined that the real-scene coordinate data and the coordinate data in the other real-scene data source are in an inclusion relationship, the real-scene code field and the inclusion status of the other real-scene data source are stored in the first real-scene association field; or, when it is determined that the real-scene coordinate data and the coordinate data in the other real-scene data source are in an intersection relationship, it is determined whether the intersection area of the real-scene coordinate data and the coordinate data in the other real-scene data source is greater than a first threshold, and if so, the real-scene code field and the intersection status of the other real-scene data source are stored in the first real-scene association field; When it is determined that the real-scene coordinate data and the point cloud coordinate data are in an inclusion relationship, the point cloud classification data and the inclusion status are stored in the second real-scene associated field; or, when it is determined that the real-scene coordinate data and the point cloud coordinate data are in an intersection relationship, it is determined whether the intersection area of the real-scene coordinate data and the point cloud coordinate data is greater than the first threshold. If so, the real-scene coding field and the intersection status of the other real-scene data source are stored in the second real-scene associated field. When loading the real-scene three-dimensional image data through the second real-scene associated field, the laser point cloud data associated with it is called through the database, thereby facilitating local loading and display.

2. The method for constructing a high-precision map based on laser point cloud and real-scene 3D according to claim 1, characterized in that: Obtaining full-factor terrain data based on the laser point cloud data and the real-scene three-dimensional image data, including: Read the real scene core data from the second database; Classifying the real scene core data to obtain full-factor grid data, wherein the full-factor grid data includes point factor data, line factor data, surface factor data, and volume factor data; Reading point cloud core data from the first database to generate full-factor elevation data corresponding to the full-factor grid data through a preset association relationship between the first database and the second database; The full-factor grid data and the full-factor elevation data are superimposed to obtain full-factor terrain data.

3. The method for constructing a high-precision map based on laser point cloud and real-scene 3D according to claim 1, characterized in that: Reading corresponding element data from the full-element terrain data according to preset layer logic to construct a high-precision map and a road component-level real-scene 3D model, including: Obtain the topological relationships between layers, within layers, and elements of a high-precision map based on the preset layer logic; According to the topological relationship between layers, the topological relationship within layers and the topological relationship of elements, the corresponding full-element terrain data are extracted from the first database and the second database in sequence to construct a high-precision map, or the corresponding real-scene semantic data are read to construct a road component-level real-scene three-dimensional model based on the real-scene semantic data.

4. A high-precision map construction system based on laser point cloud and real-scene 3D, characterized in that: include: Data acquisition module, used to acquire laser point cloud data and real-scene 3D image data; A full-factor terrain data generation module, configured to obtain full-factor terrain data based on the laser point cloud data and the real-scene three-dimensional image data; A map model construction module, configured to read corresponding element data from the full-element terrain data according to a preset layer logic to construct a high-precision map and a road component-level real-scene three-dimensional model; The data acquisition module includes: A database construction unit, configured to construct a first database and a second database; a laser point cloud data processing unit configured to, in response to receiving raw laser point cloud data sent by a plurality of point cloud data sources, extract point cloud core data from the raw laser point cloud data of each point cloud data source, classify the core data of each point cloud data source, and store the data in a first database as laser point cloud data, wherein the first database includes a laser point cloud data index; a real-scene 3D image data processing unit configured to, in response to receiving raw real-scene 3D image data sent by a plurality of real-scene data sources, extract real-scene core data from the raw real-scene 3D image data from each real-scene data source, classify the real-scene core data from each real-scene data source, and store the data in a second database as real-scene 3D image data, wherein the second database includes a real-scene 3D image data index; The first database and the second database have a preset association relationship, and the preset association relationship is determined according to the laser point cloud data and the real-scene three-dimensional image data; The point cloud core data includes point cloud coordinate data, point cloud RGB data and point cloud classification data; the real scene core data includes real scene coordinate data, real scene texture data and real scene semantic data; the real scene semantic data includes a real scene coding field, a real scene attribute field, a first real scene association field and a second real scene association field; The real scene 3D image data processing unit specifically includes: A real scene coordinate reading subunit, used for reading real scene coordinate data; A first real-scene associated field acquisition subunit is configured to, when determining that the real-scene coordinate data and the coordinate data in another real-scene data source are in an inclusion relationship, store the real-scene code field and the inclusion status of the other real-scene data source in the first real-scene associated field; or, when determining that the real-scene coordinate data and the coordinate data in another real-scene data source are in an intersection relationship, determine whether an intersection area between the real-scene coordinate data and the coordinate data in the other real-scene data source is greater than a first threshold, and if so, store the real-scene code field and the intersection status of the other real-scene data source in the first real-scene associated field; The second real scene associated field acquisition subunit is used to store the point cloud classification data and the inclusion status in the second real scene associated field when it is determined that the real scene coordinate data and the point cloud coordinate data are in an inclusion relationship; or, when it is determined that the real scene coordinate data and the point cloud coordinate data are in an intersection relationship, determine whether the intersection area of the real scene coordinate data and the point cloud coordinate data is greater than a first threshold. If so, store the real scene coding field and the intersection status of the other real scene data source in the second real scene associated field. When loading the real scene three-dimensional image data through the second real scene associated field, the laser point cloud data associated with it is called through the database, thereby facilitating local loading and display.

5. The high-precision map construction system based on laser point cloud and real-scene 3D according to claim 4 is characterized in that: The full-factor terrain data generation module includes: A real scene core data reading unit, configured to read the real scene core data from the second database; A full-factor grid data processing unit, configured to classify the real scene core data to obtain full-factor grid data, wherein the full-factor grid data includes point element data, line element data, surface element data, and volume element data; a full-factor elevation data processing unit, configured to read point cloud core data from the first database, so as to generate full-factor elevation data corresponding to the full-factor grid data through a preset association relationship between the first database and the second database; The full-factor terrain data generating unit is configured to superimpose the full-factor grid data and the full-factor elevation data to obtain full-factor terrain data.

6. The high-precision map construction system based on laser point cloud and real-scene 3D according to claim 4 is characterized in that: The map model construction module includes: A topological relationship determination unit is used to obtain the inter-layer topological relationship, intra-layer topological relationship and element topological relationship of the high-precision map according to the preset layer logic; The map model construction module unit is used to extract the corresponding full-factor data from the first database and the second database in sequence according to the topological relationship between layers, the topological relationship within the layer and the topological relationship of the elements to construct a high-precision map, or to read the corresponding real-scene semantic data to construct a road component-level real-scene three-dimensional model based on the real-scene semantic data.

Images