Map conversion method and device, electronic equipment and storage medium

Abstract

The invention discloses a map conversion method and device, electronic equipment and a storage medium. Comprising the following steps: analyzing multiple pieces of Internet of Vehicles map information to obtain a plurality of intersections and road information of a driving-in direction road of each intersection; performing interpolation resampling on point location data of a lane center line of a lane in each road in the road information to obtain resampling point locations; obtaining the lane width of each re-sampling point, and obtaining a road corresponding to an opposite lane of each road according to the re-sampling points; fitting the road between every two sampling points according to the lane width by adopting a section-by-section fitting mode to obtain an Opendrive format map of the road in the driving-in direction; generating an Opendrive format map in the intersection according to the driving-in direction road of each intersection and the road of the opposite lane of each intersection; and according to the driving-in direction road of each intersection and the Opendrive format map in the intersection, generating the Opendrive format maps of all intersections.

Description

Technical Field

[0001] The present application relates to the field of autonomous driving technology, and in particular to a map conversion method, device, electronic device and storage medium. Background Art

[0002] At present, with the rapid development of C-V2X (Cellular Vehicle-to-Everything) technology, the number of demonstration areas in various places is increasing, and the actual road tests of C-V2X applications are also showing explosive growth. A large number of actual road tests have generated massive data resources, mainly including five types of messages: BSM (Basic Safety Message), SPAT (Signal Phase and Timing Message), MAP (Map Information), RSI (Road Side Information) and RSM (Road Side Message). If these raw data resources can be used efficiently, real traffic scenes can be realistically reproduced in a virtual environment, thereby providing more accurate traffic state simulation for simulation tests.

[0003] The Opendrive high-precision map format is one of the most widely used map formats in autonomous driving technology. As an open source standard map format, it can effectively achieve resource sharing.

[0004] At present, how to convert V2X map (vehicle network map) messages into OPENDRIVE format high-precision maps is an urgent problem to be solved. Summary of the invention

[0005] The present application provides a map conversion method, device, electronic device and storage medium to achieve the purpose of converting a V2X map (vehicle-to-everything map) message into an OPENDRIVE format high-precision map.

[0006] In a first aspect, an embodiment of the present application provides a map conversion method, the method comprising:

[0007] Parsing multiple vehicle network map information to obtain road information of multiple intersections and the approaching direction road of each of the intersections;

[0008] Interpolate and resample the point data of the lane centerline of each lane in the road in the road information to obtain resampled points of the lane;

[0009] Obtaining a lane width corresponding to each of the resampling points, and obtaining a road corresponding to an opposite lane of each of the roads according to the resampling points;

[0010] Using a segment-by-segment fitting method, the roads between the resampled points are fitted according to the lane width to obtain Opendrive format maps of the entry direction roads of the plurality of intersections;

[0011] Generate an Opendrive format map of the interior of each intersection according to the roads in the respective entry directions and the roads in the opposite lanes of each intersection;

[0012] According to the Opendrive format map of the approaching direction road of each intersection and the Opendrive format map inside each intersection, the Opendrive format map of all intersections is generated.

[0013] In a second aspect, an embodiment of the present application provides a map conversion device, the device comprising:

[0014] A road information acquisition module is used to parse multiple vehicle network map information to obtain road information of several intersections and the approaching direction road of each of the intersections;

[0015] A resampling point acquisition module, used for performing interpolation resampling on the point data of the lane centerline of each lane in the road in the road information to obtain the resampling point of the lane;

[0016] A width road acquisition module, used to acquire the lane width corresponding to each of the resampling points, and acquire the road corresponding to the opposite lane of each of the roads according to the resampling points;

[0017] An entry direction map acquisition module is used to fit the roads between the resampled points in pairs according to the lane width in a segment-by-segment fitting manner to obtain Opendrive format maps of the entry direction roads of the plurality of intersections;

[0018] An intersection internal map generation module, used to generate an Opendrive format map of the interior of each intersection according to each of the entry direction roads and the roads of the opposite lanes of each intersection;

[0019] The Opendrive map generation module is used to generate Opendrive format maps of all intersections based on the Opendrive format map of the entrance direction road of each intersection and the Opendrive format map inside each intersection.

[0020] In a third aspect, an embodiment of the present application provides an electronic device, including:

[0021] A processor, a memory, and a computer program stored in the memory and executable on the processor, wherein the map conversion method described in any one of the above items is implemented when the processor executes the program.

[0022] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium. When instructions in the storage medium are executed by a processor of an electronic device, the electronic device is enabled to execute any of the map conversion methods described above.

[0023] Compared with the prior art, this application has the following advantages:

[0024] The solution provided in the embodiment of the present application parses multiple vehicle network map information to obtain road information of several intersections and the approaching roads of each intersection. The point data of the lane centerline of the lane in each road in the road information is interpolated and resampled to obtain the resampled points of the lane. The lane width corresponding to each resampled point is obtained, and the road corresponding to the opposite lane of each road is obtained based on the resampled point. Using a segment-by-segment fitting method, the roads between the two resampled points are fitted according to the lane width to obtain Opendrive format maps of the approaching roads of several intersections. Based on the various approaching roads and the roads of the opposite lanes of each intersection, an Opendrive format map inside the intersection of each intersection is generated. Based on the Opendrive format map of the approaching roads of each intersection and the Opendrive format map inside the intersection of each intersection, an Opendrive format map of all intersections is generated. The embodiment of the present application obtains vehicle network map data by parsing the vehicle network map information and resampling it, and then samples it segment by segment by fitting it, and constructs the final Opendrive format high-precision map segment by segment through the line segment roads between the point data, thereby achieving the purpose of converting the V2X map (vehicle network map) message into an Opendrive format high-precision map. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 A flowchart of a map conversion method provided in an embodiment of the present application;

[0026] Figure 2 A schematic diagram of a vehicle network map data acquisition process provided in an embodiment of the present application;

[0027] Figure 3 A schematic diagram of a process for generating a high-precision map in Opendrive format based on vehicle network map data provided in an embodiment of the present application;

[0028] Figure 4 A schematic diagram of the structure of a map conversion device provided in an embodiment of the present application;

[0029] Figure 5 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0030] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0031] Reference Figure 1 , shows a flowchart of the steps of a map conversion method provided by an embodiment of the present application. Figure 1 As shown, the map conversion method may specifically include:

[0032] Step 101: parse multiple vehicle network map information to obtain road information of multiple intersections and the approach direction road of each of the intersections.

[0033] The embodiments of the present application can be applied to the scenario of converting a connected vehicle map into a high-precision map in the Opendrive format.

[0034] The vehicle network map information is the V2X MAP information. In this example, a parsing script, such as a Python script, needs to be written to parse multiple vehicle network map information to obtain road information of several intersections and the approaching roads of each intersection. Specifically, the XML (Extensible Markup Language) file of the V2X MAP can be read and parsed by the parsing script to obtain multiple intersections (Node) and extract the road information of the approaching roads of each intersection.

[0035] In a specific implementation, the Link information (i.e., road information) of multiple incoming roads (i.e., In direction) at each intersection (in the V2X MAP, the In direction road and the opposite road (i.e., Out direction) are two isolated Links) may include: at least one lane (Lane), each Lane contains continuous point data (point), and these point data can be used to represent the specific location information on the center line of the lane. The script (Python) parses the XML file of the V2X MAP, and can implement the parsing logic of XML entities such as Node, Link, Lane, and Point.

[0036] Step 102: interpolating and resampling the point data of the lane centerline of each lane in the road in the road information to obtain the resampled points of the lane.

[0037] After parsing multiple vehicle network map information to obtain road information of several intersections and the approaching direction roads of each intersection, the point data of the lane centerline of each road in the road information can be interpolated and resampled (such as spline interpolation, etc.) to obtain the resampled points of each lane. Specifically, when there are multiple lanes on the road, a specified number of corresponding points can be resampled from the multiple lanes to obtain resampled points. For example, if there are two lanes on a road, 50 point data can be resampled from the point data of the lane centerlines of the two lanes by spline interpolation. These 50 point data are the point data resampled from the two lanes with a specified interval distance, that is, the two groups of 50 point data are one-to-one corresponding.

[0038] Of course, if there is only one lane on the road and the number of lane points is relatively small (usually no more than 30), the point data of the lane can be spline interpolated to increase the number of points. Then, the specified number of point data can be resampled according to the preset interval distance.

[0039] In one implementation of the present application, when there are merged lanes and / or split lanes on a road, spline interpolation processing can be performed on the lanes on the road to obtain spline interpolation lanes on the road. The spline interpolation lanes on the road are resampled into a specified number of points to obtain resampled points of the spline interpolation lanes.

[0040] The lane spline interpolation process of the lane centerline of each lane in the road may be as follows: obtain a reference lane in the road, where the reference lane is the longest lane in the road. Based on the point data of the reference lane, perform spline interpolation on lane portions of other lanes in the road that do not have point data, obtain a point interpolation lane with the same length as the reference lane, and the reference lane and the point interpolation lane have the same curvature, and use the reference lane and the point interpolation lane as spline interpolation lanes.

[0041] In this example, for each lane, the total length of its point data can be calculated, which can be achieved by accumulating the straight-line distances between adjacent points, and then finding the longest lane as the reference lane. Then, the point data of the non-reference lane and the reference lane can be compared to find the interval that needs to be interpolated. In each interpolation interval, spline interpolation can be performed on the non-reference lane based on the point data of the reference lane to generate new point data. Finally, the point data obtained by spline interpolation calculation is added to the point data of the interpolation lane to generate a point interpolation lane. The reference lane and point interpolation lane of the road can be used as the spline interpolation lane of the road.

[0042] The resampling method for the spline interpolation lanes may be: uniformly resampling a specified number of corresponding points from a plurality of spline interpolation lanes to obtain resampling points.

[0043] In order to solve the problem that lanes may become narrower due to merging or wider due to splitting, resulting in inconsistent number and length of points, the embodiment of the present application adopts lane spline interpolation and resampling to ensure that the number and length of points in each Lane of the same Link are the same, which facilitates the subsequent construction of Opendrive format maps.

[0044] Step 103: Obtain the lane width corresponding to each of the resampling points, and obtain the road corresponding to the opposite lane of each of the roads according to the resampling points.

[0045] After resampling the point data of the lane centerline of each lane in the road in the road information to obtain the resampled points of the lane, the lane width corresponding to each resampled point can be obtained.

[0046] In one implementation of the present application, when there is only one lane on the road, the fixed lane width of each lane carried in the vehicle network map information can be obtained. That is, the attribute of the lane width is directly included in the XML file of the V2X map information, and the fixed lane width can be directly extracted from the XML file of the V2X map information as the lane width corresponding to the lane where the resampling point is located.

[0047] In another specific implementation of the present application, when there are at least two lanes on the road (which may be spline interpolation lanes), the lane width corresponding to each resampling point can be determined according to the distance between the resampling points of two adjacent lanes on the road. Specifically, the resampling points of each lane are sampled along the extension direction of the lane, and the number and length are the same. Then, the lane width can be calculated by calculating the distance between the sampling points of each lane with the same sequence number. Assuming that n lanes are recorded as Lane i , where i∈[1,n], Lane i The corresponding point data is Point i [j], where i∈[1,n], j∈[1,N], then Point i [j] The corresponding width calculation formula can be (distance represents distance):

[0048] Width i [j] = 0.5*(distance(Width i [j],Width i-1 [j])+distance(Width i [j],Widthi+1 [j]))

[0049] Where i∈[2,n-1], when i=1, Width1[j]=distance(Width1[j],Width2[j]), when i=n, Width n [j] = distance (Width n-1 [j],Width n [j]).

[0050] The embodiment of the present application can process lanes of different lengths and numbers of points through lane spline interpolation and resampling, and dynamically calculate the lane width to more accurately reflect the road characteristics in actual traffic scenarios.

[0051] After obtaining the resampling points, the road corresponding to the opposite lane of each road can also be obtained according to the resampling points, that is, the exit direction road corresponding to the entry direction road of the intersection.

[0052] In one implementation of the present application, the position mean of each road can be calculated based on the resampled points of the lanes in the road, and the position mean is used as the road position of each road. Then, the road corresponding to the opposite lane of each road can be obtained based on the distance between the road position of each road and the road position of other roads and the distance threshold. Specifically, the road in other roads whose distance from the road position of each road is less than the distance threshold can be obtained as the opposite road corresponding to the opposite lane of each road. When there is no road in other roads whose distance from the road position of each road is less than the distance threshold, the opposite road corresponding to the opposite lane of each road is marked as empty. That is, after the resampling is completed, the Link corresponding to the opposite lane of each Link can be found, so as to obtain a Link pair set. Considering that the relevant information of each V2X MAP may be incomplete or not matched (the upstreamNodeId in the V2X MAP contains some opposite road information, but the data of each MAP may not match this upstreamNodeId), the search method based on geometric position can be: calculate the position mean of each Link, the point data (Point) of multiple Lanes contained in the Link, directly take the point data of the first Lane to calculate the position mean, and this position can be considered as the position of the Link. Then, the positions of these Links are compared in pairs, and when the distance is less than the threshold (customizable), the two Links are considered to be adjacent roads. When a Link cannot find a Link with a distance less than the threshold, it is considered that the current V2X MAP XML file lacks the opposite road corresponding to the Link, and the Link is marked as a Road separately, and the opposite road Link is None (this opposite road is None, and when the Opendrive format map is generated later, the Link will be directly copied to the opposite road as a mirror image).

[0053] Through the above steps, the data preparation work of the Opendrive format high-precision map can be completed. The prepared data includes: multiple Nodes (intersections), each Node (intersection) contains multiple Links (roads) in the in direction. Each Link (road) in the in direction contains multiple Lane resampled point data (Point), and each Link (road) is also associated with the corresponding Link of the opposite road.

[0054] Step 104: Using a segment-by-segment fitting method, the roads between the resampled points are fitted according to the lane width to obtain Opendrive format maps of the entry-direction roads of the plurality of intersections.

[0055] The traditional method of making high-precision maps in the Opendrive format usually involves first fitting a reference line (trajectory) of the road based on the data, and then offsetting the reference line along its normal direction to create the road. In this process, it is necessary to fit the curve and solve a series of parameters, including reference line parameters, offset parameters, and lane width change parameters. Due to the complex curve fitting involved, the amount of calculation is large and the efficiency is low. At the same time, the fitting effect has a great impact on the final map accuracy, which can easily lead to error accumulation. However, there is a certain error in converting the approximate V2X MAP high-precision map based on point description to the precise Opendrive format high-precision map based on line and offset description. Therefore, the embodiment of the present application proposes a segment-by-segment fitting method. That is, the segment-by-segment fitting method is adopted to fit the roads between the two resampled points according to the lane width to obtain the Opendrive format maps of the entry direction roads of several intersections. The implementation process of segment-by-segment fitting can be described in detail in conjunction with the following specific implementation method.

[0056] In a specific implementation of the present application, the above step 104 may include:

[0057] Sub-step S1: determining a reference line trajectory of each road according to the resampled points of the lanes in each road of the plurality of intersections.

[0058] In this embodiment, in the process of converting to the Opendrive format map, the reference line trajectory refers to the extension path of the center reference line of the road in space, which is the basic and necessary element of all roads in Opendrive, represents the geometric shape of the road, and all features of the road (such as lane lines) are extended based on the road reference line trajectory.

[0059] After obtaining the resampled points of the lanes in each road, the reference line trajectory of each road can be determined based on the resampled points in each road at several intersections. Specifically, for the first lane of each road, the point data (Point) is offset by 0.5 times × the lane width of the Lane (of course, the lane width can be appropriately increased by 20cm, etc., considering the width of the lane marking line) along the right normal of the point tangent direction (the tangent direction is the direction from the Point to the next Point, indicating the slope of the line segment formed by the two points) to obtain the reference line position corresponding to the point. Then, by traversing all the points in the first Lane, the reference line trajectory (i.e., point set) of the road can be generated, recorded as: Reference_Line_Points, and the number of point data is N (consistent with the number of resampled points).

[0060] Sub-step S2: Obtain the line segment, line segment length and reserved curve length between every two points on the reference line trajectory.

[0061] After obtaining the reference line trajectory, the line segment, line segment length and reserved curve length between every two points on the reference line trajectory can be obtained. Specifically, the above generated Reference_Line_Points (road reference line trajectory) can be traversed, and N points correspond to N-1 line segments, and each line segment is recorded as Reference_Line_Points. i ->Reference_Line_Points i+1 , where i∈[1,N-1]. Then, we can calculate Reference_Line_Points i ->Reference_Line_Points i+1 The length of the line segment is Distance i , in order to make the line segment connection smoother, you can reserve (1-Factor)*Distance i The length is connected by adaptive curves, where Factor is the ratio of the actual used line segments. For example, if there are 100 resampling points on the lane, there can be 99 line segments on the lane, and 99 curves are reserved. If the distance between every two points is 100 meters, then a line segment with a length of 90 meters can be created between every two points, and a curve with a reserved curve length of 10 meters can be reserved, etc.

[0062] It can be understood that the above examples are merely examples listed for a better understanding of the technical solutions of the embodiments of the present application, and are not intended to be the sole limitation to the embodiments.

[0063] The embodiment of the present application can effectively adjust the shape of the connecting line segment by reserving the corresponding adaptive curve, thereby ensuring smooth transition at turns and positions with large changes.

[0064] Sub-step S3: Create an Opendrive straight segment road according to the position information of the starting point of the line segment, the length of the line segment, the line segment slope of the line segment, the number of lanes of the road and the lane width of each of the resampling points.

[0065] After obtaining the length of the line segment, you can create the corresponding Opendrive straight line segment road according to the position of the starting point of the line segment (i.e. the starting resampling point of the line segment), the length of the line segment, the line slope of the line segment, the number of lanes on the road, and the lane width of each resampling point. i Location, Reference_Line_Points i ->Reference_Line_Points i+1The slope of the line segment and the actual length of the line segment used are Factor*Distance i , the number of lanes and the width of each lane j [i] (j is the lane number), when creating an Opendrive straight segment road, the newly created Opendrive straight segment road can be saved to Link_Parts (i.e., a road collection). In the specific implementation, you can use Python's scenariogeneration library (or custom function) to create a new road object, and add a line segment to the road object based on the starting point, the end point (which can be calculated based on the line segment slope and the actual line segment length used) and the line segment type (straight line). And set the lane information for each line segment based on the number of lanes and the lane width of each resampling point. In Opendrive, lane information can include lane ID, type, width and other information. At this point, you can get the Opendrive straight segment road for each lane in the road.

[0066] Of course, if the opposite road of the Link exists (i.e. not None), only the right road needs to be generated. If the opposite road does not exist (i.e. None), both the left and right roads need to be generated based on the same reference line trajectory, and the parameters of the left road should be consistent with those of the right road.

[0067] Sub-step S4: creating an Opendrive curve segment road according to the reserved curve length, the number of lanes of the road and the lane width of each of the resampling points.

[0068] After obtaining the reserved curve length, the Opendrive curve segment road can be created based on the reserved curve length, the number of lanes on the road, and the lane width of each resampling point. Specifically, the adaptive curve length (1-Factor)*Distance i , the number of lanes and the width of each lane j [i] (j is the lane number), create an Opendrive curve segment road, and save the newly created Opendrive curve segment road to Link_Parts.

[0069] Of course, if the opposite road of the Link exists (i.e. not None), only the right road needs to be generated; if the opposite road does not exist (i.e. None), both the left and right roads need to be generated based on the same reference line trajectory, and the parameters of the left road should be consistent with those of the right road.

[0070] The embodiment of the present application can handle different situations of whether there are oncoming roads or not, and automatically generate an Opendrive format map within the intersection to adapt to various complex road network structures.

[0071] Sub-step S5: adding a predecessor road and a successor road respectively to the Opendrive straight segment roads and the Opendrive curve segment roads between the resampled points, and generating an Opendrive format map of the approaching direction roads of each intersection.

[0072] After obtaining the Opendrive straight segment roads and Opendrive curved segment roads, predecessor roads and successor roads can be added to the Opendrive straight segment roads and Opendrive curved segment roads between the two resampled points, respectively, to generate an Opendrive format map of the entry direction roads of each intersection. For example, straight segment roads and curved segment roads are placed together alternately, and it is necessary to add predecessors and successors to each road segment. For example, the predecessor of the straight segment road is the road curve on its left side, and the successor is the road curve on the right side. The predecessor of the curved segment road is the road line segment on its left side, and the successor is the road line segment on its right side, etc.

[0073] In this embodiment, after obtaining the Opendrive road segments and Opendrive road curves, an Opendrive road set Link_Parts of the road can be formed, in which the line segment roads and the adaptive curve roads are cross-arranged, and the number is 2N (N is the number of resampling points for each lane). Then, traverse Link_Parts and add the leading road and the trailing road for each road. For example, the leading road of Link_Parts[i] is Link_Parts[i-1], and the trailing road is Link_Parts[i+1]. In this process, the adaptive road curve will automatically adapt to the line segment roads between the two fixed positions, and its length will be adjusted accordingly to ensure a smooth transition.

[0074] In the previous step, the roads in the in direction (i.e. the direction of entry) have been added, and now the roads in the out direction (i.e. the direction of exit) need to be added. Specifically, all the in-direction Links under the Node can be traversed, and the corresponding out-direction Link can be found according to the Link pair set established previously, and then its Link_Parts[0] can be added to the Junction_Link corresponding to the Node. In this way, the Junction_Link can completely contain the roads in the in direction and the out direction.

[0075] The preceding road refers to the road (or lane) immediately preceding the road (or lane) in the driving direction. In other words, it is the road that the vehicle traveled on before entering the current road. In Opendrive, this relationship is usually expressed through <link> Inside the element <predecessor>The preceding road is crucial for determining the vehicle's path and navigation, especially in complex road networks.

[0076] The subsequent road is the road (or lane) that follows a road (or lane) in the driving direction. It is the road that the vehicle will enter after the current road is completed. In Opendrive, the subsequent road is defined by <link> Inside the element <successor>The definition of the rear road is also important for path planning and navigation because it helps the system understand where the vehicle might go after leaving the current road.

[0077] The embodiment of the present application constructs the Opendrive high-precision map by adopting a segment-by-segment fitting method, avoiding the complex curve fitting process, improving the conversion efficiency, and ensuring the map accuracy. At the same time, adaptive curves are used to connect road segments to ensure smooth transitions at turns and locations with large changes, thereby improving the quality of the generated map.

[0078] Step 105: Generate an Opendrive format map of the interior of each intersection according to the roads in the respective entry directions of each intersection and the roads in the opposite lanes.

[0079] In the above steps, the roads corresponding to the opposite lanes of each road have been obtained. When generating the intersection map, the Opendrive format map inside each intersection can be generated according to the roads in the direction of each intersection and the roads in the opposite lanes. When constructing the Opendrive format map of each intersection, the ending line segment road of each entrance road and the starting line segment road of each exit road can be obtained. In the specific implementation, for each Link, its starting line segment road Link_Parts[0] and the last ending line segment road Link_Parts[-1] can be recorded for subsequent intersection generation. The last ending line segment road Link_Parts[-1] of each entrance direction Link needs to be added to the Junction_Link set of the Node corresponding to the Link's entrance direction, and the first line segment road Link_Parts[0] of the Link corresponding to the opposite road of the Link needs to be added to the Junction_Link set of the same Node. This Junction_Link is used for the Node to automatically generate the Opendrive format map inside the intersection.

[0080] Step 106: Generate Opendrive format maps of all intersections based on the Opendrive format map of the approaching road of each intersection and the Opendrive format map inside each intersection.

[0081] After obtaining the Opendrive format map of the approaching direction road of each intersection and the Opendrive format map of the interior of each intersection, the Opendrive format map of all intersections can be generated according to the Opendrive format map of the approaching direction road of each intersection and the Opendrive format map of the interior of each intersection. In a specific implementation, the initial Opendrive format map of all intersections can be composed according to the Opendrive format map of the approaching direction road of each intersection and the Opendrive format map of the interior of each intersection.

[0082] After obtaining the initial Opendrive format map of all intersections, the initial Opendrive format map can be optimized based on the vehicle network map information to obtain the Opendrive format map. Specifically, the high-precision map generated based on the compensation optimization of multiple V2X map information data does not fully consider the situation where roads merge or split, and the automatically generated intersections may be inaccurate, so later optimization is required. In order to ensure the accuracy and practicality of the map, it may be necessary to import the point data in the Opendrive format map and V2X map information into the simulation software or Opendrive format map editor for manual optimization.

[0083] The embodiment of the present application obtains vehicle network map data by parsing vehicle network map information and resampling, and then samples the road between two point data in a segment-by-segment fitting manner, and constructs the final Opendrive format high-precision map segment by segment, thereby achieving the purpose of converting the V2X map (vehicle network map) message into an Opendrive format high-precision map. At the same time, the Opendrive high-precision map is constructed in a segment-by-segment fitting manner, which avoids the complex curve fitting process, improves the conversion efficiency, and ensures the map accuracy. The above method can efficiently utilize massive road test data: it can fully tap the potential value of a large amount of actual road test data, and provide higher authenticity and data-driven simulation support for virtual simulation.

[0084] The above technical solution provided by the embodiment of the present application can be divided into two processes: 1. Preparation of Internet of Vehicles map data; 2. Generation of Opendrive format high-precision map based on Internet of Vehicles map data. Figure 2 and Figure 3 These two processes are described in detail.

[0085] Reference Figure 2 , showing a schematic diagram of a vehicle network map data acquisition process provided in an embodiment of the present application.

[0086] V2X MAP is an approximate high-precision map based on point description, while Opendrive is an accurate high-precision map based on line and offset description. From point to line, from approximate to accurate, some special methods are needed to process and optimize the data. The processing flow can be as follows: Figure 2 As shown:

[0087] 1. Read and load multiple V2X MAP XML files (can be read and loaded through Python scripts). To extract Node, Link, Lane, and Point information. Specifically, multiple Nodes can be obtained by loading XML files, and Link information corresponding to multiple in-direction roads of each Node can be extracted (in V2X MAP, the in-direction road and the opposite road (out direction) are two isolated Links). Each Link road contains multiple Lanes, and each Lane contains continuous point data (points), which represent the specific location information on the center line of the lane. The script (Python) parses the XML file of V2X MAP, and needs to implement the parsing logic of XML entities such as Node, Link, Lane, and Point by itself.

[0088] 2. Process each Link. The processing flow can be:

[0089] 1) Spline interpolation and resampling are performed on the Lanes of each Link. Spline interpolation is performed on the point data (Point) of multiple Lanes contained in each Link. Since the length and number of point data of each Lane are different, the reason is that the number and length of points may be inconsistent due to merging and narrowing or splitting and widening. In order to simplify the processing, the spline interpolation method is used to extend the lane and resample it. This method ensures that the number and length of points of each Lane in the same Link are the same. When the Opendrive format map is generated later, these compensated roads will be cropped. The specific method of spline interpolation to extend the lane and resample can be: (1) Select the longest Lane as the reference (this is also a simplified processing. In the case of no coverage, the road will be optimized when the Opendrive format map is generated later). The extension direction of the reference Lane is considered to be the extension direction of the Link. (2) For the other shorter Lanes without point data, spline interpolation is performed to extend them to the same length as the reference Lane, and the curvature is maintained to be the same as the reference Lane. (3) Each lane is uniformly resampled to a specified number N of points along the lane extension direction (reference Lane extension direction), so that the number and length of points in each lane under the same link can be the same.

[0090] 2) Calculate the lane width corresponding to the resampling points of each Lane in the direction of Link extension. Each sampling point corresponds to a specific lane width. This method can dynamically reflect the lane width that changes in real time along the lane direction. Although the lane width has been defined in V2X MAP, the width of each lane is a fixed value and cannot reflect the change in width along the lane direction. Therefore, through the combination of the above resampling points and lane width, the dynamic characteristics of lane width in actual traffic scenarios can be more accurately described. The lane width calculation method for the resampling points of each Lane is as follows: The resampling points of each Lane are sampled along the direction of lane extension, and the number and length are the same, then the lane width can be calculated by calculating the distance between the sampling points of each Lane with the same serial number.

[0091] 3) Find the opposite lane of each Link. Considering that the relevant information of each V2X MAP may be incomplete or inconsistent (the upstreamNodeId in the V2X MAP contains some opposite road information, and the data of each MAP may not match this upstreamNodeId), the search method based on geometric position can be: (1) Calculate the position mean of each Link. For the point data (Point) of multiple Lanes contained in the Link, directly take the point data of the first Lane to calculate the position mean. This position can be considered as the position of the Link. (2) Compare the positions of these Links in pairs. When the distance is less than the threshold (which can be customized), the two Links are considered to be adjacent roads. When a Link cannot find a Link with a distance less than the threshold, it is considered that the current V2X MAP XML file lacks the opposite road corresponding to the Link. Then the Link is marked as a Road separately, and the opposite road Link is None (this opposite road is None. When the Opendrive format map is generated later, the Link will be directly copied to the opposite lane as a mirror image).

[0092] So far, the data preparation for Opendrive high-precision map has been completed. The data contains multiple Nodes (intersections), and each Node (intersection) contains multiple Links (roads) in the in direction. Each Link (road) in the in direction contains multiple Lane resampled point data (Point), and each Link (road) is also associated with the corresponding Link of the opposite lane.

[0093] The single XML file of the V2X MAP map only contains the road information in the in direction, while the roads in the out direction are stored in another V2X MAP XML file. It can be understood that the roads in a single direction are isolated, which is very different from the Opendrive format map. The Opendrive format map is based on the reference line. Assuming that the reference line points to the intersection, the road to the right of the reference line is the in direction road, and the road to the left is the out direction road. The roads in the in and out directions are an integrated whole. Therefore, when making an Opendrive format high-precision map based on V2XMAP data, the roads in the in and out directions need to be regarded as independent Opendrive roads, and each road only contains the right lane and does not contain the left lane.

[0094] The traditional method of making high-precision maps in the Opendrive format usually involves first fitting a reference line (trajectory) of the road based on the data, and then offsetting the reference line along its normal direction to create the road. This process requires fitting curves and solving a series of parameters, including reference line parameters, offset parameters, and lane width change parameters. Due to the complex curve fitting involved, the amount of calculation is large and the efficiency is low. At the same time, the fitting effect has a great impact on the final map accuracy, which can easily lead to error accumulation. However, there are certain errors in the conversion from the approximate V2X MAP high-precision map based on point description to the precise Opendrive format high-precision map based on line and offset description. Therefore, this embodiment adopts a segment-by-segment fitting method to fit the road between the two point data (Point), and construct the final Opendrive format high-precision map segment by segment. This method takes into account both efficiency and accuracy, and can effectively improve the quality of map conversion.

[0095] The process of generating a map based on the above optimized data can be as follows: Figure 3 As shown:

[0096] 1. Traverse and process the above optimized Node set (each Node usually corresponds to an intersection), and process each Node as follows: First, generate all the Links (roads) in the in direction corresponding to the Node. When processing other Nodes, the Links (roads) in the out direction corresponding to the Node will be generated accordingly, because the Links (roads) in the out direction of this Node are exactly the Links (roads) in the in direction of other Nodes.

[0097] 2. For each Node, it is necessary to traverse and process all the Links (roads) in the in direction corresponding to the Node. The processing process is as follows:

[0098] 1) Calculate the trajectory of the road reference line (i.e., point set): For the first Lane of each Link, its point data (Point) is offset along the right normal of the point tangent direction (the tangent direction is the direction from the Point to the next Point, indicating the slope of the line segment formed by two points) by 0.5 times the lane width of the Lane (20cm can be appropriately increased to consider the width of the lane marking line) to obtain the reference line position corresponding to the point. By traversing all the points of the first Lane, the reference line trajectory (i.e., point set) of the road can be generated, recorded as: Reference_Line_Points, and the number of point data is N (consistent with the number of points resampled previously).

[0099] 2) Traverse and process Reference_Line_Points (the trajectory of the road reference line), N points correspond to N-1 line segments, and each line segment is recorded as:

[0100] Reference_Line_Points i ->Reference_Line_Points i+1 , where i∈[1,N-1].

[0101] 3) Calculate Reference_Line_Points i ->Reference_Line_Points i+1 The length of the line segment is Distance i , in order to make the line segment connection smoother, you can reserve (1-Factor)*Distance i The length is connected using an adaptive curve, where Factor is the proportion of the actual line segment used. This method can effectively adjust the shape of the connecting line segment to ensure smooth transitions at turns and locations with large changes.

[0102] 4) Based on the data: Reference_Line_Points i Location, Reference_Line_Points i ->Reference_Line_Points i+1 The slope of the line segment and the actual length of the line segment used are Factor*Distance i , the number of lanes and the width of each lane j [i] (j is the lane number), when creating an Opendrive road segment, you can save the newly created Opendrive road segment to Link_Parts. If the opposite road of the Link exists (not None), only the right road needs to be generated; if the opposite road does not exist (None), both the left and right roads need to be generated based on the same reference line, and the parameters of the left road should be consistent with those of the right road.

[0103] 5) According to the data: Adaptive curve length (1-Factor)*Distance i , the number of lanes and the width of each lane j [i] (j is the lane number), when creating an Opendrive adaptive curve road (refer to the Python library scenariogeneration, some functions require secondary development), save the newly created Opendrive adaptive curve road to Link_Parts. If the opposite road of the Link exists (that is, not None), only the right road needs to be generated; if the opposite road does not exist (that is, None), both the left and right roads need to be generated based on the same reference line, and the parameters of the left road should be consistent with those of the right road.

[0104] 6) After traversing and processing Reference_Line_Points, we get the Opendrive road set Link_Parts, where line roads and adaptive curve roads are arranged crosswise, with a number of 2N. Next, we traverse Link_Parts and add the leading and trailing roads for each road, for example: the leading road of Link_Parts[i] is Link_Parts[i-1], and the trailing road is Link_Parts[i+1]. In this process, the adaptive curve road will automatically fit between the two fixed line roads, and its length will be adjusted accordingly to ensure a smooth transition.

[0105] 7) For each Link, it is also necessary to record the first segment road Link_Parts[0] and the last segment road Link_Parts[-1]. The last segment road Link_Parts[-1] should be added to the Junction_Link collection of the Node corresponding to the Link. This Junction_Link is used by the Node to automatically generate the Opendrive format map within the intersection.

[0106] 3. Generate the intersection map. After processing all nodes in the previous step, the non-junction roads of the entire map have been generated. Next, generate the intersection map corresponding to each node.

[0107] 1) Supplement the Junction_Link data corresponding to each Node. In the previous step, the roads in the in direction have been added. Now the roads in the out direction need to be supplemented. The supplement method can be: traverse all the in-direction Links under the Node, find the corresponding out-direction Link according to the Link pair set established previously, and then take its Link_Parts[0] and add it to the Junction_Link corresponding to the Node. In this way, the Junction_Link can completely contain the roads in the in direction and the out direction.

[0108] 2) Create an Opendrive junction road based on the Junction_Link data corresponding to each Node.

[0109] After the Opendrive format map is generated, the Opendrive format map can be cropped and optimized: The high-precision map generated based on the compensation optimization of multiple V2X MAP data does not fully consider the situation of road merging or splitting, and the automatically generated intersections may be inaccurate, so later optimization is required. In order to ensure the accuracy and practicality of the map, it may be necessary to import the point data in the Opendrive format map and V2X MAP into the simulation software or Opendrive format map editor for manual optimization.

[0110] The map conversion method provided in the embodiment of the present application parses multiple vehicle network map information to obtain road information of several intersections and the approaching roads of each intersection. The point data of the lane centerline of the lane in each road in the road information is interpolated and resampled to obtain the resampled points of the lane. The lane width corresponding to each resampled point is obtained, and the road corresponding to the opposite lane of each road is obtained based on the resampled point. Using a segment-by-segment fitting method, the roads between the two resampled points are fitted according to the lane width to obtain Opendrive format maps of the approaching roads of several intersections. Based on the various approaching roads and the roads of the opposite lanes of each intersection, an Opendrive format map inside the intersection of each intersection is generated. Based on the Opendrive format map of the approaching roads of each intersection and the Opendrive format map inside the intersection of each intersection, an Opendrive format map of all intersections is generated. The embodiment of the present application obtains vehicle network map data by parsing the vehicle network map information and resampling it, and then samples it segment by segment by fitting it, and constructs the final Opendrive format high-precision map segment by segment through the line segment roads between the point data, thereby achieving the purpose of converting the V2X map (vehicle network map) message into an Opendrive format high-precision map.

[0111] Reference Figure 4 , shows a schematic diagram of the structure of a map conversion device provided in an embodiment of the present application. Figure 4 As shown, the map conversion device 400 may specifically include the following modules:

[0112] A road information acquisition module 410 is used to parse multiple vehicle network map information to obtain road information of multiple intersections and the approaching direction road of each of the intersections;

[0113] A resampling point acquisition module 420 is used to interpolate and resample the point data of the lane centerline of each lane in the road in the road information to obtain the resampling point of the lane;

[0114] A width road acquisition module 430 is used to acquire the lane width corresponding to each of the resampling points, and acquire the road corresponding to the opposite lane of each of the roads according to the resampling points;

[0115] The entry direction map acquisition module 440 is used to fit the roads between the resampled points in pairs according to the lane width in a segment-by-segment fitting manner to obtain Opendrive format maps of the entry direction roads of the plurality of intersections;

[0116] The intersection internal map generation module 450 is used to generate an Opendrive format map of the interior of each intersection according to each of the entry direction roads and the roads of the opposite lanes of each intersection;

[0117] The Opendrive map generation module 460 is used to generate Opendrive format maps of all intersections based on the Opendrive format map of the approaching direction road of each intersection and the Opendrive format map inside each intersection.

[0118] Optionally, the resampling point acquisition module includes:

[0119] A lane interpolation unit, configured to perform spline interpolation processing on lanes in the road to obtain spline interpolated lanes in the road when there are merged lanes and / or split lanes on the road;

[0120] The resampling point acquisition unit is used to resample the spline interpolation lane in the road into a specified number of points to obtain the resampling points of the spline interpolation lane.

[0121] Optionally, the lane interpolation unit includes:

[0122] A reference lane acquisition subunit, used to acquire a reference lane in the road, wherein the reference lane is the longest lane in the road;

[0123] The lane interpolation subunit is used to perform spline interpolation on lane portions without point data in other lanes of the road based on the point data of the reference lane to obtain a point interpolation lane with the same length as the reference lane, and use the reference lane and the point interpolation lane as the spline interpolation lane.

[0124] Optionally, the width road acquisition module includes:

[0125] A first width acquisition unit is used to acquire the fixed lane width carried in the vehicle network map information when there is only one lane on the road, and determine it as the lane width corresponding to each of the resampling points;

[0126] The second width acquisition unit is used to determine the lane width corresponding to each resampling point according to the distance between the resampling points of two adjacent lanes on the road when there are at least two lanes on the road.

[0127] Optionally, the width road acquisition module includes:

[0128] A road position acquisition unit, configured to calculate a position mean of each of the roads according to the resampled points of the lanes in the road to serve as the road position of each of the roads;

[0129] The opposite road acquisition unit is used to acquire the road corresponding to the opposite lane of each road according to the distance between the road position of each road and the road position of other roads and the distance threshold.

[0130] Optionally, the oncoming road acquisition unit includes:

[0131] A first road acquisition subunit is used to acquire a road from among the other roads whose road position is less than the distance threshold, as a road corresponding to an opposite lane of each of the roads;

[0132] The second road acquisition subunit is configured to mark the road corresponding to the opposite lane of each of the roads as empty when there is no road in the other roads whose distance to the road position of each of the roads is less than the distance threshold.

[0133] Optionally, the entry direction map acquisition module includes:

[0134] A reference line trajectory determination unit, configured to determine a reference line trajectory of each road according to the resampled points of the lanes in each road of the plurality of intersections;

[0135] A trajectory data acquisition unit, used to acquire the line segment, the line segment length and the reserved curve length between every two points on the reference line trajectory;

[0136] A straight line segment road creation unit, used to create an Opendrive straight line segment road according to the position information of the starting point of the line segment, the length of the line segment, the line segment slope of the line segment, the number of lanes of the road and the lane width of each resampling point;

[0137] A curve segment road creation unit, used to create an Opendrive curve segment road according to the reserved curve length, the number of lanes of the road and the lane width of each resampling point;

[0138] The entry direction map generating unit is used to add a predecessor road and a successor road respectively to the Opendrive straight segment road and the Opendrive curve segment road between the two resampling points, and generate an Opendrive format map of the entry direction road of each intersection.

[0139] Optionally, the intersection internal map generation module includes:

[0140] An opposite road acquisition unit, used for acquiring the roads of each approaching direction and the opposite lane of each intersection;

[0141] The intersection internal map generating unit is used to generate an Opendrive format map of the interior of each intersection according to the end segment road of each entry direction road and the start segment road of the opposite lane road of each intersection.

[0142] Optionally, the Opendrive map generation module includes:

[0143] An initial map composition unit, configured to compose initial Opendrive format maps of all intersections based on an Opendrive format map of the approaching road of each intersection and an Opendrive format map of the interior of each intersection;

[0144] The Opendrive map generation unit is used to optimize the initial Opendrive format map based on the Internet of Vehicles map information to obtain the Opendrive format maps of all intersections.

[0145] The map conversion device provided in the embodiment of the present application obtains road information of several intersections and the approaching roads of each intersection by parsing multiple vehicle network map information. The point data of the lane centerline of the lane in each road in the road information is interpolated and resampled to obtain the resampled points of the lane. The lane width corresponding to each resampled point is obtained, and the road corresponding to the opposite lane of each road is obtained based on the resampled point. Using a segment-by-segment fitting method, the roads between the two resampled points are fitted according to the lane width to obtain the Opendrive format maps of the approaching roads of several intersections. According to the roads of each approaching direction and the roads of its opposite lane of each intersection, an Opendrive format map inside the intersection of each intersection is generated. According to the Opendrive format map of the approaching direction road of each intersection and the Opendrive format map inside the intersection of each intersection, an Opendrive format map of all intersections is generated. The embodiment of the present application obtains vehicle network map data by parsing the vehicle network map information and resampling it, and then samples it segment by segment by fitting it, and constructs the final Opendrive format high-precision map segment by segment through the line segment roads between the point data, thereby achieving the purpose of converting the V2X map (vehicle network map) message into an Opendrive format high-precision map.

[0146] The present application also provides an electronic device, such as Figure 5 As shown, it includes a processor 501, a communication interface 502, a memory 503 and a communication bus 504, wherein the processor 501, the communication interface 502, and the memory 503 communicate with each other through the communication bus 504.

[0147] Memory 503, used for storing computer programs;

[0148] The processor 501 is used to implement the above map conversion method when executing the program stored in the memory 503.

[0149] The communication bus mentioned above can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. The communication bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, only one thick line is used in the figure, but it does not mean that there is only one bus or one type of bus.

[0150] The communication interface is used for communication between the above terminal and other devices.

[0151] The memory may include a random access memory (RAM) or a non-volatile memory, such as at least one disk memory. Optionally, the memory may also be at least one storage device located away from the aforementioned processor.

[0152] 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, discrete hardware components.

[0153] In another embodiment provided in the present application, a computer-readable storage medium is provided, in which instructions are stored. When the computer-readable storage medium is run on a computer, the computer executes the map conversion method described in any one of the above embodiments.

[0154] In another embodiment provided by the present application, a computer program product including instructions is also provided, on which a computer program is stored. When the computer program is run on a computer, the computer executes any of the map conversion methods described above.

[0155] For the aforementioned method embodiments, for the sake of simplicity, they are all described as a series of action combinations, but those skilled in the art should be aware that the present application is not limited by the order of the actions described, because according to the present application, some steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present application.

[0156] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0157] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of further restrictions, the elements defined by the sentence "including a..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0158] The above is a preferred embodiment of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles described in the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.< / successor> < / predecessor>

Claims

1. A map conversion method, characterized in that: The method comprises: Parsing multiple vehicle network map information to obtain road information of multiple intersections and the approaching direction road of each of the intersections; Interpolate and resample the point data of the lane centerline of each lane in the road in the road information to obtain resampled points of the lane; Obtaining a lane width corresponding to each of the resampling points, and obtaining a road corresponding to an opposite lane of each of the roads according to the resampling points; Using a segment-by-segment fitting method, the roads between the resampled points are fitted according to the lane width to obtain Opendrive format maps of the entry direction roads of the plurality of intersections; Generate an Opendrive format map of the interior of each intersection according to the roads in the respective entry directions and the roads in the opposite lanes of each intersection; According to the Opendrive format map of the approaching direction road of each intersection and the Opendrive format map inside each intersection, the Opendrive format map of all intersections is generated.

2. The method according to claim 1, characterized in that The interpolation resampling of the point data of the lane centerline of each lane in the road in the road information to obtain the resampled points of the lane includes: In the case where there are merged lanes and / or split lanes on the road, performing spline interpolation processing on the lanes on the road to obtain spline interpolated lanes on the road; The spline interpolation lanes in the road are resampled into a specified number of points to obtain the resampled points of the spline interpolation lanes.

3. The method according to claim 2, characterized in that The performing spline interpolation processing on the lanes in the road to obtain the spline interpolation lanes in the road includes: Acquire a reference lane in the road, where the reference lane is the longest lane in the road; Based on the point data of the reference lane, spline interpolation is performed on lane portions of other lanes in the road that do not have point data to obtain a point interpolation lane with the same length as the reference lane, and the reference lane and the point interpolation lane are used as the spline interpolation lane.

4. The method according to claim 1, characterized in that: The obtaining of the lane width corresponding to each of the resampling points includes: In the case where there is only one lane on the road, obtaining a fixed lane width carried in the vehicle networking map information, and determining the lane width corresponding to each of the resampling points; In the case that there are at least two lanes on the road, the lane width corresponding to each resampling point is determined according to the distance between the resampling points of two adjacent lanes on the road.

5. The method according to claim 1, characterized in that The acquiring, according to the resampling points, a road corresponding to an opposite lane of each of the roads comprises: Calculating the position mean of each road according to the resampled points of the lanes in the road to serve as the road position of each road; The road corresponding to the opposite lane of each road is obtained according to the distance between the road position of each road and the road position of other roads and a distance threshold.

6. The method according to claim 5, characterized in that The acquiring the road corresponding to the opposite lane of each road according to the distance between the road position of each road and the road position of other roads and the distance threshold comprises: Acquire a road from among the other roads, the distance between which and the road position of each of the roads is less than the distance threshold, as the road corresponding to the opposite lane of each of the roads; When there is no road in the other roads whose distance from the road position of each of the roads is less than the distance threshold, the road corresponding to the opposite lane of each of the roads is marked as empty.

7. The method according to claim 1, characterized in that The step-by-step fitting method is adopted to fit the roads between the resampled points in pairs according to the lane width to obtain Opendrive format maps of the entry direction roads of the plurality of intersections, including: Determining a reference line trajectory of each road according to the resampled points of the lanes in each road of the plurality of intersections; Obtaining the line segment, the line segment length and the reserved curve length between every two points on the reference line trajectory; Create an Opendrive straight segment road according to the position information of the starting point of the line segment, the length of the line segment, the line segment slope of the line segment, the number of lanes of the road and the lane width of each resampling point; Creating an Opendrive curve segment road according to the reserved curve length, the number of lanes of the road, and the lane width of each resampling point; Predecessor roads and successor roads are added to the Opendrive straight segment roads and Opendrive curve segment roads between the two resampling points, respectively, to generate an Opendrive format map of the approaching direction road of each intersection.

8. The method according to claim 1, characterized in that The method of generating an Opendrive format map of the interior of each intersection according to the roads in the respective directions of entry and the roads in the opposite lanes of each intersection comprises: Obtain the roads in the directions of each of the intersections and the roads in the opposite lanes; According to the end segment road of each approaching direction road of each intersection and the start segment road of the opposite lane road, an Opendrive format map inside each intersection is generated.

9. The method according to claim 1, characterized in that: The step of generating Opendrive format maps of all intersections according to the Opendrive format map of the road in the direction of entry of each intersection and the Opendrive format map inside each intersection comprises: According to the Opendrive format map of the approaching direction road of each intersection and the Opendrive format map inside each intersection, an initial Opendrive format map of all intersections is formed; The initial Opendrive format map is optimized based on the Internet of Vehicles map information to obtain Opendrive format maps of all intersections.

10. A map conversion device, characterized in that: The device comprises: A road information acquisition module is used to parse multiple vehicle network map information to obtain road information of several intersections and the approaching direction road of each of the intersections; A resampling point acquisition module, used for performing interpolation resampling on the point data of the lane centerline of each lane in the road in the road information to obtain the resampling point of the lane; A width road acquisition module, used to acquire the lane width corresponding to each of the resampling points, and acquire the road corresponding to the opposite lane of each of the roads according to the resampling points; An entry direction map acquisition module is used to fit the roads between the resampled points in pairs according to the lane width in a segment-by-segment fitting manner to obtain Opendrive format maps of the entry direction roads of the plurality of intersections; An intersection internal map generation module, used to generate an Opendrive format map of the interior of each intersection according to each of the entry direction roads and the roads of the opposite lanes of each intersection; The Opendrive map generation module is used to generate Opendrive format maps of all intersections based on the Opendrive format map of the entrance direction road of each intersection and the Opendrive format map inside each intersection.

11. An electronic device, comprising: A transceiver, a processor, a memory, and a program or instruction stored in the memory and executable on the processor, wherein the processor implements the map conversion method as described in any one of claims 1 to 9 when executing the program or instruction.

12. A readable storage medium having a program or instruction stored thereon, characterized in that: When the program or instruction is executed by a processor, the map conversion method according to any one of claims 1 to 9 is implemented.

Images