Method and device for generating road centerline in virtual scene, and storage medium

Abstract

This invention relates to a method, apparatus, and storage medium for generating road centerlines in a virtual scene. The method includes: generating discrete coordinate points of the road centerline in a third coordinate system based on vehicle position information in a first coordinate system and road information in a second coordinate system; decomposing the discrete coordinate points of the road centerline into several road elements; and generating the road centerline based on the geometric parameters and relationships of each road element. Embodiments of this invention enable the generation of road centerlines using basic road elements, ensuring that the generated road closely resembles the planar structure of a real road, maintaining the continuity of the generated road, and reducing the complexity of the generated road. This improves the accuracy of generating road centerlines in a virtual scene based on discrete coordinate points of the real road centerline.

Description

Technical Field

[0001] This invention relates to the automotive field, and more specifically, to a method, apparatus, and storage medium for generating road centerlines in a virtual scene. Background Technology

[0002] The performance of advanced driver assistance systems (ADAS) and autonomous driving systems heavily relies on the amount of testing mileage during their development and verification process. The construction of intelligent driving virtual scenarios can significantly improve the efficiency of these systems' development and verification. As the most crucial component of the virtual scenario, the accuracy of road data is paramount, and the fitting and generation of road centerlines is fundamental to determining this data.

[0003] Existing technologies combine global navigation satellite systems (e.g., GPS) and environmental perception systems (e.g., cameras, LiDAR) to fit road centerlines using cubic polynomials. However, road centerlines fitted by cubic polynomials not only do not conform to the geometric characteristics of the road itself, but also increase road complexity and computational costs when virtual vehicles are driving in the scene.

[0004] Therefore, a technology is needed that can fit the road centerline into simpler and more reasonable road elements. Summary of the Invention

[0005] The method, apparatus, and storage medium for generating road centerlines in a virtual scene according to the present invention can generate road centerlines using the most basic road elements, namely straight lines, circular curves, and spirals, by performing piecewise linear fitting on the local curvature of the road. The road centerlines generated by the method and apparatus according to one or more embodiments of the present invention conform to the geometric characteristics of the road itself while reducing the complexity of the road.

[0006] To achieve one or more of the above objectives, the present invention provides the following technical solutions. Specifically:

[0007] According to a first aspect of the present invention, a method for generating a road centerline in a virtual scene is provided. The method includes: generating discrete coordinate points of the road centerline in a third coordinate system based on vehicle position information in a first coordinate system and road information in a second coordinate system; decomposing the discrete coordinate points of the road centerline into a plurality of road elements; and generating the road centerline based on the geometric parameters and related relationships of each road element.

[0008] According to an embodiment of the present invention, the method further includes: determining the discrete coordinate points of the vehicle in the third coordinate system and the local heading angle based on the vehicle position information in the first coordinate system.

[0009] According to another embodiment or any of the above embodiments of the present invention, the method further includes: determining the discrete coordinate points of the road centerline in the second coordinate system based on road information in the second coordinate system.

[0010] According to another embodiment or any of the above embodiments of the present invention, the first coordinate system is an Earth coordinate system based on information collected by a global navigation satellite system; the second coordinate system is a vehicle coordinate system based on information collected by an environmental perception system; and the third coordinate system is a scene coordinate system with the initial position of the vehicle as the origin.

[0011] According to another embodiment or any of the above embodiments of the present invention, the type of road element includes straight lines, circular curves and spirals.

[0012] According to another embodiment or any of the above embodiments of the present invention, the method further includes: determining the distance and local curvature of discrete coordinate points along the road route direction.

[0013] According to another embodiment or any of the above embodiments of the present invention, the method further includes: obtaining segment breakpoint information of the piecewise linear fitting by performing piecewise linear fitting on the local curvature.

[0014] According to another embodiment or any of the above embodiments of the present invention, the method further includes: obtaining line segment breakpoint information by performing multi-objective optimization on the piecewise linear fitting result and the root mean square error of the local curvature, and minimizing the root mean square error.

[0015] According to another embodiment or any of the above embodiments of the present invention, the method further includes: obtaining line segment breakpoint information by determining a compensation error proportional to the number of line segments and minimizing the sum of the root mean square error and the compensation error.

[0016] According to another embodiment or any of the above embodiments of the present invention, the method further includes: updating the line segment breakpoint information according to a preset threshold, wherein the preset threshold includes a first threshold for the local curvature at the initial point and the termination point of the line segment; the preset threshold includes a second threshold for the difference between the local curvature at the initial point and the termination point of the line segment.

[0017] According to another embodiment or any of the above embodiments of the present invention, the method further includes: determining the road element type based on the line segment breakpoint information of each line segment.

[0018] According to another embodiment or any of the above embodiments of the present invention, the method further includes: determining translation parameters and rotation parameters, splicing two adjacent road elements; determining the overall translation parameters and rotation parameters of the spliced ​​road elements through bi-objective optimization, reconstructing the road centerline; and updating the information of each road element according to the overall translation parameters and rotation parameters.

[0019] According to a second aspect of the present invention, an apparatus for generating a road centerline in a virtual scene is provided. The apparatus includes: a calculation unit configured to generate discrete coordinate points of a road centerline in a third coordinate system based on vehicle position information in a first coordinate system and road information in a second coordinate system; a decomposition unit configured to decompose the discrete coordinate points of the road centerline into a plurality of road elements; and a generation unit configured to generate a road centerline based on the geometric parameters and related relationships of each road element.

[0020] According to an embodiment of the present invention, the calculation unit is further configured to: determine discrete coordinate points of the vehicle in a third coordinate system and local heading angle based on vehicle position information in a first coordinate system; and determine discrete coordinate points of the road centerline in a second coordinate system based on road information in a second coordinate system.

[0021] According to another embodiment or any of the above embodiments of the present invention, the first coordinate system is an Earth coordinate system based on information collected by a global navigation satellite system; the second coordinate system is a vehicle coordinate system based on information collected by an environmental perception system; and the third coordinate system is a scene coordinate system with the initial position of the vehicle as the origin.

[0022] According to another embodiment or any of the embodiments above, the types of road elements include straight lines, circular curves, and spirals.

[0023] According to another embodiment or any of the above embodiments of the apparatus, the decomposition unit is further configured to: obtain segment breakpoint information by performing piecewise linear fitting on the local curvature; and obtain segment breakpoint information by performing multi-objective optimization on the piecewise linear fitting result and the root mean square error of the local curvature, and minimizing the root mean square error; and obtain segment breakpoint information by determining a compensation error proportional to the number of segment, and minimizing the sum of the root mean square error and the compensation error.

[0024] According to another embodiment or any of the above embodiments of the present invention, the decomposition unit is further configured to: determine the road element type based on the line segment breakpoint information of each line segment.

[0025] According to another embodiment or any of the above embodiments of the apparatus, the generation unit is further configured to: determine translation parameters and rotation parameters, and splice two adjacent road elements; determine the overall translation parameters and rotation parameters of the spliced ​​road elements through bi-objective optimization, and reconstruct the road centerline; and update the information of each road element according to the overall translation parameters and rotation parameters.

[0026] According to a third aspect of the present invention, a computer-readable storage medium is provided, wherein instructions are stored therein, characterized in that, when executed by a processor, the instructions cause the processor to perform the method according to a first aspect of the present invention.

[0027] By utilizing one or more aspects of the present invention, the following effects can be achieved:

[0028] a) Road construction in traditional virtual scenes is configured manually. This invention proposes one or more embodiments that can be used for the construction of automated field roads;

[0029] (b) Traditional methods for generating road centerlines involve cubic linear fitting of discrete data points. One or more embodiments proposed in this invention fit discrete data points using three basic road elements: straight lines, circular curves, and spirals. This conforms to the real road structure specified in the national standard JTG B01-2014 "Technical Standards for Highway Engineering" and also satisfies the OpenDRIVE open road format, while simultaneously reducing the complexity of the generated road.

[0030] c) Traditional methods for generating road centerlines are not optimized and are usually segmented and fitted at certain distances. One or more embodiments proposed in this invention use an optimization algorithm in the process of generating road centerlines to ensure that the total error with the discrete coordinate points of the road centerlines actually collected is minimized, and the roads in the virtual scene can be accurately generated.

[0031] d) By using piecewise linear fitting of local curvature, the generated road is guaranteed to have continuity in G0 (point continuity), G1 (tangential continuity), and G2 (curvature continuity). Attached Figure Description

[0032] The above and / or other aspects and advantages of the present invention will become clearer and more readily understood from the following description taken in conjunction with the accompanying drawings, in which like or similar elements are denoted by the same reference numerals. The drawings include:

[0033] Figure 1 A flowchart illustrating a method for generating road centerlines for virtual scene construction according to an embodiment of the present invention is shown.

[0034] Figure 2 A schematic flowchart of a method for calculating discrete coordinate points of a road centerline according to an embodiment of the present invention is shown.

[0035] Figure 3 A schematic flowchart of a method for decomposing a road centerline into road elements according to an embodiment of the present invention is shown;

[0036] Figure 4 A schematic flowchart of a method for generating roads from road elements according to an embodiment of the present invention is shown;

[0037] Figure 5 A schematic diagram of coordinates and coordinate points according to an embodiment of the present invention is shown;

[0038] Figure 6 A schematic diagram showing the result of piecewise linear fitting of local curvature according to an embodiment of the present invention is illustrated. Detailed Implementation

[0039] In this specification, the invention is described more fully with reference to the accompanying drawings, which illustrate exemplary embodiments of the invention. However, the invention may be implemented in various forms and should not be construed as being limited to the embodiments given herein. The given embodiments are intended to make the disclosure herein complete and thorough, so as to more fully convey the scope of protection of the invention to those skilled in the art.

[0040] Terms such as "comprising" and "including" indicate that, in addition to the units and steps that are directly and explicitly stated in the specification and claims, the technical solution of the present invention does not exclude the presence of other units and steps that are not directly or explicitly stated. Terms such as "first" and "second" do not indicate the order of the units in terms of time, space, size, etc., but are merely used to distinguish the units.

[0041] The invention is described below with reference to flowchart illustrations, block diagrams, and / or flowcharts of methods and systems according to embodiments of the invention. It will be understood that each block of these flowchart illustrations and / or block diagrams, and combinations thereof, 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, or other programmable data processing apparatus to form a machine such that these instructions, executable by the processor of the computer or other programmable data processing apparatus, create components for implementing the functions / operations specified in these flowchart illustrations and / or blocks and / or one or more flowchart illustrations. It should also be noted that in some alternative implementations, the functions / operations shown in the blocks may not occur in the order shown in the flowchart. For example, two blocks shown sequentially may actually be executed substantially simultaneously, or these blocks may sometimes be executed in reverse order, depending on the functions / operations involved.

[0042] These computer program instructions may be loaded onto a computer or other programmable data processor to cause a series of operational steps to be performed on the computer or other programmable processor to constitute a computer-implemented process, such that the instructions, which execute on the computer or other programmable data processor, provide steps for implementing the functions or operations specified in one or more boxes of this flowchart and / or block diagram.

[0043] It should be noted that in this article, the term "multiple" refers to two or more. The term "lane" in this article refers to a path marked by lane lines, which typically only allows one vehicle to travel at a time (e.g., multiple vehicles are not allowed to travel side-by-side). The term "road" can include one or more lanes traveling in the same or opposite directions. The term "road width" refers to the cross-sectional area of ​​a road, consisting of the carriageway, median strip, and shoulders; that is, the width of the road surface. The term "road centerline" refers to a virtual, widthless, continuous line used to reflect the road's planar position and curvature. From a spatial geometric perspective, it is the characteristic line formed by connecting the center points of the road width sequentially from the road's starting point to its ending point, and generally coincides with the marking lines used to distinguish the travel spaces for vehicles traveling in both directions.

[0044] Now for reference Figure 1 , Figure 1 A flowchart illustrating a method 100 for generating road centerlines for virtual scene construction according to an embodiment of the present invention is shown.

[0045] Figure 1 In this process, the process begins at step S101. At step S102, based on the vehicle position information in the first coordinate system (e.g., Earth coordinate system) and the road information in the second coordinate system (e.g., Vehicle coordinate system), discrete coordinate points of the road centerline in the third coordinate system (e.g., Scene coordinate system) are generated.

[0046] Specifically, step S102 may optionally include, for example: Figure 2 One or more of the steps shown in the diagram.

[0047] In step S201, the discrete coordinates of the vehicle in a third coordinate system (e.g., the scene coordinate system) can be determined using the vehicle's longitude and latitude location information in a first coordinate system (e.g., the Earth coordinate system). The first coordinate system can be an Earth coordinate system based on information collected by a global navigation satellite system.

[0048] In practice, the Global Navigation Satellite System used can be any system, including GPS, and the acquired vehicle longitude and latitude information can be in any Earth coordinate system, including WGS84. After transformation from the Earth coordinate system to, for example, the scene coordinate system, using any method, the vehicle's longitude and latitude information can be converted into discrete coordinate points in, for example, the scene coordinate system. Generally, the scene coordinate system is a special case of a planar coordinate system, with its origin at the vehicle's initial position. Figure 5As shown, 501 is a schematic diagram of the coordinate axes of the scene coordinate system, and 502 is a schematic diagram of a discrete coordinate point of the vehicle after transformation from the first coordinate system (e.g., the Earth coordinate system) to the third coordinate system (e.g., the scene coordinate system).

[0049] In step S202, the local heading angle of the vehicle can be determined by the discrete coordinate points of the vehicle in the first coordinate system.

[0050] Specifically, the specified coordinate point and its N adjacent discrete coordinate points (e.g., positive integers greater than or equal to 1) are extracted. A linear fitting method is used to calculate a fitting formula in, for example, the scene coordinate system. The slope of this formula, expressed as an angle in two-dimensional space, represents the local heading angle at the specified coordinate point. This process can be repeated until the local heading angles of all discrete coordinate points are calculated. It should be noted that "before" and "after" a specified coordinate point refer to the front and rear directions according to the vehicle's direction of travel; if there are fewer than N data points before or after a specified coordinate point, all remaining coordinate points are extracted.

[0051] In step S203, the discrete coordinate points of the road centerline in the second coordinate system are calculated using the road information in the second coordinate system (e.g., the vehicle coordinate system).

[0052] Optionally, the environmental perception system employed can be any of the following: a camera, a lidar sensor, and one or more sensors. This system is used to acquire the coordinates of the road centerline in a second coordinate system. For example... Figure 5 As shown, 503 represents the coordinate axes of the second coordinate system (e.g., the vehicle coordinate system). For example, if a vehicle is traveling on the leftmost lane in the same direction, the marker point of the left lane line it collects is considered the center line of the road; if the vehicle is traveling on the second lane from the left in the same direction, the marker point of the second lane from the left it collects is generally considered the center line of the road, and so on.

[0053] In step S204, the discrete coordinates of the road centerline in the third coordinate system (e.g., the scene coordinate system) are calculated using the discrete coordinates of the vehicle and the local heading angle in the third coordinate system (e.g., the scene coordinate system) and the discrete coordinates of the road centerline in the second coordinate system (e.g., the vehicle coordinate system).

[0054] Optionally, the discrete vehicle coordinates in the third coordinate system obtained in step S201 can represent the translation relationship between the second and third coordinate systems, and the local heading angle obtained in step 202 can represent the rotation relationship between the second and third coordinate systems. Based on the translation and rotation relationships, the discrete coordinates of the road centerline in the second coordinate system obtained in step 203 can be transformed to the third coordinate system. Figure 5As shown, 504 is a schematic diagram of a discrete coordinate point of the road centerline in a third coordinate system (e.g., the scene coordinate system).

[0055] Continue back Figure 1 In step S103, the discrete coordinate points of the road centerline are decomposed into several road elements. The type of these road elements may optionally include straight lines, circular curves, and spirals. Optionally, step S103 may include... Figure 3 One or more of the steps shown in the diagram.

[0056] In step S301, the distance and local curvature of discrete coordinate points along the road route are determined using the discrete coordinate points of the road centerline. Optionally, the relative distance between two adjacent coordinate points is calculated from the start to the end of the road centerline, and then the relative distances between adjacent points are accumulated and summed to calculate the distance from each discrete coordinate point to the start point along the road route. The local curvature near a specified coordinate point can be calculated by specifying the coordinate point and the data of M (e.g., positive integers greater than or equal to 1) adjacent discrete coordinate points before and after it. It should be noted that "before" and "after" a specified coordinate point refer to the front and rear directions according to the vehicle's travel direction; if there are fewer than M data points before or after a specified coordinate point, all remaining coordinate points are extracted. The above operation can be repeated until the distance and local curvature of all discrete coordinate points along the road route are calculated. The distance along the road route can be labeled as X-axis data or X value, and the local curvature along the road route can be labeled as Y-axis data or Y value. Based on the X and Y values, an XY plane can be established, and the relationship between the local curvature and the distance along the road route can be obtained on the XY plane.

[0057] It should be noted that in this article, "segment of piecewise linear fitting" or "segment" refers to a line segment on the XY plane used to fit the XY relationship; while "road element" refers to a small section of road that can be transformed from the fitted line segment and used to describe the geometric characteristics of the road. The combination of multiple road elements can constitute the entire road.

[0058] In step S302, the number of line segments for piecewise linear fitting is initialized. Here, i can be used to refer to the number of line segments for piecewise linear fitting, and i is initialized to 1.

[0059] In step S303, the breakpoints of the piecewise linear fitting are calculated. Based on the current number of line segments i, the X value of each breakpoint is used as a variable. Multi-objective optimization is performed, using the root mean square error between the piecewise linear fitting result and the local curvature data as the optimization objective, to determine the X value of the breakpoint that minimizes the optimization objective. It should be noted that the piecewise linear fitting must ensure that adjacent line segments are connected end-to-end, guaranteeing the existence of a unique breakpoint between line segments.

[0060] In step S304, the error E1 of the piecewise linear fitting result is calculated. The result can be the root mean square error between the piecewise linear fitting result after multi-objective optimization and the local curvature data.

[0061] In step S305, a compensation error E2 proportional to the number of line segments is calculated. This compensation error is used to attempt to reduce the number of line segments used in the calculation process, so that the final fitting result maintains a balance between the number of line segments and the fitting result. According to a preset constant C1, the compensation error is proportional to i, and can be expressed as:

[0062] E2 = i × C1

[0063] In step S306, the total error is calculated. The total error is the sum of E1 and E2.

[0064] In step S307, it is determined whether the current total error is less than the total error of the previous cycle. If the determination result is "yes", it means that the total error can be further reduced, and then proceed to step S308; if the determination result is "no", it means that the number of line segments used in the current cycle and the calculated line segment breakpoints can minimize the total error, and then exit the calculation cycle and proceed to step S309.

[0065] It should be noted that when i equals 1, this judgment is not performed, and the result is directly assumed to be "yes," proceeding to step S308. It should also be noted that for some complex calculation cases, after satisfying the result of "no" in step S307, multiple additional loop calculations are needed to ensure that the total error remains minimum within a certain number of line segments. The aim is to minimize the total error by optimizing the number of line segments in the piecewise linear fitting; the specific method can be modified.

[0066] In step S308, the number of line segments is increased so that the current number of line segments is i+1. An attempt is made to further reduce the total error by increasing the number of line segments, and then steps S303 to S307 are repeated in the next loop.

[0067] In step S309, the segment breakpoint information of the piecewise fitting is recorded. For example... Figure 6 The diagram shows the result of a piecewise linear fitting, where 601 represents a line segment in the result; 602 represents a breakpoint in the result, which is also the starting point of 601; and 603 represents a breakpoint in the result, which is also the ending point of 601. Based on the piecewise linear fitting results, the X and Y values ​​of the starting and ending points of each line segment can be recorded, and each line segment can be numbered starting from 1, such as 1, 2, 3, etc., according to the order of the entire road from the starting point to the ending point.

[0068] In step S310, the line segment information is updated according to preset thresholds. The first preset threshold can be labeled T1. The Y-values ​​of the initial and final points of each line segment are compared with T1 (e.g., subtracted). Y-values ​​with an absolute value less than T1 are set to 0. Then, the second preset threshold can be labeled T2. The absolute value of the difference between the Y-values ​​of the initial and final points of each line segment is compared with T2 (e.g., subtracted). The Y-value of the final point of line segments with an absolute value less than T2 is set to the same value as the Y-value of the line segment's termination point.

[0069] It should be noted that, except for the starting point of the first line segment and the ending point of the last line segment, all other breakpoints are the ending point of the previous line segment and the starting point of the next line segment. When performing step S310, the line segments can be numbered in ascending order. When updating the Y value of the ending point of the previous line segment, the Y value of the starting point of the next line segment will also be updated to the same value.

[0070] In step S311, each line segment is traversed. One or more of the following parameters can be used as the values ​​extracted for each line segment during traversal:

[0071] a) Line segment number, labeled as ID;

[0072] b) The distance along the road route from the starting point of the line segment, which is the X value of the starting point, is marked as X1;

[0073] c) The distance along the road route from the end point of the line segment, which is the X value of the end point, is marked as X2;

[0074] d) The local curvature of the starting point of the line segment along the road route, that is, the Y value of the starting point, is marked as Y1;

[0075] e) The local curvature of the line segment termination point along the road route, which is the Y value of the termination point, is compared to Y2.

[0076] In step S312, it is determined whether the local curvature of the initial point and the ending point of the line segment is 0. If the determination result is "yes", proceed to step S313; if the determination result is "no", proceed to step S314.

[0077] In step S313, the road element type is determined to be a straight line. This means that the line segment with this ID represents a straight road element with a length of X2-X1 and arbitrary direction and starting point, and this road element segment will also be assigned the same ID.

[0078] In step S314, it is determined whether the local curvatures of the initial point and the ending point of the line segment are equal. If the determination result is "yes", proceed to step S315; if the determination result is "no", proceed to step S316.

[0079] In step S315, the road element type is determined to be a circular curve. This means that the line segment of this ID represents a curve with an arc length of X2-X1 and a radius of Y1. -1 A circular curve road element with an arbitrary center and starting point, and the number of this road element will also be assigned the same ID.

[0080] In step S316, the road element type is determined to be a spiral. This means that the line segment of this ID represents a curve with an arc length of X2-X1 and a starting curve radius of Y1. -1 The radius of the endpoint curve is Y2 -1 The spiral road element will also be assigned the same ID.

[0081] Specifically, the spiral can be an Euler spiral, and its basic formula is:

[0082] r × L = A 2

[0083] In the formula: r is the radius of the curve (m) at a point on the spiral.

[0084] L is the length of the curve (m) from a point on the spiral line to the origin.

[0085] A is the characteristic parameter of the spiral (m);

[0086] The characteristic parameter A of the spiral can be determined by X2-X1 and Y2-Y1, thus fixing the spiral's shape. Then, based on the values ​​of X1, X2, Y1, and Y2 and their interrelationships, a unique curve segment on the spiral can be calculated. The parameters of this curve segment represent the geometric parameters of the spiral's road elements.

[0087] In step S104, the road centerline is generated based on the geometric parameters and relationships of each road element. Optionally, step S104 includes... Figure 4 One or more of the steps shown.

[0088] In step S401, each road element is traversed. This involves obtaining the ID of each road element and its geometric parameters derived from 313, 315, and 316. Simultaneously, the road element information is initialized. The known data includes: the type of the road element, the curvature of its start and end points, and the distance between the start and end points along the road route. The unknown data includes: the coordinates of the start point (initial value: (0,0)); the coordinates of the end point (initial value calculated from the known data and the start point coordinates); the direction of the start point (initial value: 0); and the direction of the end point (initial value calculated from the known data and the direction of the start point).

[0089] In step S402, the translation and rotation parameters are calculated, and adjacent road elements are spliced ​​together.

[0090] Calculate the translation parameter T j To ensure that the starting point coordinates of the road element with ID j+1 (j>=1) are the same as the ending point coordinates of the road element j, making the road element j and the road element j+1 form a G0 continuum. Then, according to T... j Calculate and update the coordinates of the endpoint of road element j+1.

[0091] Calculate the rotation parameter R j To ensure that the starting direction of the road element with ID j+1 (j>=1) is the same as the ending direction of the road element j, making the road element j and the road element j+1 form a G1 continuum. Then, according to R... j Calculate and update the direction of the end point of the j+1 road element.

[0092] It should be noted that, according to the result of step S310, the curvature of the starting point of the road element with ID j+1 (j>=1) is the same as the curvature of the ending point of the road element j. Therefore, the road element j and the road element j+1 have formed a G2 continuity.

[0093] Optionally, step S402 can be repeated until all road elements are updated.

[0094] In step S403, the overall translation and rotation parameters are calculated to reconstruct the road centerline. Specifically, the entire road can be reconstructed based on the road elements pieced together in S402. The variables are the overall translation parameter T and the overall rotation parameter R. The optimization objective is the root mean square error between the reconstructed road centerline after overall translation and rotation and the discrete coordinate points of the original road centerline. A dual-objective optimization is performed to determine T and R, minimizing the value of the optimization objective.

[0095] In step S404, the information of each road element is updated based on the overall translation and rotation parameters. Specifically, the information of each road element in the entire road is updated according to the T and R values ​​obtained in S403. The obtained road element information can be output in OpenDRIVE format to represent a road in a virtual scene.

[0096] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein instructions are stored therein, which, when executed by a processor, cause the processor to perform a method for generating a road centerline according to one or more embodiments of the present invention.

[0097] The foregoing disclosure is not intended to limit this disclosure to the precise form disclosed or any particular field of use. Therefore, it is contemplated that various alternative embodiments and / or modifications of this disclosure are possible, whether expressly described or implied herein. Given that embodiments of this disclosure have been described as such, those skilled in the art will recognize that changes in form and detail may be made without departing from the scope of this disclosure. Therefore, this disclosure is limited only by the claims.

Claims

1. A method for generating road centerlines in a virtual scene, comprising: Based on the vehicle position information in the Earth coordinate system and the road information in the vehicle coordinate system, generate discrete coordinate points of the road centerline in the scene coordinate system; Determine the distance and local curvature of the discrete coordinate points along the road route; The local curvature is piecewise linearly fitted, and the root mean square error of the piecewise linear fitting result and the local curvature is optimized by multiple objectives to minimize the root mean square error to obtain the line segment breakpoint information. Based on the line segment breakpoint information, the discrete coordinate points of the road centerline are decomposed into several road elements, wherein the types of road elements include straight lines, circular curves, and spirals; and The road centerline is generated based on the geometric parameters and relationships of each road element.

2. The method according to claim 1, further comprising: Based on the vehicle position information in the Earth coordinate system, determine the discrete coordinate points and local heading angle of the vehicle in the scene coordinate system.

3. The method according to claim 1, further comprising: The discrete coordinate points of the road centerline in the vehicle coordinate system are determined based on the road information in the vehicle coordinate system.

4. The method according to claim 1, wherein, The Earth coordinate system is based on information collected by the Global Navigation Satellite System; The vehicle coordinate system is based on information collected by the environmental perception system; and The scene coordinate system uses the vehicle's initial position as its origin.

5. The method according to claim 1, further comprising: The line segment breakpoint information is obtained by determining a compensation error proportional to the number of line segments and minimizing the sum of the root mean square error and the compensation error.

6. The method according to claim 1 or 5, further comprising: The line segment breakpoint information is updated according to a preset threshold, wherein, The preset threshold includes a first threshold for the local curvature at the initial and final points of the line segment. The preset threshold includes a second threshold for the difference in local curvature at the initial and final points of a line segment.

7. The method according to claim 1, further comprising: Determine the translation and rotation parameters, and then merge two adjacent road elements; Through bi-objective optimization, the overall translation and rotation parameters of the spliced ​​road elements are determined, and the road centerline is reconstructed; and The information for each road element is updated based on the overall translation and rotation parameters.

8. An apparatus for generating road centerlines in a virtual scene, comprising: The computing unit is configured to generate discrete coordinate points of the road centerline in the scene coordinate system based on vehicle position information in the Earth coordinate system and road information in the vehicle coordinate system; Decomposition unit, configured as follows: Determine the distance and local curvature of the discrete coordinate points along the road route; The local curvature is piecewise linearly fitted, and the root mean square error of the piecewise linear fitting result and the local curvature is optimized by multiple objectives to minimize the root mean square error to obtain the line segment breakpoint information. Based on the line segment breakpoint information, the discrete coordinate points of the road centerline are decomposed into several road elements, wherein the types of road elements include straight lines, circular curves, and spirals; and The generation unit is configured to generate the road centerline based on the geometric parameters and relationships of each road element.

9. The apparatus of claim 8, wherein, The computing unit is also configured to: Based on the vehicle position information in the Earth coordinate system, determine the discrete coordinate points and local heading angle of the vehicle in the scene coordinate system; and The discrete coordinate points of the road centerline in the vehicle coordinate system are determined based on the road information in the vehicle coordinate system.

10. The apparatus according to claim 8, wherein, The Earth coordinate system is based on information collected by the Global Navigation Satellite System; The vehicle coordinate system is based on information collected by the environmental perception system; and The scene coordinate system uses the vehicle's initial position as its origin.

11. The apparatus of claim 8, wherein, The decomposition unit is further configured to: The line segment breakpoint information is obtained by determining a compensation error proportional to the number of line segments and minimizing the sum of the root mean square error and the compensation error.

12. The apparatus of claim 8, wherein, The generation unit is further configured to: Determine the translation and rotation parameters, and then merge two adjacent road elements; Through bi-objective optimization, the overall translation and rotation parameters of the spliced ​​road elements are determined, and the road centerline is reconstructed; and The information for each road element is updated based on the overall translation and rotation parameters.

13. A computer-readable storage medium storing instructions, characterized in that, When the instruction is executed by the processor, it causes the processor to perform the method as described in any one of claims 1-7.

Images