Lane representation method, lane representation device, storage medium, and program

Abstract

Embodiments of the present application provide a lane representation method, a lane representation device, a computer readable storage medium and a program. The lane representation method according to an embodiment of the present application comprises: obtaining a plurality of initial lane description curves describing a first lane on a first road in map data; extracting a plurality of initial sampling points on the plurality of initial lane description curves, the plurality of initial sampling points comprising at least a first end point of a first initial lane description curve and a last end point of a last initial lane description curve; performing curve fitting using at least the first end point and the last end point to generate a lane fitting curve; determining whether to update the generated lane fitting curve based on an error between each initial sampling point and a sampling point on the generated lane fitting curve to which the initial sampling point is respectively mapped, and obtaining a lane description curve of the first lane according to a determination result.

Description

Technical Field

[0001] This invention relates to the field of image processing, and more particularly to a lane representation method, a lane representation device, a computer-readable storage medium, and a computer program product. Background Technology

[0002] High-precision maps are primarily used in autonomous driving and advanced driver assistance systems (ADAS) for automobiles. They can accurately represent information such as the latitude and longitude coordinates of roads and other buildings, the number and width of lanes, road shapes, road markings, traffic signs, and roadblocks. In automotive electronic control units (ECUs) with high-precision map data processing capabilities, this data needs to be processed in real-time. In scenarios where real road data is used to construct virtual simulation test environments, especially in hardware-in-the-loop (HIL) testing simulations, the road scenes described by the high-precision map data need to be processed, constructed, and rendered in real-time by a real-time processor.

[0003] In existing technologies, the geometric features of road elements in high-definition maps are typically described using numerous and complex mathematical functions. When high-definition map data is processed in real-time hardware such as onboard electronic control units and processors in hardware-in-the-loop simulation devices, the cubic curves used to describe the road geometry consume significant computational resources, thus limiting the amount of high-definition map data the processor can handle and impacting the response speed and processing capabilities of real-time applications. Therefore, combining and simplifying high-definition map data while maintaining its geometric features and accuracy is crucial for improving its real-time performance on real-time hardware and expanding the road network range that real-time hardware can process.

[0004] Therefore, there is a need for a lane representation method, apparatus, computer-readable storage medium, and program that provides an accurate and concise representation of lane curves. Summary of the Invention

[0005] To address the aforementioned technical problems, according to one aspect of the present invention, a lane representation method is provided, comprising: acquiring a series of initial lane description curves describing a first lane on a first road from map data; extracting a plurality of initial sampling points on the plurality of initial lane description curves, wherein the plurality of initial sampling points at least include the first endpoint of the first initial lane description curve and the last endpoint of the last initial lane description curve; performing curve fitting using at least the first endpoint and the last endpoint to generate a lane fitting curve; determining whether to update the generated lane fitting curve based on the error between each initial sampling point and its respective sampling point mapped onto the generated lane fitting curve; and obtaining the lane description curve of the first lane according to the determination result.

[0006] According to another aspect of the present invention, a lane representation method is provided, comprising: acquiring road description curves describing a first road in map data; and splitting the road description curves of the first road in response to differences in the number of lanes, lane connection relationships, or lane description curve properties; or acquiring road description curves describing a first road and a connected second road in map data; and merging the road description curves of the first road and the second road in response to the same number of lanes, lane connection relationships, and lane description curve properties between the road description curves of the first road and the road description curves of the second road.

[0007] According to another aspect of the present invention, a lane representation device is provided, comprising: an acquisition unit configured to acquire a plurality of consecutive initial lane description curves describing a first lane on a first road from map data; an extraction unit configured to extract a plurality of initial sampling points on the plurality of initial lane description curves, wherein the plurality of initial sampling points include at least the first endpoint of the first initial lane description curve and the last endpoint of the last initial lane description curve; a fitting unit configured to perform curve fitting using at least the first endpoint and the last endpoint to generate a lane fitting curve; and a description unit configured to determine whether to update the generated lane fitting curve based on the error between each initial sampling point and a sampling point mapped to the generated lane fitting curve, and to acquire the lane description curve of the first lane according to the determination result.

[0008] According to another aspect of the present invention, a lane representation device is provided, comprising: a processor; and a memory storing computer program instructions, wherein, when the computer program instructions are executed by the processor, the processor performs the following steps: acquiring a series of initial lane description curves describing a first lane on a first road from map data; extracting a plurality of initial sampling points on the plurality of initial lane description curves, the plurality of initial sampling points including at least the first endpoint of the first initial lane description curve and the last endpoint of the last initial lane description curve; performing curve fitting using at least the first endpoint and the last endpoint to generate a lane fitting curve; determining whether to update the generated lane fitting curve based on the error between each initial sampling point and a sampling point mapped to the generated lane fitting curve, and obtaining the lane description curve of the first lane according to the determination result.

[0009] According to another aspect of the present invention, a computer-readable storage medium is provided having computer program instructions stored thereon, wherein the computer program instructions, when executed by a processor, perform the following steps: acquiring a series of initial lane description curves describing a first lane on a first road from map data; extracting a plurality of initial sampling points on the plurality of initial lane description curves, the plurality of initial sampling points including at least the first endpoint of the first initial lane description curve and the last endpoint of the last initial lane description curve; performing curve fitting using at least the first endpoint and the last endpoint to generate a lane fitting curve; determining whether to update the generated lane fitting curve based on the error between each initial sampling point and its respective sampling point mapped onto the generated lane fitting curve; and obtaining the lane description curve of the first lane according to the determination result.

[0010] According to another aspect of the present invention, a computer program product is provided, comprising computer program instructions, wherein the computer program instructions, when executed by a processor, implement the steps of the lane representation method as described in any of the preceding claims.

[0011] According to the lane representation method, apparatus, computer-readable storage medium, and program of the present invention, a lane fitting curve can be generated by using multiple initial sampling points on multiple initial lane description curves. Therefore, high-precision map data can be processed offline by computer software or programs, effectively reducing the amount of high-precision map data and improving its efficiency while ensuring a set accuracy threshold. In applications such as autonomous driving where high-precision map real-time processing is required, the computational resources required for data processing can be effectively reduced.

[0012] In software-in-the-loop testing, model-in-the-loop testing, or hardware-in-the-loop testing scenarios that use high-precision map data to generate virtual simulation scenarios, the high-precision map data generated by the above processing can significantly reduce the configuration time of the simulation environment and the computing resources of real-time or non-real-time processors occupied by high-precision map generation and rendering, thereby improving the efficiency of simulation testing while increasing the range of roads available for simulation testing. Attached Figure Description

[0013] The above and other objects, features, and advantages of the present invention will become clearer from the detailed description of the embodiments of the present invention in conjunction with the accompanying drawings.

[0014] Figure 1 A flowchart of a lane representation method according to an embodiment of the present invention is shown;

[0015] Figure 2 A flowchart of a lane representation method according to an embodiment of the present invention is shown;

[0016] Figure 3A schematic diagram illustrating the lane representation of a first road according to an embodiment of the present invention is shown;

[0017] Figure 4 The process of generating a lane description curve for a first lane according to an embodiment of the present invention is illustrated;

[0018] Figure 5 The process of generating a width description curve for a first lane according to an embodiment of the present invention is illustrated;

[0019] Figure 6 This diagram illustrates a road description curve processing method for a first road according to an embodiment of the present invention.

[0020] Figure 7 This illustration shows a schematic diagram of processing a first lane and left and right shoulders on a first road, according to an embodiment of the present invention.

[0021] Figure 8 A schematic diagram of a road processing procedure is shown as an example of an embodiment of the present invention.

[0022] Figure 9 A schematic diagram of road processing according to an embodiment of the present invention is shown;

[0023] Figure 10 A block diagram of a lane indication device according to an embodiment of the present invention is shown;

[0024] Figure 11 A block diagram of a lane indication device according to an embodiment of the present invention is shown. Detailed Implementation

[0025] Lane indication methods, apparatus, computer-readable storage media, and programs according to embodiments of the present invention will now be described with reference to the accompanying drawings. In the drawings, the same reference numerals denote the same elements throughout. It should be understood that the embodiments described herein are merely illustrative and should not be construed as limiting the scope of the invention.

[0026] In this embodiment of the invention, geometric features such as coordinate systems, road reference lines, lanes, and lane offsets in high-precision maps can be defined using data format standards including but not limited to ASAM OpenDRIVE. Typically, multiple contiguous cubic curves can be used to describe road reference lines, and cubic polynomials can be used to describe, for example, the lane offset of the center lane (with zero width on the far left of the road) relative to the road reference line. For the first lane, its shape can be characterized by its width from the center lane along the road reference line direction. Subsequent lanes can be characterized by their width from the right boundary of the preceding lane along the road reference line direction. A road in the map data can contain one or more lanes.

[0027] Regarding the specific format of map data, particularly road and lane data, the complex polynomial curves, width formulas, and various offset representations consume a large amount of computing resources, thus greatly limiting the efficiency and quantity of data processing.

[0028] To address the issues of high processor performance requirements and large computational resource consumption when processing raw high-precision map data in real-time hardware, this invention proposes a lane representation method and apparatus that can process a larger amount of high-precision map data without increasing processor resources or consume less processor resources with a specific amount of high-precision map data.

[0029] Figure 1 A flowchart of a lane indication method 100 according to an embodiment of the present invention is shown. Referring below... Figure 1 A lane representation method according to an embodiment of the present invention is described.

[0030] In step S101, multiple consecutive initial lane description curves describing the first lane on the first road are obtained from the map data.

[0031] In this embodiment of the invention, the initial lane description curves of the first lane of the first road in the high-precision map data can be obtained first. For example, each segment of the initial lane description curve can be represented by a parametric cubic curve.

[0032] In step S102, multiple initial sampling points are extracted from the multiple initial lane description curves, and the multiple initial sampling points include at least the first endpoint of the first initial lane description curve and the last endpoint of the last initial lane description curve.

[0033] In this embodiment of the invention, extracting multiple initial sampling points on the multiple initial lane description curves may include: extracting the multiple initial sampling points on the multiple initial lane description curves at equal intervals based on curve interpolation parameters; or, extracting the multiple initial sampling points on the multiple initial lane description curves at equal intervals based on distance.

[0034] In this embodiment of the invention, since the multiple initial lane description curves are multiple lane initial description curves connected end to end, the multiple initial sampling points may include at least the first endpoint and the last endpoint of these curves. That is, the multiple initial sampling points may include at least the first endpoint of the first initial lane description curve and the last endpoint of the last initial lane description curve.

[0035] In step S103, at least the first endpoint and the last endpoint are used to perform curve fitting to generate a lane fitting curve.

[0036] In an embodiment of the present invention, curve fitting using at least the first endpoint and the last endpoint to generate a lane fitting curve may include: using at least the first endpoint and the last endpoint as the endpoints of the lane fitting curve, converting the global coordinate system to a local coordinate system with the first endpoint as the origin, and performing curve fitting by calculating the coordinates of the endpoints of the lane fitting curve in the local coordinate system of the lane fitting curve and the derivatives of the coordinates.

[0037] In step S104, based on the error between each initial sampling point and its respective sampling point mapped to the generated lane fitting curve, it is determined whether to update the generated lane fitting curve, and the lane description curve of the first lane is obtained according to the determination result.

[0038] In one example of an embodiment of the present invention, iteratively determining whether initial sampling points need to be used as endpoints of the lane fitting curve for lane fitting can be performed. Optionally, the steps may include: Step S1: For the generated lane fitting curve, obtaining the lane error between each initial sampling point and the corresponding mapped sampling point; Step S2: In response to the lane error between each initial sampling point and the corresponding mapped sampling point being less than a lane threshold, determining not to update the generated lane fitting curve, and using the generated lane fitting curve as the lane description curve of the first lane; Step S3: In response to the existence of a lane error value greater than the lane threshold between each initial sampling point and the corresponding mapped sampling point, supplementing the initial sampling point corresponding to the largest lane error value as an endpoint of the lane fitting curve; Step S4: Using the endpoint as a supplemented endpoint and the first endpoint of its respective lane fitting curve as the endpoint of the lane fitting curve for curve fitting, and using the endpoint as a supplemented first endpoint and the end point of its respective lane fitting curve as the endpoint of the lane fitting curve for curve fitting, and performing steps S1-S4 on the generated lane fitting curve.

[0039] In another example of this embodiment of the invention, optionally, the width description curve of the first lane can be obtained by simplifying the process using an initial width description curve in the map data used to describe the width of the first lane. Specifically, this may include: obtaining multiple consecutive initial width description curves in the map data describing the width of the first lane; extracting multiple initial sampling points on the multiple initial width description curves, wherein the multiple initial sampling points include at least the first endpoint of the first initial width description curve and the last endpoint of the last initial width description curve; performing curve fitting using at least the first endpoint and the last endpoint to generate a width fitting curve; determining whether to update the generated width fitting curve based on the error between each initial sampling point and the sampling point mapped to the generated width fitting curve, and obtaining the width description curve of the first lane based on the determination result.

[0040] Similar to the lane description curve described above, the fitting process for the width description curve can also be iteratively determined whether the initial sampling points need to be used as the endpoints of the width fitting curve. Optionally, the steps may include: Step S5: For the generated width fitting curve, obtain the width error between each initial sampling point and the corresponding mapped sampling point; Step S6: In response to the width error between each initial sampling point and the corresponding mapped sampling point being less than the width threshold, determine not to update the generated width fitting curve, and use the generated width fitting curve as the width description curve of the first lane; Step S7: In response to the width error between each initial sampling point and the corresponding mapped sampling point having a width error value greater than the width threshold, supplement the initial sampling point corresponding to the largest width error value as the endpoint of the width fitting curve; Step S4: Perform curve fitting with the endpoint as the supplemented end point and the first endpoint of its respective width fitting curve as the endpoint of the width fitting curve, and perform curve fitting with the endpoint as the supplemented first endpoint and the end point of its respective width fitting curve as the endpoint of the width fitting curve, and perform steps S5-S8 on the generated width fitting curve.

[0041] In another embodiment of the present invention, optionally, in response to the fact that the representation data of the first lane's continuous multiple initial lane description curves includes lane offset data components relative to the reference lane line in the map data, the initial lane description curve of the first lane is corrected to the reference lane line in the map data to remove the lane offset data components. That is, if the high-precision map data does not use the aforementioned center lane (such as the first lane in this embodiment of the present invention) as a road reference line, the lane offset data components of the first lane relative to the road reference line, described by multiple cubic polynomials, can be read from the original high-precision map data. Combined with this offset data component, the coordinate information of each sampling point on the first lane can be calculated using the coordinate information of the sampling points on the road reference line. Thus, parametric cubic curve fitting is performed using the coordinate information of each sampling point on the first lane to realize the use of the first lane as a road reference line. At this time, when the first lane is used as a road reference line, the first lane coincides with the road reference line, thereby removing all cubic polynomial data in the high-precision map data used to describe the lane offset data components of the first lane relative to the road reference line.

[0042] Furthermore, in this embodiment of the invention, after obtaining the lane description curve and width description curve of the first lane, optionally, a road description curve of the first lane can be generated based on the lane description curve and width description curve of the first lane on the first road. Additionally, for cases where the first road has multiple lanes, the width description curves of multiple lanes can be obtained separately, so that a road description curve of the first road can be generated based on the width description curves corresponding to one or more lanes adjacent to the first lane on the first road.

[0043] In another embodiment of the present invention, considering the map scenario for autonomous driving, the left and right shoulders that are inaccessible on the first road may not need to be included in the scenario, thus the road description curve of the first road can be further simplified. Therefore, the method may further include: removing the description curves of the first shoulder and / or the second shoulder on both sides of the first road from the road description curve of the first road. In one example, when the first lane in the first road is the left shoulder lane of the first road, and the lane description curve of the first lane is the left boundary of the left shoulder lane, when deleting the left shoulder lane, it is necessary to shift the lane description curve of the first lane from the left boundary of the left shoulder lane to the right boundary of the left shoulder lane using the width information of the left shoulder lane. In this example, the right shoulder lane can be directly deleted without affecting other limitations in the first road.

[0044] In another example, when the first lane in the first road is the right shoulder lane and the lane description curve of the first lane is the right boundary of the right shoulder lane, deleting the right shoulder lane requires shifting the lane description curve of the first lane from the right boundary of the right shoulder lane to the left boundary of the right shoulder lane using the width information of the right shoulder lane. In this example, the left shoulder lane can be deleted directly without affecting other constraints in the first road.

[0045] In another example, when the lane description curve of the first lane in the first road is not the left boundary of the left shoulder lane or the right boundary of the right shoulder lane, both the left shoulder lane and the right shoulder lane can be directly deleted.

[0046] In this embodiment of the invention, when high-precision map data contains one or more roads connected sequentially between two adjacent intersections, these roads can be combined into one road to achieve road combination and compression, thereby reducing the number of roads between two adjacent intersections and the number of parametric cubic curves used for description. Furthermore, when features such as the number of lanes, lane connectivity, or lane description curve properties within a road change, the road can be segmented into multiple roads at the point of feature change for processing, ensuring the accuracy of the road description curves.

[0047] According to the lane representation method of the present invention, a lane fitting curve can be generated by using multiple initial sampling points on multiple initial lane description curves. Therefore, high-precision map data can be processed offline by computer software or programs, effectively reducing the amount of high-precision map data and improving its efficiency while ensuring a set accuracy threshold. In applications such as autonomous driving where high-precision map real-time processing is required, the computational resources required for data processing can be effectively reduced.

[0048] In software-in-the-loop testing, model-in-the-loop testing, or hardware-in-the-loop testing scenarios that use high-precision map data to generate virtual simulation scenarios, the high-precision map data generated by the above processing can significantly reduce the configuration time of the simulation environment and the computing resources of real-time or non-real-time processors occupied by high-precision map generation and rendering, thereby improving the efficiency of simulation testing while increasing the range of roads available for simulation testing.

[0049] Figure 2 A flowchart of a lane representation method 200 according to an embodiment of the present invention is shown. Referring below... Figure 2 A lane representation method according to an embodiment of the present invention is described.

[0050] In step S201, the road description curve describing the first road is obtained from the map data.

[0051] In this embodiment of the invention, the road description curve for the first road can be obtained first from the high-precision map data. For example, the road description curve can be represented by a parametric cubic curve.

[0052] In step S202, in response to differences in the number of lanes, lane connection relationships, or lane description curve properties of the first road, the road description curve of the first road is split; or the road description curve of the second road connected to the first road is obtained from the map data, and in response to the fact that the number of lanes, lane connection relationships, and lane description curve properties of the road description curves of the first road and the second road are the same, the road description curves of the first road and the second road are merged.

[0053] In this embodiment of the invention, if the number of lanes, lane connection relationships, or lane description curve properties of the first road differ, the road description curve of the first road can be split at feature transformation points. For example, when the number of lanes in the road description curve of the first road changes, the first road can be split into a second road and a third road at the location where the number of lanes changes. Furthermore, when the lane connection relationships in the road description curve of the first road change, such as when a second lane connects to a third lane in the next segment, the road can be split at the location where the lane connection relationships change. As another example, when the properties of the lane description curve in the road description curve of the first road change, such as when the lane description curve changes from a dashed line to a straight line, the road can be split at the location where the properties of the lane description curve change.

[0054] Furthermore, in this embodiment of the invention, when the number of lanes, lane connection relationship, and lane description curve properties are the same between the second road and the first road, the first road and the second road can be merged and treated as a single road for subsequent operations.

[0055] Figure 2 The specific operation steps in the lane representation method of the present invention shown in this embodiment can be applied to... Figure 1 Before the lane representation method shown, in order to Figure 1 Before performing lane curve fitting, roads are first split or merged to obtain clearer road division relationships, which facilitates subsequent lane curve fitting operations. Of course, Figure 2 The specific operational steps in the lane representation method of the present invention shown in this embodiment can also be applied to... Figure 1 Following the lane representation method shown, or during its operation, to obtain more concise and accurate lane representation data.

[0056] According to the lane representation method of the present invention, roads can be merged or divided according to various properties of lane lines, so as to effectively reduce the amount of data of high-precision maps and improve their usage efficiency while ensuring the set accuracy threshold. In situations such as autonomous driving where there are high-precision map and high real-time processing requirements, the computing resources required for data processing can be effectively reduced.

[0057] The following provides an example of a lane representation method according to an embodiment of the present invention.

[0058] Figure 3 A schematic diagram illustrating the lane representation of a first road according to an embodiment of the present invention is shown. Figure 3As shown, the road reference line on the first road includes parametric cubic curves 221 and 222, which are located in their respective local UV coordinate systems. Curve 221 is located in local UV coordinate system 210 and is described by a cubic polynomial in both the u and v directions. This local UV coordinate system 210 is located in global XY coordinate system 200 and has a certain rotation angle relative to global XY coordinate system 200. The first and last endpoints of the road reference line are 201 and 203, respectively, and the connection point of curves 221 and 222 is 202. The parametric cubic curve contains a cubic polynomial in both the u and v directions.

[0059]

[0060] Where aU, bU, cU, and dU are the coefficients of the cubic polynomial in the u direction, and aV, bV, cV, and dV are the coefficients of the cubic polynomial in the v direction. A p value of 0 indicates the first endpoint of the parametric cubic curve, and a p value of 1 indicates the last endpoint of the parametric cubic curve.

[0061] exist Figure 3 In the diagram, the first lane is described by two cubic polynomials 223 and 224. For the portion of the first lane described by 223, the input of the cubic polynomial is the distance of any point on the first lane relative to the starting endpoint 204 along the road reference line, and the output is the offset of the point on the road reference line from the first lane. The lane offset is described by a cubic polynomial:

[0062] Offset(s) = a + b * Δs + c * Δs 2 +d*Δs 3 (2)

[0063] Where Δs = s - sOffset i Let be the distance from any point on the lane offset cubic polynomial segment relative to the starting point of that segment along the road reference line, and let s be the distance from any point on the lane offset cubic polynomial segment relative to the starting point of the first lane offset cubic polynomial segment along the road reference line. Let sOffset be the distance from any point on the lane offset cubic polynomial segment relative to the starting point of the first lane offset cubic polynomial segment along the road reference line. i This is the distance along the road reference line from the starting point of the cubic polynomial segment of lane offset to the starting point of the first cubic polynomial segment of lane offset.

[0064] The left boundary of the first lane is composed of curves 223 and 224, and its right boundary is determined by a multi-segment cubic polynomial describing the width of the lane from the left boundary along the road reference line normal.

[0065] The left boundary of the second lane adjacent to the first lane is the right boundary of the first lane, such as... Figure 3As shown, the right boundary of the second lane is described by cubic polynomials 225 and 226. For the lane width portion described by 225, the input of the cubic polynomial is the distance of any point on the second lane relative to the starting point 207 along the road reference line, and the output is the width of the second lane at that point.

[0066] Width(s) = a + b * Δs + c * Δs 2 +d*Δs 3 (3)

[0067] Where Δs = s - sWidth i Let s be the distance from any point on the second lane width cubic polynomial segment relative to the starting point of the second lane width cubic polynomial segment along the road reference line, and let s be the distance from any point on the second lane width cubic polynomial segment relative to the starting point of the first second lane width cubic polynomial segment along the road reference line. i The distance along the road reference line is the distance from the starting point of the cubic polynomial segment of the second lane width to the starting point of the first cubic polynomial segment of the second lane width.

[0068] In one example of an embodiment of the present invention, the lane description curve of the first lane on the first road in the embodiment of the present invention can be optimized by compressing multiple parametric cubic curves into a new and fewer parametric cubic curves.

[0069] Figure 4 The process of generating a lane description curve for a first lane according to an embodiment of the present invention is illustrated. Figure 4 As shown, in step S401, the starting coordinates and coefficients, local coordinate system orientation angle, and curve length of each parametric cubic curve segment of the first lane are read from the high-precision map data. Specifically, information describing each parametric cubic curve segment of the first lane can be read from the high-precision map data. For the i-th parametric cubic curve segment, the information may include: the distance s from the starting point of the parametric cubic curve segment along the first lane relative to the starting point of the entire road segment. i The coordinates (x, y) of the origin of the local UV coordinate system of the cubic curve with respect to the global XY coordinate system. i_0 ,y i_0 The rotation angle θ of the local UV coordinate system relative to the global XY coordinate system of the cubic curve with respect to the parameter. i_heading The length l of the parametric cubic curve i The cubic polynomial coefficients aU of the parametric cubic curve in the u direction. i bU i cU i dU i The cubic polynomial coefficients aV of the parametric cubic curve in the v direction i bVi cV i dV i .

[0070] In step S402, sampling points can be extracted along the first lane according to a specific pattern, and the coordinate information of the sampling points in the global XY coordinate system can be obtained. Optionally, sampling can be performed along the first lane to generate a series of initial sampling points for the first lane. To ensure the continuity of connected roads, the starting and ending points of the first lane can optionally be always included in the above sampling points.

[0071] Optionally, the generated series of initial sampling points for the first lane can be selected at equal intervals in the interval from 0 to 1 for the independent variable p of the aforementioned cubic curve. In this case, all sampling points are not equally spaced along the road reference line. Alternatively, the generated series of initial sampling points for the first lane can be interpolated by the parameter p using the length function of the cubic curve, thereby ensuring that the p values ​​of the initial sampling points are not equally spaced in the interval from 0 to 1, but all initial sampling points are equally spaced along the road reference line.

[0072] Based on the extracted initial sampling points of the first lane, two sets of sampling points can be created: the endpoint set of the lane fitting curve for the first lane and the error verification point set for the first lane. The initial endpoint set contains at least two sampling points: the first endpoint and the last endpoint. The error verification point set for the first lane consists of all the initial sampling points of the first lane except for the first and last endpoints. That is:

[0073] In step S403, a set of endpoints for the lane fitting curve is created, which initially includes the first endpoint and the last endpoint of the first lane.

[0074] In step S404, an error verification point set is created, initially consisting of all initial sampling points other than the endpoint set of the lane fitting curve.

[0075] Subsequently, in step S405, the coordinate information of the sampling points in the endpoint set is used to create a parametric cubic curve segment by segment.

[0076] Specifically, the coordinates of each road reference line sampling point in the local UV coordinate system can be calculated based on the coefficients of the parametric cubic curve. For the sampling point on the i-th segment of the parametric cubic curve, its coordinates in the local UV coordinate system are:

[0077]

[0078] Where 0≤p≤1, p=0 represents the first endpoint of the parametric cubic curve segment, and p=1 represents the last endpoint of the parametric cubic curve segment.

[0079] To ensure the smoothness of each curve segment, the derivatives of the horizontal and vertical coordinates of each initial sampling point in the local UV coordinate system can be calculated simultaneously based on the parametric cubic curve corresponding to that initial sampling point. For a sampling point on the i-th parametric cubic curve, its derivative in the local UV coordinate system is:

[0080]

[0081] Based on the coordinates (u) of each initial sampling point in the local UV coordinate system i (p),v i (p) The coordinates (x) of the origin of the local UV coordinate system of the cubic curve corresponding to the initial sampling point in the global XY coordinate system. i_0 ,y i_0 and rotation angle θ i_headi The coordinates (x, y) of the sampling point in the global XY coordinate system can be calculated.

[0082]

[0083] The rotation angle θ of the local UV coordinate system relative to the global XY coordinate system based on the parametric cubic curve. i_heading The derivative of the initial sampling point on the curve in the local UV coordinate system can be used to calculate the slope of the tangent line at the initial sampling point on the curve in the global XY coordinate system.

[0084]

[0085] If the first lane contains n parametric cubic curves, then during the initial parametric cubic curve fitting, the set of endpoints of the fitted curve for the first lane only includes the first endpoint and the last endpoint of the first lane. These two initial sampling points are also the first endpoint of the first parametric cubic curve and the last endpoint of the nth parametric cubic curve.

[0086] At this point, the first endpoint of the initial sampling points can be used as the origin of the fitted lane curve, i.e., the local UV coordinate system of the parametric cubic curve. The slope of the tangent line at the first endpoint in the global XY coordinate system is used as the rotation angle of the local UV coordinate system of the fitted parametric cubic curve relative to the global XY coordinate system. Then, the coordinates and derivatives of the first endpoint are respectively...

[0087]

[0088] Transforming the coordinates and derivatives of the end points in the initial sampling points from the global XY coordinate system to the local UV coordinate system defined by the first endpoint, the coordinates and derivatives of the end points in the local UV coordinate system of the fitted parametric cubic curve are as follows:

[0089]

[0090] Where Δθ heading =θ heading_start -θ heading_end .

[0091] Using the coordinates and derivatives of two initial sampling points (the first and last endpoints), a new cubic polynomial can be constructed in the u and v directions respectively using cubic Schmidt spline interpolation to represent the fitted lane curve:

[0092]

[0093] In step S406, the error between each initial sampling point in the error verification point set and the sampling point mapped on the lane fitting curve is calculated.

[0094] At this point, all initial sampling points in the first lane error verification point set can be transformed to the local UV coordinate system defined by the first endpoint using the method described in equation (10). If the transformed coordinates are (u actual ,v actual If so, the p-value can be obtained using the following method:

[0095] u(p) = aU + bU*p + cU*p 2 +dU*p 3 =u actual (13)

[0096] If the value of p obtained according to the method in equation (13) is p actual The error between the initial sampling point and the sampling point mapped to the lane fitting curve is:

[0097] error = |v(p) acutal )-v actual | (14)

[0098] In step S407, it is determined whether there are any initial sampling points with errors greater than a threshold. If the errors of all initial sampling points in the first lane error verification point set are not greater than the set lane threshold, then proceed to step S409, where the fitted lane curve is used as the lane description curve to represent the original cubic curve of multiple parameters.

[0099] If, in step S407, there are sampling points in the error verification point set of the first lane whose error is greater than the set lane threshold, then proceed to step S408, remove the initial sampling point with the largest error value from the error verification point set of the first lane, add the initial sampling point to the endpoint set of the lane fitting curve of the first lane, and return to step S405 to re-perform curve fitting.

[0100] During the re-curve fitting process, a parametric cubic curve can be generated from every two adjacent initial sampling points, which serve as endpoints. This curve uses the previous initial sampling point as the origin of the local UV coordinate system of the newly generated parametric cubic curve, and uses the slope of the tangent line of the previous initial sampling point in the global XY coordinate system as the rotation angle of the local UV coordinate system of the newly generated parametric cubic curve relative to the global XY coordinate system. When generating a parametric cubic curve from every two adjacent initial sampling points in the endpoint set, the coordinates and derivatives of the latter initial sampling point need to be transformed from the global XY coordinate system to the local UV coordinate system defined by the former initial sampling point. After generating a parametric cubic curve from every two adjacent initial sampling points in the endpoint set, the coordinates of all initial sampling points in the error verification point set located between these two initial sampling points need to be transformed to the local UV coordinate system defined by the newly generated parametric cubic curve, and then error verification is performed.

[0101] In step S408, selecting the initial sampling point with the largest error as the endpoint for continued fitting is merely an example. In practical applications, multiple initial sampling points can be selected as endpoints for curve fitting based on the relationship between the mapped sampling points and the lane threshold; this is not a limitation.

[0102] Finally, until step S407 determines that the errors of all initial sampling points are not greater than the set lane threshold, then proceed to step S409 and end.

[0103] According to the method of this example embodiment of the present invention, the number of curves included in the generated lane description curve of the first lane will be less than the number of initial lane description curves used to describe the first lane in the original map data.

[0104] Because the newly generated lane description curves differ from the parametric cubic curves in the original high-precision map data, it is necessary to recalculate the length l of each lane description curve segment. i :

[0105]

[0106] As mentioned earlier, each parametric cubic curve in the first lane of each road in the high-precision map data contains the distance s from its starting point along the first lane relative to the starting point of the entire road. i For the newly generated lane description curve, combined with equation (15), the starting point of each cubic polynomial is the distance s along the first lane from the starting point of the entire road segment. i for:

[0107]

[0108] For each parametric cubic curve segment in the newly generated lane description curve, the distance s from the starting point of that parametric cubic curve segment along the first lane relative to the starting point of the entire road segment can be obtained through equation (16). i The coordinates of the first and last endpoints of the parametric cubic curve can be obtained through equation (6) and used as the coordinates of the origin of the local UV coordinate system of the parametric cubic curve in the global XY coordinate system (x i_0 ,y i_0 The slope of the tangent line at the first endpoint of the parametric cubic curve in the global XY coordinate system can be obtained through equation (7) as the rotation angle θ of the local UV coordinate system of the parametric cubic curve relative to the global XY coordinate system. i_headi The length l of the parametric cubic curve can be obtained through equation (15). i The cubic polynomial coefficients aU of the cubic curve in the u direction can be obtained through equation (12). i bU i cU i dU i and the cubic polynomial coefficients aV in the v direction i bV i cV i dV i .

[0109] Based on obtaining the lane description curve of the first lane on the first road, alternatively, similar optimizations can be performed on the width description curve describing the width of the first lane.

[0110] Figure 5 The process of generating a width description curve for a first lane according to an embodiment of the present invention is illustrated. Figure 5 As shown, in step S501, the starting distance and parameters of each segment of the cubic polynomial are read from the initial width description curve of the first lane in the high-precision map data. Specifically, information for each segment of the cubic polynomial used for lane width can be read from the high-precision map data. For the i-th segment of the cubic polynomial, the information may include: the distance sWidth of the starting point of the cubic polynomial relative to the starting point of the entire first lane along the lane description curve of the first lane. i The coefficients a, b, c, and d of the cubic polynomial in this segment.

[0111] In step S502, parameter sampling points can be extracted along the first lane according to a specific pattern, and the width of the parameter sampling points on that lane can be obtained. Optionally, sampling can be performed along the first lane to generate a series of initial sampling points on the initial width description curve of the first lane. To ensure the continuity of connected roads, the first and last endpoints of the initial width description curve can optionally be always included in the above sampling points.

[0112] Based on the initial sampling points extracted above, two sets of sampling points can be created: the endpoint set of the width fitting curve for the first lane and the error verification point set of the width of the first lane. The initial endpoint set contains at least two sampling points: the first endpoint and the last endpoint. The error verification point set consists of all initial sampling points other than the first and last endpoints. That is:

[0113] In step S503, a set of endpoints for the width-fitted curve is created, which initially includes the first endpoint and the last endpoint.

[0114] In step S504, an error verification point set is created, initially consisting of all initial sampling points except for the endpoint set of the initial width fitting curve.

[0115] Subsequently, in step S505, a cubic polynomial is created piecewise using the width information of the initial sampling points in the endpoint set.

[0116] Specifically, the lane width at each initial sampling point can be calculated using a cubic polynomial describing the lane width, thereby obtaining the (s,Width) of each initial sampling point. actua l(s)):

[0117] Width actual (s)=a+b*Δs+c*Δs 2 +d*Δs 3 (17)

[0118] Where Δs = s - sWidth i , which is the displacement increment of the sampling point of this parameter relative to the starting point of the cubic polynomial of the fitted curve for this width.

[0119] To ensure the continuity of lane width changes described by the subsequent cubic polynomials, the lane width change rate at each sampling point can be calculated simultaneously:

[0120] Width'(s)=b+2*c*Δs+3*d*Δs 2 (18)

[0121] The endpoints of the fitted width curve can contain only two initial sampling points: the first endpoint and the last endpoint. Therefore, a cubic polynomial can be constructed using the first endpoint and the last endpoint and the cubic Schmidt spline interpolation method.

[0122] The width-fitted curve represented by a cubic polynomial can have the following form:

[0123] Width fitted (s)=a+b*Δs+c*Δs 2 +d*Δs 3 (19)

[0124] In step S506, the error between each initial sampling point in the error verification point set and the sampling point mapped on the width fitting curve is calculated.

[0125] At this point, the distance Δs of the initial sampling point in the error verification point set relative to the first endpoint of the lane description curve along the first lane can be used as input to calculate the output of the newly generated width fitting curve, and the error between this output and the original road width Width(Δs) of the sampling point can be calculated:

[0126] error = |Width fitted (Δs)-Width dctual (s)| (20)

[0127] In step S507, it is determined whether there are any initial sampling points with errors greater than the width threshold. If the errors of all initial sampling points in the error verification point set are not greater than the set width threshold, then proceed to step S509, where the fitted width curve is used as the width description curve to represent the original cubic curve of multiple parameters.

[0128] If, in step S507, there are sampling points in the error verification point set whose errors are greater than the set width threshold, then proceed to step S508, remove the initial sampling point with the largest error value from the error verification point set, add the initial sampling point to the endpoint set of the width fitting curve of the first lane, and return to step S505 to re-perform curve fitting.

[0129] In step S508, selecting the initial sampling point with the largest error as the endpoint for continued fitting is merely an example. In practical applications, multiple initial sampling points can be selected as endpoints for curve fitting based on the relationship between the mapped sampling points and the width threshold; this is not a limitation here.

[0130] Finally, until step S507 determines that the errors of all initial sampling points are not greater than the set lane threshold, then proceed to step S509 and end.

[0131] For each segment of the cubic polynomial in the newly generated width description curve, the distance *s* of the starting point of that segment of the cubic polynomial relative to the starting point of the first lane along the first lane can be used as the distance *s* of the starting point of that segment of the cubic polynomial relative to the entire starting point along the first lane. i The coefficients of the cubic polynomial in this segment can be obtained through equation (19).

[0132] When the first road includes multiple lanes, a road description curve for the first road can be generated based on the width description curves corresponding to one or more lanes adjacent to the first lane on the first road.

[0133] According to one example of an embodiment of the present invention, in response to the fact that the representation data of a series of initial lane description curves of the first lane contains a lane offset data component relative to a reference lane line in the map data, the initial lane description curve of the first lane is corrected to the reference lane line in the map data to remove the lane offset data component. Furthermore, the description curves of the first shoulder and / or the second shoulder on both sides of the first road can also be removed from the road description curve of the first road.

[0134] Figure 6 This diagram illustrates a road description curve processing method for a first road according to an embodiment of the present invention. Without altering the road geometry, if the first lane is not defined as a road reference line in the original high-precision map data, the road reference line can be offset to the first lane, thereby removing the cubic polynomial in the high-precision map data used to describe the lane offset data component of the first lane relative to the road reference line.

[0135] like Figure 6 As shown, in step S601, the road reference line, the first lane offset, and the width data of each lane can be read from the original high-precision map data. Initial sampling points can be extracted from the road reference line and their coordinate information can be obtained.

[0136] Subsequently, in step S602, it can be selected whether to use the first lane as a road reference line as needed. If the first lane is to be used as a road reference line, i.e., the first lane coincides with the road reference line, step S603 can be entered to extract the first lane offset at each initial sampling point of the first lane and remove the cubic polynomial data component used to describe the first lane offset. If the first lane is not to be used as a road reference line, then there is a cubic polynomial of the lane offset of the first lane relative to the road reference line, and step S608 can be entered to regenerate one or more parametric cubic polynomial curves using the sampling points of the first lane for the lane description curve of the first lane.

[0137] Specifically, in an example of an embodiment of the present invention, sampling can be performed along the road reference line to generate a series of road reference line sampling points; the lane offset of the sampling point is calculated based on the cubic polynomial of the lane offset of the first lane corresponding to each sampling point.

[0138] Offset(s) = a + b * Δs + c * Δs 2 +d*Δs 3 (twenty one)

[0139] Therefore, equations (4) and (6) can be used to calculate the coordinates (x, y) of each sampling point on the road reference line in the global XY coordinate system, equation (21) can be used to calculate the offset of the first lane relative to the road reference line at the sampling point, and equation (7) can be used to calculate the tangent slope θ at the curve sampling point in the global XY coordinate system. headi .

[0140] The coordinates of the sampling point on the first lane corresponding to each road reference line sampling point in the global XY coordinate system can be represented as:

[0141]

[0142] Where θ normal The orientation angle of the left normal of this sampling point in the global XY coordinate system can be calculated using the following formula:

[0143]

[0144] Therefore, a new parametric cubic curve can be constructed using the coordinates and derivatives of points on the first lane through cubic Schmidt spline interpolation. This constructed parametric cubic curve located on the first lane can then be used as a new road reference line. Since the first lane coincides with the road reference line, the cubic polynomial data describing the lane offset of the first lane relative to the road reference line in the high-precision map data can be removed.

[0145] Furthermore, the original high-precision map data may describe shoulder lanes by defining lane widths. If the definition of shoulder lanes has no impact on the use of high-precision map data (such as autonomous driving), then the shoulder lanes on both sides of the road can be removed from the high-precision map data. Therefore, in the example of this embodiment of the invention, the left and right shoulder lanes can also be removed from the generated road description curves respectively. In step S604, it can be determined whether to remove the left shoulder lane of the first road.

[0146] For the left shoulder lane of the road, when it is not necessary to offset the road reference line to the leftmost first lane and the left shoulder needs to be removed, step S605 can be entered to extract the width of the left shoulder lane at each initial sampling point, and then the width description data of the left shoulder lane can be removed. For example, the lane offset cubic polynomial can be refitted using a similar processing method as the lane width cubic polynomial, where the offset of each initial sampling point can be the difference between the original offset of the first lane relative to the road reference line and the width of the left shoulder lane.

[0147] Furthermore, when the first lane is the left edge of the left shoulder lane of the first road, removing it will require moving the first lane from the left edge of the shoulder lane to the right edge of the shoulder lane. To address this, step S606 can be performed, using the first lane offset at each initial sampling point and the width of the removed left shoulder lane of the first road to calculate the coordinates and other information of the corresponding points on the first lane. In other words, before offsetting the road reference line to the first lane, the lane width corresponding to the left shoulder lane should be calculated using equation (17), and the offset of the first lane relative to the road reference line should be calculated using equation (21). The Offset in equation (22) can be modified to be the difference between the offset of the first lane relative to the road reference line and the width of the shoulder lane.

[0148] After completing the above operations, proceed to step S607, where the initial sampling point corresponding to the first lane is used as the initial sampling point of the first lane that coincides with the road reference line.

[0149] Then, proceed to step S608, where one or more parametric cubic curves are regenerated using the initial sampling points of the first lane to serve as the lane description curve for the first lane. Then proceed to step S609, where one or more cubic polynomials are regenerated using the initial width sampling points of the first lane to serve as the width description curve for the first lane.

[0150] Following this, it can be determined whether the right shoulder lane of the first road needs to be removed in step S610. If removal is required, proceed to step S611 to directly remove the width description data of the right shoulder lane of the first road. If removal is not required, proceed to step S612 to end the process.

[0151] Figure 7 This diagram illustrates an example of processing a first lane and left and right shoulders on a first road according to an embodiment of the present invention. Figure 7 As shown, the road reference line of the first road in the initial map data can be represented by curves 105, 106, and 107 connected by points 101, 102, 103, and 104. The left shoulder lane 131 can have the left edge 115 and the right edge 116, while the right shoulder lane can have the right shoulder lane 132. In addition, the first road may also include lane description curves for other lanes (such as curve 117 or curves 112, 113, and 114 connected by points 108, 109, 110, and 111) and spaced-out lanes such as 118.

[0152] After shifting to the first lane and removing the shoulder, as Figure 7As shown, the first lane can be offset to coincide with the road reference line, represented by 167, while the remaining lane description curves can include curve 167 with endpoints 161 and 162, curve 166 with endpoints 163 and 164, etc.

[0153] In one example of this invention, when each pair of adjacent intersections in high-precision map data contains one or more sequentially connected roads, it can be determined whether to combine multiple roads into one road based on conditions, thereby simplifying the representation of lane description curves and road description curves. Furthermore, when a road contains changing lane-related properties, it can also be considered to split the road to more accurately represent lane description curves and road description curves.

[0154] Figure 8 A schematic flowchart illustrating road processing according to an embodiment of the present invention is shown. First, in step S801, multiple sequentially connected roads can be found from high-precision map data, wherein the multiple roads do not include roads at intersections.

[0155] Subsequently, in step S802, sequentially connected roads can be combined into a single road to obtain a combined road. In step S803, data from all sequentially connected roads can be read and a lane description curve for the first lane of the combined road can be generated. At this point, road reference line sampling points can be extracted for each road between two adjacent intersections; the initial sampling points of the first lane of each road are added to the initial sampling point set of the first lane of the combined road in sequence according to the road connection order; then, one or more parametric cubic curves of the combined road can be generated to describe the lane description curve of the first lane of the combined road.

[0156] In step S804, the lane characteristics of each road in the sequentially connected roads can be read, including the number of lanes, lane connection relationships, and the properties of the lane description curves. In step S805, the roads in the sequentially connected roads can be combined or divided according to the lane characteristics, forming one or more processed roads in the combined roads.

[0157] Subsequently, width description curves can be generated accordingly for the processed roads. In step S806, the initial sampling points for the width of each lane are obtained based on the distance range of the lane description curve of the processed road along the first lane in the combined road. In step S807, width description curves corresponding to each lane of each processed road in the combined road are generated.

[0158] Subsequently, the road description curves at intersections connecting multiple roads can be processed separately. That is, in step S808, the roads at intersections connecting the sequentially connected roads can be compressed individually, which may include processing the road reference lines, lane widths, lane offsets, shoulder lanes, etc. of the road separately. The processing ends in step S809.

[0159] Figure 9 A schematic diagram illustrating an example of road processing according to an embodiment of the present invention is shown. Specifically, in such... Figure 9 In the example shown, there are six roads connected sequentially between the two intersections: S719, S718, S717, S716, S715, and S714. The lane description curves for the first lane of each road are RL707, RL708, RL709, RL711, and RL712, respectively. Road S719 contains five lanes: L720, L721, L722, L723, and L724. The right lane markings of lane L723 end somewhere in the middle of road S719. S718, S717, S716, and S715 each contain four lanes: L720, L721, L722, and L723. Road S714 contains five lanes: L720, L721, L722, L723, and L713. The right lane marking for lane L723 begins somewhere in the middle of road S714. The right lane markings for lanes L720 and L721 in road S716 are dashed lines, while the right lane markings for lanes L720 and L721 in road S715 are solid lines.

[0160] According to the method in the example of the embodiment of the present invention, the lane description curves of the first lanes corresponding to the six sequentially connected roads S719, S718, S717, S716, S715, and S714 can be combined into RL726. Subsequently, the sequentially connected roads S719, S718, S717, S716, S715, and S714 can be combined or split into roads S732, S731, S730, S729, S728, and S727. Among them, since the right lane marking of lane L723 in S719 ends somewhere in the middle of S719, S719 is divided into S732 and S731; similarly, S714 can be divided into S728 and S727.

[0161] Furthermore, since S718, S717, and S716 have the same number of lanes, lane description curve properties, and lane connection relationships, they are combined into S730. Because the right-side lane markings of lanes L720 and L721 in S715 differ from those in S716, S715 is not combined into S730. Because S719 and S718, and S715 and S714 have different numbers of lanes, they are not combined.

[0162] Below, refer to Figure 10 The lane indication device according to an embodiment of the present invention will be described. Figure 10 A block diagram of a lane indication device 1000 according to an embodiment of the present invention is shown. Figure 10 As shown, the lane indicating device 1000 includes an acquisition unit 1010, an extraction unit 1020, a fitting unit 1030, and a description unit 1040. Besides these units, the lane indicating device 1000 may also include other components; however, since these components are not relevant to the content of this embodiment, their illustrations and descriptions are omitted here. Furthermore, the specific details of the following operations performed by the lane indicating device 1000 according to this embodiment are consistent with those described above. Figure 1 The details described are the same, so repeated descriptions of the same details are omitted here to avoid repetition.

[0163] Figure 10 The acquisition unit 1010 of the lane representation device 1000 in the map acquires multiple consecutive initial lane description curves describing the first lane on the first road from the map data.

[0164] In this embodiment of the invention, the initial lane description curves of the first lane of the first road in the high-precision map data can be obtained first. For example, each segment of the initial lane description curve can be represented by a parametric cubic curve.

[0165] The extraction unit 1020 extracts multiple initial sampling points on the multiple initial lane description curves, and the multiple initial sampling points include at least the first endpoint of the first initial lane description curve and the last endpoint of the last initial lane description curve.

[0166] In this embodiment of the invention, extracting multiple initial sampling points on the multiple initial lane description curves may include: extracting the multiple initial sampling points on the multiple initial lane description curves at equal intervals based on curve interpolation parameters; or, extracting the multiple initial sampling points on the multiple initial lane description curves at equal intervals based on distance.

[0167] In this embodiment of the invention, since the multiple initial lane description curves are multiple lane initial description curves connected end to end, the multiple initial sampling points may include at least the first endpoint and the last endpoint of these curves. That is, the multiple initial sampling points may include at least the first endpoint of the first initial lane description curve and the last endpoint of the last initial lane description curve.

[0168] The fitting unit 1030 performs curve fitting using at least the first endpoint and the last endpoint to generate a lane fitting curve.

[0169] In an embodiment of the present invention, curve fitting using at least the first endpoint and the last endpoint to generate a lane fitting curve may include: using at least the first endpoint and the last endpoint as the endpoints of the lane fitting curve, converting the global coordinate system to a local coordinate system with the first endpoint as the origin, and performing curve fitting by calculating the coordinates of the endpoints of the lane fitting curve in the local coordinate system of the lane fitting curve and the derivatives of the coordinates.

[0170] The description unit 1040 determines whether to update the generated lane fitting curve based on the error between each initial sampling point and the sampling points mapped to the generated lane fitting curve, and obtains the lane description curve of the first lane based on the determination result.

[0171] In one example of this invention, iteratively determining whether initial sampling points should be used as endpoints of the lane fitting curve for lane fitting can be achieved. Optionally, the iterative steps may include: Step S1: For the generated lane fitting curve, obtaining the lane error between each initial sampling point and the corresponding mapped sampling point; Step S2: In response to the lane error between each initial sampling point and the corresponding mapped sampling point being less than a lane threshold, determining not to update the generated lane fitting curve, and using the generated lane fitting curve as the lane description curve of the first lane; Step S3: In response to the existence of a lane error value greater than the lane threshold between each initial sampling point and the corresponding mapped sampling point, supplementing the initial sampling point corresponding to the largest lane error value as an endpoint of the lane fitting curve; Step S4: Performing curve fitting with the endpoint as a supplemented endpoint and the first endpoint of its respective lane fitting curve as the endpoint of the lane fitting curve, and performing curve fitting with the endpoint as a supplemented first endpoint and the end point of its respective lane fitting curve as the endpoint of the lane fitting curve, and performing steps S1-S4 on the generated lane fitting curve.

[0172] In another example of this embodiment of the invention, optionally, the width description curve of the first lane can be obtained by simplifying the process using an initial width description curve in the map data used to describe the width of the first lane. Specifically, this may include: obtaining multiple consecutive initial width description curves in the map data describing the width of the first lane; extracting multiple initial sampling points on the multiple initial width description curves, wherein the multiple initial sampling points include at least the first endpoint of the first initial width description curve and the last endpoint of the last initial width description curve; performing curve fitting using at least the first endpoint and the last endpoint to generate a width fitting curve; determining whether to update the generated width fitting curve based on the error between each initial sampling point and the sampling point mapped to the generated width fitting curve, and obtaining the width description curve of the first lane based on the determination result.

[0173] Similar to the lane description curve described above, the fitting process for the width description curve can also be iteratively determined whether the initial sampling points need to be used as the endpoints of the width fitting curve. Optionally, the steps may include: Step S5: For the generated width fitting curve, obtain the width error between each initial sampling point and the corresponding mapped sampling point; Step S6: In response to the width error between each initial sampling point and the corresponding mapped sampling point being less than the width threshold, determine not to update the generated width fitting curve, and use the generated width fitting curve as the width description curve of the first lane; Step S7: In response to the width error between each initial sampling point and the corresponding mapped sampling point having a width error value greater than the width threshold, supplement the initial sampling point corresponding to the largest width error value as the endpoint of the width fitting curve; Step S4: Perform curve fitting with the endpoint as the supplemented end point and the first endpoint of its respective width fitting curve as the endpoint of the width fitting curve, and perform curve fitting with the endpoint as the supplemented first endpoint and the end point of its respective width fitting curve as the endpoint of the width fitting curve, and perform steps S5-S8 on the generated width fitting curve.

[0174] In another embodiment of the present invention, optionally, in response to the fact that the representation data of the first lane's continuous multiple initial lane description curves includes lane offset data components relative to the reference lane line in the map data, the initial lane description curve of the first lane is corrected to the reference lane line in the map data to remove the lane offset data components. That is, if the high-precision map data does not use the aforementioned center lane (such as the first lane in this embodiment of the present invention) as a road reference line, the lane offset data components of the first lane relative to the road reference line, described by multiple cubic polynomials, can be read from the original high-precision map data. Combined with this offset data component, the coordinate information of each sampling point on the first lane can be calculated using the coordinate information of the sampling points on the road reference line. Thus, parametric cubic curve fitting is performed using the coordinate information of each sampling point on the first lane to realize the use of the first lane as a road reference line. At this time, when the first lane is used as a road reference line, the first lane coincides with the road reference line, thereby removing all cubic polynomial data in the high-precision map data used to describe the lane offset data components of the first lane relative to the road reference line.

[0175] Furthermore, in this embodiment of the invention, after obtaining the lane description curve and width description curve of the first lane, optionally, a road description curve of the first lane can be generated based on the lane description curve and width description curve of the first lane on the first road. Additionally, for cases where the first road has multiple lanes, the width description curves of multiple lanes can be obtained separately, so that a road description curve of the first road can be generated based on the width description curves corresponding to one or more lanes adjacent to the first lane on the first road.

[0176] In another embodiment of the present invention, considering the map scenario for autonomous driving, the left and right shoulders that are inaccessible on the first road may not need to be included in the scenario, thus the road description curve of the first road can be further simplified. Therefore, the method may further include: removing the description curves of the first shoulder and / or the second shoulder on both sides of the first road from the road description curve of the first road. In one example, when the first lane in the first road is the left shoulder lane of the first road, and the lane description curve of the first lane is the left boundary of the left shoulder lane, when deleting the left shoulder lane, it is necessary to shift the lane description curve of the first lane from the left boundary of the left shoulder lane to the right boundary of the left shoulder lane using the width information of the left shoulder lane. In this example, the right shoulder lane can be directly deleted without affecting other limitations in the first road.

[0177] In another example, when the first lane in the first road is the right shoulder lane and the lane description curve of the first lane is the right boundary of the right shoulder lane, deleting the right shoulder lane requires shifting the lane description curve of the first lane from the right boundary of the right shoulder lane to the left boundary of the right shoulder lane using the width information of the right shoulder lane. In this example, the left shoulder lane can be deleted directly without affecting other constraints in the first road.

[0178] In another example, when the lane description curve of the first lane in the first road is not the left boundary of the left shoulder lane or the right boundary of the right shoulder lane, both the left shoulder lane and the right shoulder lane can be directly deleted.

[0179] In this embodiment of the invention, when high-precision map data contains one or more roads connected sequentially between two adjacent intersections, these roads can be combined into one road to achieve road combination and compression. This reduces the number of roads between two adjacent intersections and the number of parametric cubic curves in the lane description curve used to describe the first lane. Furthermore, when features such as the number of lanes, lane connectivity, or lane description curve properties within a road change, the road can be segmented into multiple roads at the point of feature change for processing, ensuring the accuracy of the road description curve.

[0180] According to the lane representation device of the present invention, a lane fitting curve can be generated by using multiple initial sampling points on multiple initial lane description curves. Therefore, high-precision map data can be processed offline by computer software or programs, effectively reducing the amount of high-precision map data and improving its efficiency while ensuring a set accuracy threshold. In applications such as autonomous driving where high-precision map real-time processing is required, the computational resources required for data processing can be effectively reduced.

[0181] In software-in-the-loop testing, model-in-the-loop testing, or hardware-in-the-loop testing scenarios that use high-precision map data to generate virtual simulation scenarios, the high-precision map data generated by the above processing can significantly reduce the configuration time of the simulation environment and the computing resources of real-time or non-real-time processors occupied by high-precision map generation and rendering, thereby improving the efficiency of simulation testing while increasing the range of roads available for simulation testing.

[0182] Below, refer to Figure 11 The lane indication device according to an embodiment of the present invention will be described. Figure 11 A block diagram of a lane indication device 1100 according to an embodiment of the present invention is shown. Figure 11 As shown, the device 1100 can be a computer or a server.

[0183] like Figure 11 As shown, the lane indicating device 1100 includes one or more processors 1110 and a memory 1120. In addition, the lane indicating device 1100 may also include input devices, output devices (not shown), etc., and these components can be interconnected via a bus system and / or other forms of connection mechanisms. It should be noted that... Figure 11 The components and structure of the lane indication device 1100 shown are merely exemplary and not limiting; the lane indication device 1100 may also have other components and structures as needed.

[0184] The processor 1110 may be a central processing unit (CPU) or other form of processing unit with data processing capabilities and / or instruction execution capabilities, and may utilize computer program instructions stored in memory 1120 to perform desired functions, including: acquiring multiple consecutive initial lane description curves describing a first lane on a first road from map data; extracting multiple initial sampling points on the multiple initial lane description curves, the multiple initial sampling points including at least the first endpoint of the first initial lane description curve and the last endpoint of the last initial lane description curve; performing curve fitting using at least the first endpoint and the last endpoint to generate a lane fitting curve; determining whether to update the generated lane fitting curve based on the error between each initial sampling point and its respective sampling point mapped to the generated lane fitting curve; and obtaining the lane description curve of the first lane based on the determination result.

[0185] The memory 1120 may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. One or more computer program instructions may be stored on the computer-readable storage medium, which the processor 1110 may execute to implement the functions of the lane indication device according to the embodiments of the present invention described above, and / or other desired functions, and / or to execute the lane indication method according to the embodiments of the present invention. Various application programs and various data may also be stored in the computer-readable storage medium.

[0186] The following describes a computer-readable storage medium according to embodiments of the present invention, having stored thereon computer program instructions, wherein the computer program instructions, when executed by a processor, perform the following steps: acquiring a series of initial lane description curves describing a first lane on a first road from map data; extracting a plurality of initial sampling points on the plurality of initial lane description curves, the plurality of initial sampling points including at least the first endpoint of the first initial lane description curve and the last endpoint of the last initial lane description curve; performing curve fitting using at least the first endpoint and the last endpoint to generate a lane fitting curve; determining whether to update the generated lane fitting curve based on the error between each initial sampling point and its respective sampling point mapped onto the generated lane fitting curve; and acquiring the lane description curve of the first lane according to the determination result.

[0187] The following describes a computer program product according to an embodiment of the present invention, including computer program instructions, wherein when the computer program instructions are executed by a processor, they implement the steps of the lane representation method as described above.

[0188] Of course, the specific embodiments described above are merely examples and not limitations. Those skilled in the art can combine and integrate some steps and devices from the various embodiments described separately above to achieve the effects of the present invention. Such combined and integrated embodiments are also included in the present invention, but will not be described one by one here.

[0189] Note that the advantages, benefits, and effects mentioned in this invention are merely examples and not limitations, and should not be considered as essential features of every embodiment of the invention. Furthermore, the specific details described above are for illustrative and illustrative purposes only, and are not intended to limit the invention. These details do not limit the invention from being implemented solely by employing these specific details.

[0190] The block diagrams of devices, apparatuses, devices, and systems involved in this invention are merely illustrative examples and are not intended to require or imply that they must be connected, arranged, or configured in the manner shown in the block diagrams. As those skilled in the art will recognize, these devices, apparatuses, devices, and systems can be connected, arranged, and configured in any manner. Words such as “comprising,” “including,” “having,” etc., are open-ended terms meaning “including but not limited to,” and are used interchangeably with them. The terms “or” and “and” as used herein refer to the terms “and / or,” and are used interchangeably with them unless the context clearly indicates otherwise. The term “such as” as used herein refers to the phrase “such as but not limited to,” and is used interchangeably with it.

[0191] The flowcharts and method descriptions in this invention are merely illustrative examples and are not intended to require or imply that the steps of the various embodiments must be performed in the given order. As those skilled in the art will recognize, the steps in the above embodiments can be performed in any order. Words such as "then," "next," etc., are not intended to limit the order of steps; these words are only used to guide the reader through the description of these methods. Furthermore, any reference to a singular element, such as the use of the articles "a," "one," or "the," is not to be construed as limiting that element to the singular.

[0192] Furthermore, the steps and apparatus in the various embodiments herein are not limited to any one embodiment. In fact, new embodiments can be conceived by combining relevant steps and apparatus in the various embodiments herein with the concepts of the present invention, and these new embodiments are also included within the scope of the present invention.

[0193] Each operation described above can be performed by any suitable means capable of performing the corresponding function. Such means may include various hardware and / or software components and / or modules, including but not limited to circuits, application-specific integrated circuits (ASICs), or processors.

[0194] The various exemplified logic blocks, modules, and circuits described herein can be implemented or performed using a general-purpose processor, digital signal processor (DSP), ASIC, field-programmable gate array (FPGA) or other programmable logic device (PLD), discrete gate or transistor logic, discrete hardware components, or any combination thereof designed to perform the functions described herein. The general-purpose processor may be a microprocessor, but alternatively, it may be any commercially available processor, controller, microcontroller, or state machine. The processor may also be implemented as a combination of computing devices, such as a combination of a DSP and a microprocessor, multiple microprocessors, one or more microprocessors cooperating with a DSP core, or any other such configuration.

[0195] The steps of the methods or algorithms described in this invention can be directly embedded in hardware, in a software module executed by a processor, or a combination of both. The software module can reside in any form of tangible storage medium. Some examples of usable storage media include random access memory (RAM), read-only memory (ROM), flash memory, EPROM, EEPROM, registers, hard disks, removable disks, CD-ROMs, etc. The storage medium can be coupled to the processor so that the processor can read information from and write information to the storage medium. Alternatively, the storage medium can be integral with the processor. The software module can be a single instruction or many instructions, and can be distributed across several different code segments, different programs, and across multiple storage media.

[0196] The method of this invention includes one or more actions for implementing the method. The methods and / or actions may be interchanged without departing from the scope of the claims. In other words, unless a specific order of actions is specified, the order and / or use of specific actions may be modified without departing from the scope of the claims.

[0197] The described functionality can be implemented in hardware, software, firmware, or any combination thereof. If implemented in software, the functionality can be stored as one or more instructions on a tangible computer-readable medium. The storage medium can be any available tangible medium that can be accessed by a computer. By way of example, and not limitation, such a computer-readable medium can include RAM, ROM, EEPROM, CD-ROM or other optical disc storage, disk storage or other magnetic storage devices, or any other tangible medium that can be used to carry or store desired program code in the form of instructions or data structures and that can be accessed by a computer. As used herein, a disc includes a compact disc (CD), a laser disc, an optical disc, a digital universal disc (DVD), a floppy disk, and a Blu-ray disc.

[0198] Therefore, a computer program product can perform the operations described herein. For example, such a computer program product can be a computer-readable tangible medium having instructions tangibly stored (and / or encoded) thereon, which can be executed by one or more processors to perform the operations described herein. The computer program product may include packaging materials.

[0199] Software or instructions can also be transmitted via a transmission medium. For example, software can be transmitted from a website, server, or other remote source using transmission media such as coaxial cable, fiber optic cable, twisted pair, digital subscriber line (DSL), or wireless technologies such as infrared, radio, or microwave.

[0200] Furthermore, modules and / or other suitable means for carrying out the methods and techniques described herein can be downloaded and / or obtained by user terminals and / or base stations as appropriate. For example, such a device can be coupled to a server to facilitate the transmission of means for carrying out the methods described herein. Alternatively, the various methods described herein can be provided via storage components (e.g., RAM, ROM, physical storage media such as CDs or floppy disks) so that user terminals and / or base stations can obtain the various methods when coupled to the device or when providing storage components to the device. Furthermore, any other suitable techniques for providing the methods and techniques described herein to the device can be utilized.

[0201] Other examples and implementations are within the scope and spirit of this invention and the appended claims. For example, due to the nature of software, the functions described above can be implemented using software executed by a processor, hardware, firmware, hardwired, or any combination thereof. Features implementing the functions can also be physically located in various places, including being distributed so that parts of the functions are implemented at different physical locations. Moreover, as used herein, including as used in the claims, the "or" used in a list of items beginning with "at least one" indicates a separate list, such that a list of, for example, "at least one of A, B, or C" means A or B or C, or AB or AC or BC, or ABC (i.e., A and B and C). Furthermore, the word "exemplary" does not mean that the described examples are preferred or better than other examples.

[0202] Various changes, substitutions, and modifications can be made to the technology described herein without departing from the teachings defined by the appended claims. Furthermore, the scope of the claims is not limited to the specific aspects of the processes, machines, manufacturing processes, events, means, methods, and actions described above. Currently existing or later-developed processes, machines, manufacturing processes, events, means, methods, or actions that perform substantially the same function or achieve substantially the same result as the corresponding aspects described herein can be utilized. Therefore, the appended claims include such processes, machines, manufacturing processes, events, means, methods, or actions within their scope.

[0203] The above description of aspects of the invention is provided to enable any person skilled in the art to make or use the invention. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein can be applied to other aspects without departing from the scope of the invention. Therefore, the invention is not intended to be limited to the aspects shown herein, but rather to be carried out within the widest scope consistent with the principles and novel features of the invention herein.

[0204] The above description has been given for illustrative and descriptive purposes. Furthermore, this description is not intended to limit the embodiments of the invention to the forms described herein. Although numerous exemplary aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, alterations, additions, and sub-combinations therein.

Claims

1. A lane representation method, comprising: Obtain multiple consecutive initial lane description curves from the map data that describe the first lane on the first road; Multiple initial sampling points are extracted from the multiple initial lane description curves, and the multiple initial sampling points include at least the first endpoint of the first initial lane description curve and the last endpoint of the last initial lane description curve. At least the first and last endpoints are used to perform curve fitting to generate a lane fitting curve; Based on the error between each initial sampling point and its respective sampling point mapped onto the generated lane fitting curve, it is determined whether to update the generated lane fitting curve, and the lane description curve of the first lane is obtained according to the determination result.

2. The method according to claim 1, wherein, The initial lane description curves for each segment are represented by parametric cubic curves.

3. The method according to claim 1, wherein, Extracting multiple initial sampling points from the multiple initial lane description curves includes: Based on curve interpolation parameters or distance, the multiple initial sampling points are extracted at equal intervals on the multiple initial lane description curves.

4. The method according to claim 1, wherein, Based on the error between each initial sampling point and its corresponding sampling point mapped onto the generated lane fitting curve, it is determined whether to update the generated lane fitting curve. Based on this determination, the lane description curve of the first lane is obtained, including: Step S1: For the generated lane fitting curve, obtain the lane error between each initial sampling point and the corresponding mapped sampling point; Step S2: In response to the fact that the lane error between each initial sampling point and the corresponding mapped sampling point is less than the lane threshold, it is determined not to update the generated lane fitting curve, and the generated lane fitting curve is used as the lane description curve of the first lane. Step S3: In response to the existence of a lane error value greater than the lane threshold between each initial sampling point and the corresponding mapped sampling point, the initial sampling point corresponding to the largest lane error value is added as the endpoint of the lane fitting curve. Step S4: Perform curve fitting by taking the endpoint as a supplementary end point and the first endpoint of the lane fitting curve to which it belongs as the endpoint of the lane fitting curve, and perform curve fitting by taking the endpoint as a supplementary first endpoint and the end point of the lane fitting curve to which it belongs as the endpoint of the lane fitting curve, and perform steps S1-S4 on the generated lane fitting curve.

5. The method according to any one of claims 1-4, wherein, At least using the first and last endpoints to perform curve fitting to generate a lane fitting curve includes: At least the first and last endpoints are used as the endpoints of the lane fitting curve. Curve fitting is performed by calculating the coordinates of the endpoints of the lane fitting curve in the local coordinate system of the lane fitting curve and the derivative of the coordinates.

6. The method according to claim 1, wherein, The method further includes: Obtain multiple consecutive initial width description curves from the map data that describe the width of the first lane; Multiple initial sampling points are extracted from the multiple initial width description curves, and the multiple initial sampling points include at least the first endpoint of the first initial width description curve and the last endpoint of the last initial width description curve. At least the first and last endpoints are used to perform curve fitting to generate a width-fitted curve; Based on the error between each initial sampling point and its respective sampling point mapped onto the generated width fitting curve, it is determined whether to update the generated width fitting curve, and the width description curve of the first lane is obtained according to the determination result.

7. The method according to claim 1, wherein, The method further includes: In response to the fact that the representation data of the first lane's consecutive initial lane description curves contains a lane offset data component relative to a reference lane line in the map data, the initial lane description curve of the first lane is corrected to the reference lane line in the map data to remove the lane offset data component.

8. The method according to claim 6, wherein, The method further includes: Based on the lane description curve and the width description curve of the first lane on the first road, a road description curve for the first lane is generated; Based on the width description curves corresponding to one or more lanes adjacent to the first lane on the first road, a road description curve for the first road is generated.

9. The method according to claim 8, wherein, The method further includes: In the road description curve of the first road, remove the description curves of the first shoulder and / or the second shoulder on both sides of the first road.

10. A lane representation method, comprising: Obtain the road description curve describing the first road from the map data; In response to differences in the number of lanes, lane connection relationships, or lane description curve properties of the first road, the road description curve of the first road is split; or the road description curve of the second road connected to the first road is obtained from the map data, and in response to the fact that the number of lanes, lane connection relationships, and lane description curve properties of the road description curves of the first road and the second road are the same, the road description curves of the first road and the second road are merged.

11. A lane indication device, comprising: The acquisition unit is configured to acquire multiple consecutive initial lane description curves describing the first lane on the first road from the map data; The extraction unit is configured to extract multiple initial sampling points on the multiple initial lane description curves, wherein the multiple initial sampling points include at least the first endpoint of the first initial lane description curve and the last endpoint of the last initial lane description curve. The fitting unit is configured to perform curve fitting using at least the first endpoint and the last endpoint to generate a lane fitting curve; The description unit is configured to determine whether to update the generated lane fitting curve based on the error between each initial sampling point and the sampling points mapped to the generated lane fitting curve, and to obtain the lane description curve of the first lane based on the determination result.

12. A lane indication device, comprising: processor; and a memory, in which computer program instructions are stored. When the computer program instructions are executed by the processor, the processor performs the following steps: Obtain multiple consecutive initial lane description curves from the map data that describe the first lane on the first road; Multiple initial sampling points are extracted from the multiple initial lane description curves, and the multiple initial sampling points include at least the first endpoint of the first initial lane description curve and the last endpoint of the last initial lane description curve. At least the first and last endpoints are used to perform curve fitting to generate a lane fitting curve; Based on the error between each initial sampling point and its respective sampling point mapped onto the generated lane fitting curve, it is determined whether to update the generated lane fitting curve, and the lane description curve of the first lane is obtained according to the determination result.

13. A computer-readable storage medium having stored thereon computer program instructions, wherein, When the computer program instructions are executed by the processor, the following steps are performed: Obtain multiple consecutive initial lane description curves from the map data that describe the first lane on the first road; Multiple initial sampling points are extracted from the multiple initial lane description curves, and the multiple initial sampling points include at least the first endpoint of the first initial lane description curve and the last endpoint of the last initial lane description curve. At least the first and last endpoints are used to perform curve fitting to generate a lane fitting curve; Based on the error between each initial sampling point and its respective sampling point mapped onto the generated lane fitting curve, it is determined whether to update the generated lane fitting curve, and the lane description curve of the first lane is obtained according to the determination result.

14. A computer program product comprising computer program instructions, wherein, When the computer program instructions are executed by the processor, they implement the steps of the method according to any one of claims 1-9.

Images