Lane centerline smoothing method and apparatus

Abstract

The application provides a lane center line smoothing method and device, first lane center lines obtained by interpolating discrete lane center line data are processed in segments, and the curves of different segments are respectively subjected to B-spline fitting smoothing processing, and then the fitting curves obtained by different segments are spliced to obtain second lane center lines. The application can improve the speed and efficiency of fitting the first lane center line of the lane center line, the obtained second lane center line can also ensure smoothness and improve the accuracy of the fitted curve, and provides a more accurate basis for subsequent vehicle driving along the second lane center line as a driving track.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, and particularly relates to a lane center line smoothing method and device. BACKGROUND

[0002] The vehicle with the automatic driving function can generate a safe, reasonable and comfortable path according to the external environment information and the state of the vehicle itself, and control the vehicle to travel according to the generated path. In order to determine the driving path of the vehicle according to the center line data of the lane, the vehicle performs quadratic interpolation processing on the lane center line data composed of discrete points after obtaining the lane center line data, to obtain a continuous curve form of the center line, and then performs B-spline fitting processing on the center line to obtain a smooth lane center line, which can be used as the trajectory of the subsequent vehicle driving.

[0003] However, the B-spline fitting processing in the prior art has a large amount of calculation and low precision when fitting the center line, so that the speed and efficiency of the vehicle in smoothing the curve of the lane center line are low, and the precision of the smooth curve is low. SUMMARY

[0004] The present application provides a lane center line smoothing method and device to solve the technical problems of slow speed, low efficiency and low precision of the vehicle in smoothing the curve of the lane center line.

[0005] The first aspect of the present application provides a lane center line smoothing method, comprising: performing quadratic interpolation processing on a plurality of discrete data of a lane center line to be smoothed to obtain a first lane center line; dividing the first lane center line into a plurality of category intervals according to the curvature of the first lane center line; performing fitting processing on the plurality of category intervals respectively to obtain a plurality of fitting curves, wherein different fitting parameters are used when fitting each category interval; and performing splicing processing on the plurality of fitting curves to obtain a second lane center line.

[0006] The second aspect of the present application provides a lane center line smoothing device, which can be used to execute the lane center line smoothing method provided in the first aspect of the present application. The device comprises: an interpolation module configured to perform quadratic interpolation processing on a plurality of discrete data of a lane center line to be smoothed to obtain a first lane center line; a classification module configured to divide the first lane center line into a plurality of category intervals according to the curvature of the first lane center line; a fitting module configured to perform fitting processing on the plurality of category intervals respectively to obtain a plurality of fitting curves, wherein different fitting parameters are used when fitting each category interval; and a splicing module configured to perform splicing processing on the plurality of fitting curves to obtain a second lane center line.

[0007] The third aspect of the present application provides an electronic device, comprising a processor and a memory; wherein the memory stores a computer program, and when the processor executes the computer program, the processor can be used to execute the lane center line smoothing method according to any one of the first aspect of the present application.

[0008] The fourth aspect of the present application provides a computer readable storage medium, characterized in that the computer readable storage medium stores a computer program, and the computer program can be used to execute the lane center line smoothing method according to any one of the first aspect of the present application when executed.

[0009] In summary, the lane center line smoothing method and device provided by the present application, by segmenting the first lane center line obtained by interpolating the discrete lane center line data, and respectively performing B-spline fitting smoothing processing on the curves of different segments, and splicing the fitting curves of multiple segments to obtain the second lane center line. Thus, the entire first lane center line does not need to be fitted, and for each type of different interval corresponding to the curve in the first lane center line, the fitting times and the sampling point interval used in B-spline fitting are different. When fitting the curves corresponding to the 90-degree turning and U-turn of the vehicle, the fitting times can be increased and the sampling point interval can be increased, while for the curves corresponding to the straight driving and small turning of the vehicle, the fitting times can be reduced and the sampling point interval can be reduced. When fitting the first lane center line, the calculation amount of processing the entire curve is reduced, and the steps in the process of smoothing the first lane center line by B-spline fitting are reduced, thereby improving the speed and efficiency of fitting the first lane center line of the lane center line. The obtained second lane center line can also ensure smoothness and improve the accuracy of the fitted curve, providing a more accurate basis for subsequent vehicle driving along the second lane center line as a driving trajectory. BRIEF DESCRIPTION OF DRAWINGS

[0010] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed in the embodiments or prior art description will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present application, and those skilled in the art can also obtain other drawings according to these drawings without creative labor.

[0011] Figure 1 It is a schematic diagram of the application application scenario;

[0012] Figure 2 It is a flowchart of a method for a vehicle to generate a driving path according to lane center line data;

[0013] Figure 3 It is a flowchart of an embodiment of the lane center line smoothing method provided by the present application;

[0014] Figure 4 Structure diagram of sliding window provided in the present application;

[0015] Figure 5 Size diagram of sliding window provided in the present application;

[0016] Figure 6 Diagram of one embodiment of the state of sliding window on the first lane center line provided in the present application;

[0017] Figure 7 Diagram of the change of the absolute value of the difference of the heading angle on the first lane center line F1 provided in the present application;

[0018] Figure 8 Diagram of another embodiment of the state of sliding window on the first lane center line provided in the present application;

[0019] Figure 9 Diagram of the change of the absolute value of the difference of the heading angle on the first lane center line provided in the present application;

[0020] Figure 10 Diagram of the absolute value of the difference of the heading angle in all sliding window on the first lane center line provided in the present application;

[0021] Figure 11 Diagram of the adjustment of the control point in the B-spline fitting process provided in the present application;

[0022] Figure 12 Diagram of the change of the fitting curve in the B-spline fitting process provided in the present application;

[0023] Figure 13 Diagram of the change of the fitting curve in the B-spline fitting process provided in the present application;

[0024] Figure 14 Diagram of the second lane center line obtained by B-spline fitting provided in the present application;

[0025] Figure 15 Diagram of the B-spline fitting result after the adjustment of the control point of the curve provided in the present application. DETAILED DESCRIPTION

[0026] The technical solutions in the embodiments of the present application will be described clearly and completely below with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all the other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0027] The terms "first", "second", "third", "fourth" and the like in the description and in the claims of the present application, and above-mentioned drawings, if any, are used to distinguish between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the use of the terms so-termed "first", "second", "third", "fourth" and the like, if any, in the description and in the claims of the present application is merely intended to distinguish between two similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used can be interchanged, where appropriate, so that the embodiments of the present application described herein can be carried out in other than the order shown or described herein. Furthermore, the terms "comprising" and "including" and any of their derivatives, are intended to be construed as encompassing not only the listed items, but also others that are similar or equivalent in function, range of use and the like, to the listed items. The use of the negative "not" is intended to exclude, for example, the listed items, but not to exclude other items that are similar or equivalent in function, range of use and the like, to the listed items.

[0028] Figure 1 For the purpose of illustrating the inventive concepts, the present application is applied to a scenario of vehicle automatic driving as shown in Figure 1 wherein a vehicle with automatic driving function can generate a safe, reasonable and comfortable path according to external environment information and the state of the vehicle itself, and control the vehicle to travel along the generated path.

[0029] In some embodiments, the vehicle can determine the travel path of the vehicle according to the center line data of the lane, for example, in the example as shown in Figure 1 when the vehicle travels from point A to point B in the lane, a path in the center of the lane can be determined according to the center line data of the lane, and the vehicle can travel along the path in the center of the lane.

[0030] In some embodiments, the center line data of the lane can be collected and made by the supplier of the high-precision map, including discrete data points on the center line of the lane, each data point can be used to indicate the position of the lane center line at the position of the data point and the like.

[0031] Figure 2 For a method flowchart of a vehicle generating a travel path according to lane center line data, the method of the vehicle determining a continuous travel path according to discrete lane center line data is shown. First, in S10, the vehicle obtains discrete lane center line data of the lane center line, which can be stored in the vehicle in advance or downloaded by the vehicle through the Internet and the like, and the lane center line data can be provided by the map data department, and then in S20, the vehicle performs secondary interpolation processing on the discrete lane center line data to obtain a continuous curve form of the first lane center line F1. However, the curvature of the first lane center line F1 obtained by secondary interpolation is not continuous, and in some cases, the curvature of the lane center line can change abruptly. For example, Figure 1The first lane center line F1 in the form of a curve obtained through twice interpolation occurs a bend at the dotted line box in the case of a jump in discrete lane center line data, so that the vehicle taking the first lane center line F1 as a driving path will occur a sudden turn and other phenomena when driving to this position, therefore, in order to ensure the smoothness and curvature continuity of the driving curve, after S20, the vehicle needs to process the lane center line F1 through curve fitting to obtain a curvature continuous second lane center line F2 through S30, so as to eliminate the bent part of the first lane center line F1, and finally take the second lane center line F2 as the driving path of the vehicle, and the vehicle can be controlled to drive according to the second lane center line F2 subsequently.

[0032] In some embodiments, B-spline fitting is a common way to smooth the driving trajectory curve due to its characteristics of local adjustment of lines, but in the process of B-spline fitting, the more times of fitting, the more accurate the obtained curve, for example, a common way of five times B-spline fitting is used to smooth the curve, and based on the mathematical characteristics of B-spline fitting, different control points need to be selected in the whole process of fitting the curve A-B as shown, and multiple iterative fitting calculations are performed according to each control point, even in some straight line parts, multiple calculations are performed, the steps are more cumbersome, the calculation amount of curve fitting is increased, and the accuracy of the fitted curve is low, so that the speed and efficiency of the curve smoothing process of the lane center line are low, and the accuracy of the smoothed curve is low. Figure 1

[0033] Therefore, the present application also provides a lane center line smoothing method and device, which segments the first lane center line obtained by interpolating the discrete data of the lane center line, and performs B-spline fitting smoothing processing on the curves of different segments of the first lane center line, to finally obtain a second lane center line after smoothing processing, so as to reduce the calculation amount of processing the whole curve, reduce the steps in the process of lane center line smoothing, and improve the speed and efficiency.

[0034] The above technical solutions of the present application will be described in detail below with specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes can not be described in some embodiments.

[0035] Figure 3 The flowchart of an embodiment of the lane center line smoothing method provided by the present application is shown in FIG. 1. Figure 3 The method shown in FIG. 1 can be executed by a computer, a server, a workstation or any electronic device with related data processing capability, and the method comprises the following steps.

[0036] ​S101: The electronic device performs quadratic interpolation on the discrete data of the lane center line to be smoothed to obtain a first curve.

[0037] Since the data of the lane center line is discrete and the driving path planned according to the lane center line data is continuous, the discrete lane center line data needs to be processed by quadratic interpolation, for example, in the plane shown in Figure 1 , the plurality of discrete data of the lane center line in the A-B interval is processed by quadratic interpolation to obtain a first lane center line F1 in the form of a continuous curve. The specific implementation of the quadratic interpolation process can refer to the prior art, which is not limited here. It should be noted that, as an example, Figure 1 , the driving interval between the A-B points in front of the vehicle is a straight line, it can be understood that in the actual situation, the vehicle can include various driving states such as straight driving, turning, U-turn, etc. between the A-B points, and the first lane center line F1 obtained in S101 can present various extension states such as straight line, curve or arc.

[0038] S102: The first lane center line is processed in sections according to the curvature of the first lane center line calculated in S101.

[0039] In particular, after obtaining the first lane center line F1 in the embodiments of the present application, when smoothing the intervals with discontinuous curvature in the first lane center line F1, the B-spline fitting process is not directly performed, but the first lane center line F1 is first processed in sections according to the curvature variation of the first lane center line F1, and then each section is smoothed.

[0040] In one specific implementation of S102, in order to segment the first lane center line F1, a sliding window of a certain size is first established on the first lane center line F1 according to the driving state information of the vehicle. Here, the sliding window is a virtual concept, and each sliding window is used to reflect the curvature variation of the first lane center line F1 within the range of the window.

[0041] For example, Figure 4 The structure of the sliding window provided in the present application is shown in the figure, wherein, on the first lane center line F1, the sliding window slides forward by a certain distance d between the A-B points, for example, a sliding window D1 is established at d1 point between A-B points, then d2 point is taken at a distance d in front of d1 point, and sliding window D2 is established, and so on.

[0042] In some embodiments, the size of the sliding window can be set according to the turning radius of the first lane center line F1 and the minimum turning radius of the vehicle. For example, Figure 5The schematic diagram of the sliding window provided in this application shows that the size of the sliding window can be determined by combining the turning radius of the road where the vehicle is located and the minimum turning radius of the vehicle. In this embodiment of the application, the length of the vehicle is denoted as m, so the size of the sliding window can be set as a rectangle of 2m*2m.

[0043] Among them, such as Figure 5 The design rules for the sliding window size are as follows: the sliding window should be positioned so that, in the case of the vehicle's minimum turning radius, the front of the vehicle enters and exits from the opposite side of the sliding window, rather than from either side. This ensures that when the vehicle is traveling straight, the sliding window can cover the past half of the vehicle body and the future half of the vehicle body. When the vehicle is turning, the sliding window can cover the past half of the vehicle body and the future half of the vehicle body when the vehicle is turning at its minimum radius.

[0044] For example, in Figure 5 The diagram shows state ① where the vehicle is centered within the sliding window. When the vehicle is traveling straight, it enters the sliding window from the left side in state ③ and exits from the right side in state ②. When the vehicle makes a minimum-radius turn in state ⑤, both the front and rear of the vehicle can enter the sliding window from the left side, and exit from the right side in state ④. This size design also considers that the vehicle body is a rigid body, and its trajectory will not change abruptly within the vehicle's dimensions (m). A sliding window size of 2m comprehensively reflects the vehicle's possible trajectory within the window. However, a sliding window size larger than 2m results in a longer trajectory, which may lead to some information loss and reduced accuracy in curvature calculation.

[0045] Subsequently, when it was determined that... Figure 5-6 After the sliding window shown, a one-dimensional parameter Δθ is obtained by observing the change in heading angle within the sliding window, which is then used to segment the two-dimensional centerline F1 of the first lane. Specifically, this application calculates the maximum heading angle θ within the sliding window. max The sum of the minimum heading angle θ min The absolute value of the difference Δθ=|θ max -θ min In one specific implementation, the preset movement length d of the sliding window can be set to any value between 0.1m and 0.5m.

[0046] For example, Figure 6 This is a schematic diagram of an embodiment of the state of a sliding window on the center line of the first lane provided in this application. Taking the d11-d16 segment on the center line F1 of the first lane as an example, a sliding window is established at a preset length d interval on d11-d16, and the absolute value Δθ of the difference in heading angles within the sliding window is calculated. Figure 7A schematic diagram illustrating a variation of an embodiment of the absolute value of the difference in heading angles on the first lane centerline F1 provided in this application, wherein... Figure 6 The variation of the absolute value Δθ of the difference in heading angles within each sliding window between d11 and d16.

[0047] More specifically, combined Figure 6 and Figure 7 It can be seen that, Figure 6 The curve between d11 and d12 is a straight line, and vehicles travel in a straight line within this interval, making... Figure 7 The value of Δθ calculated by the sliding window established within this interval is basically stable near 0, approximating a straight line; while... Figure 6 The sliding window established by D12 initiates the vehicle's 90-degree turn. Figure 7 The corresponding Δθ increases from d12 and reaches its maximum value at d13; subsequently... Figure 6 Between d13 and d14, the Δθ calculated using a sliding window remains stable near its maximum value; finally, in Figure 6 Between d14 and d15, the Δθ calculated by the sliding window gradually decreases, and finally decreases to 0 after the sliding window is established at d15. In the subsequent d15-d16, Δθ stabilizes near 0, and the vehicle ends the turn and returns to the straight state.

[0048] For example, Figure 8 This is a schematic diagram of another embodiment of the state of the sliding window on the center line of the first lane provided in this application. Taking the d21-d25 segment on the center line F1 of the first lane as an example, a sliding window is established at a preset length d interval on d21-d25, and the absolute value Δθ of the difference in heading angles within the sliding window is calculated. Figure 9 A schematic diagram illustrating another embodiment of the variation of the absolute value of the difference in heading angles along the centerline of the first lane provided in this application, wherein... Figure 8 The variation of the absolute value Δθ of the difference in heading angles within each sliding window between d21 and d25.

[0049] Combination Figure 8 and Figure 9 ,exist Figure 6 As shown, at the sliding window established at d21, the vehicle makes a U-turn along the center line of the lane corresponding to the U-turn point. The calculated Δθ gradually increases, reaching its maximum value at the sliding window established at d22, and then remains there until the sliding window established at d23. After d23, Δθ begins to decrease, and after the sliding window established at d24, Δθ drops to 0. Subsequently, between d24 and d25, the vehicle will straighten its direction, causing some fluctuations in the calculated Δθ of the sliding window between d24 and d25.

[0050] By the above Figure 5- Figure 9 , the change rule of △θ calculated by the sliding window established in the part section of the first lane center line F1 is shown in the case of straight driving, turning, U-turn, etc. according to the first lane center line F1, and finally the △θ calculated by the sliding window of all parts of the first lane center line F1 between A-B can be obtained in the same way.

[0051] Exemplarily, it is assumed that Figure 10 The absolute value of the difference between the heading angles in the entire sliding window of the first lane center line provided in the present application is shown in the figure, at this time, the first lane center line F1 can be processed by segmentation according to the change of △θ in Figure 10 . Among them, in combination with Figure 5- Figure 9 , it can be concluded that Figure 10 The change rule of △θ of the first lane center line F1 shown in has the characteristics, and the first curve F1 is classified according to different characteristics, so as to split the first lane center line F1 into different sections, and the part curve in the split section is recorded as a section. The section category provided by the present application for segmenting the first lane center line F1 at least includes:

[0052] The section of △θ as the first change rule: when an interval of the first lane center line F1 is a straight line, the running track corresponding to the vehicle in this interval is straight driving, the △θ calculated in this interval is approximately a straight line and its value is near 0, therefore a first threshold M1 can be set to determine the θ of the first lane center line F1 belonging to the straight line as the section of the first change rule, for example Figure 10 The △θ in the interval where A-d31 is located is less than the first threshold M1, so the interval where A-d31 is located in the first lane center line F1 is recorded as the section of the first change rule of △θ, and according to the △θ and the first threshold M1, the interval where d32-d33 is located and the interval where d36-B is located in the first lane center line F1 can also be recorded as the section of the first change rule of △θ. The curves in the three sections are straight lines, and the running track of the vehicle is straight driving.

[0053] The section of △θ as the second change rule: when an interval of the first lane center line F1 is a wave peak, but the peak value of the curve calculated in this interval is greater than the first threshold M1 but less than the second threshold M2, for example, Figure 10 The peak value of △θ between d33-d34 and d34-d35 in can be recorded as the section of the second change rule of △θ, and in this section, the running track of the vehicle is small-amplitude turning.

[0054] The interval of the third type of variation of Δθ is: when an interval of the first lane center line F1 is a wave peak, the interval corresponds to a 90-degree turn of the running track of the vehicle, and Δθ calculated in the interval is greater than the second threshold M2. Therefore, the interval of the first lane center line F1 in which d31-d32 is located is determined as the interval of the third type of variation of Δθ. For example, when Δθ in d31-d32 is greater than M2, the interval of the first lane center line F1 in which d31-d32 is located is determined as the interval of the third type of variation of Δθ. In the interval, the running track of the vehicle is a 90-degree turn. Figure 10 The interval of the fourth type of variation of Δθ is: when an interval of the first lane center line F1 is a wave peak and the peak value is greater than the third threshold M3, the interval corresponds to a U-turn of the running track of the vehicle, and Δθ calculated in the interval is greater than the third threshold M3. Therefore, the interval of the first lane center line F1 in which d35-d36 is located is determined as the interval of the fourth type of variation of Δθ. In the interval, the running track of the vehicle is a U-turn. Meanwhile, the peak width of Δθ in the interval of the fourth type of variation is greater than the peak width of Δθ in the interval of the third type of variation, and the peak width of Δθ in the interval of the third type of variation is greater than the peak width of Δθ in the interval of the second type of variation.

[0055] The interval of the third type of variation of Δθ is: when an interval of the first lane center line F1 is a wave peak, the interval corresponds to a 90-degree turn of the running track of the vehicle, and Δθ calculated in the interval is greater than the second threshold M2. Therefore, the interval of the first lane center line F1 in which d31-d32 is located is determined as the interval of the third type of variation of Δθ. For example, when Δθ in d31-d32 is greater than M2, the interval of the first lane center line F1 in which d31-d32 is located is determined as the interval of the third type of variation of Δθ. In the interval, the running track of the vehicle is a 90-degree turn. Figure 10 The interval of the fourth type of variation of Δθ is: when an interval of the first lane center line F1 is a wave peak and the peak value is greater than the third threshold M3, the interval corresponds to a U-turn of the running track of the vehicle, and Δθ calculated in the interval is greater than the third threshold M3. Therefore, the interval of the first lane center line F1 in which d35-d36 is located is determined as the interval of the fourth type of variation of Δθ. In the interval, the running track of the vehicle is a U-turn. Meanwhile, the peak width of Δθ in the interval of the fourth type of variation is greater than the peak width of Δθ in the interval of the third type of variation, and the peak width of Δθ in the interval of the third type of variation is greater than the peak width of Δθ in the interval of the second type of variation.

[0056] In some embodiments, when the different variation rules of Δθ of the above four types of intervals are determined, the first lane center line F1 can be segmented according to the rules, so as to obtain a plurality of intervals of the first lane center line F1 presenting different variation rules. In a specific implementation, the intervals of the first type of variation A-d31, d32-d33 and d36-B in the first lane center line F1 can be first determined by the first threshold M1. Subsequently, for the remaining intervals d31-d32, d33-d36, the local maximum values of the curve between every two intervals of the first type of variation can be calculated to obtain a local maximum value point set {Δθ1, Δθ2…Δθ5}, and then the half-peak width P of each of the five local maximum values is calculated. Since Δθ4 and Δθ5 are included in the two half-peak widths at the same time, the two peak values are merged into an interval of the same type of variation, and the peak value is max{f2(Δθ4), f2(Δθ5)}. Finally, according to Δθ4 and Δθ5 in the local maximum value point set {Δθ1, Δθ2…Δθ5} after merging, the number of turns determined is four. Subsequently, the variation rules of Δθ of the intervals of the turns are determined by the second threshold M2 and the third threshold M3.

[0057] In some embodiments, for the curve between any two intervals of the first type of variation, the starting and ending points can be determined according to the following rules: When there is a turn between two straight segments greater than 2m, the starting and ending points of the turn are the ending point of the preceding straight segment and the starting point of the following straight segment, respectively. For example, for d31 and d32 between A-d31 and d32-d33, the starting point is d31 and the ending point is d32. When there are two independent turns between two straight segments greater than 2m, and there is a straight segment less than 2m between the two turns, the left and right parts of the straight segment are merged into the adjacent turning curve segment, with the midpoint of the straight segment as the dividing point. When there are multiple consecutive turns, the half-peak width of each peak is calculated. Find the nearest left minimum value point to the peak within a 2m range to the left of the half-peak width. Use the turning trajectory point corresponding to this minimum value point as the starting point of the turn, which is also the ending point of the previous turn or straight section. For example, between d32-d33 and d36-B, three consecutive parts can be obtained, with the starting and ending points being d33-d34, d34-d35, and d35-d36, respectively.

[0058] Finally, based on the above methods, such as Figure 11 The first lane centerline F1, as shown, is segmented according to the variation law of Δθ, resulting in multiple intervals of the first lane centerline F1 with different variation laws of Δθ: These are abbreviated as follows: Category 1: A-d31, d32-d33, and d36-B; Category 2: d33-d34 and d34-d35; Category 3: d31-d32; Category 4: d35-d36. These different intervals can be used for subsequent fitting with different parameters according to the category corresponding to the interval, achieving a differentiated fitting method for intervals with different characteristics. When fitting curves corresponding to situations such as 90-degree turns and U-turns, the number of fitting operations can be increased and the sampling point spacing increased. Conversely, when fitting curves corresponding to situations such as straight travel and small turns, the number of fitting operations can be reduced and the sampling point spacing decreased. This reduces the invalid computational load when fitting the first lane centerline from straight lines and improves the accuracy of fitting the turn signal position.

[0059] S103: Perform B-spline fitting on each curve segment in the center line F1 of the first lane.

[0060] For each different type of interval, the number of fitting iterations and the sampling point spacing used during B-spline fitting are different. In some embodiments, the number of fitting iterations a for the first type of interval ≤ the number of fitting iterations b for the second type of interval ≤ the number of fitting iterations c for the third type of interval ≤ the number of fitting iterations d for the fourth type of interval, and / or, the sampling point spacing j1 for the first type of interval ≤ the sampling point spacing j2 for the second type of interval ≤ the sampling point spacing j3 for the third type of interval ≤ the sampling point spacing j4 for the fourth type of interval.

[0061] In some embodiments, before the B-spline fitting process is performed, the sampling point interval of the to-be-processed section is first determined, since the sampling density of the control points, which are parameters used to adjust the shape of the curve in the B-spline fitting process of the curve, has a significant impact on the smoothing efficiency. The greater the change in the heading angle between the control points, the greater the position deviation of the generated point from the original control point, and the more times of adjustment are needed. Therefore, in the fitting process of the third type of section and the fourth type of section, the heading angle change can be restored by increasing the sampling density, thereby reducing the number of iterations in the fitting process and improving the efficiency. However, too high control point density can cause the curvature of the generated smooth curve to have burrs. According to the tuning experience of actual data, the sampling interval can be x in the fitting process of the fourth type of section, and the sampling interval is x in the fitting process of the third type of section and rounding up x preferably 0.1-0.2 m can ensure that the curvature of the smooth curve is smooth and reasonable while improving the efficiency.

[0062] Subsequently, in the process of B-spline fitting each curve in the first lane center line, for the first type of section and the second type of section, since the curvature change is not large, a certain accuracy requirement can be achieved directly through one fitting. For the third type of section and the fourth type of section, multiple fittings are needed, and in the fitting process, the control points of the fitted curve are adjusted to make the fitted second lane center line more approximate to the first lane center line, thereby optimizing the fitting effect. The specific way of adjusting the control points in the present application will be described below in conjunction with the drawings.

[0063] Figure 12 The schematic diagram of adjusting the control points in the B-spline fitting process provided by the present application is shown, wherein the third type of section d41-d42 on the first lane center line F1 is taken as an example, which corresponds to the driving trajectory of the vehicle turning 90 degrees. In the process of fitting the first lane center line F1, the first lane center line F1 is first fitted by B-spline to obtain a fitted curve T1, then a normal line is drawn for each control point on the first lane center line F1, for example, a normal line L1 is drawn for the control point k1 to obtain an intersection point q1 of the fitted curve T1, a normal line L2 is drawn for the control point k2 to obtain an intersection point q2 of the fitted curve T1, and a normal line L3 is drawn for the control point k3 to obtain an intersection point q3 of the fitted curve T1. Then, among all the normal lines drawn, the point with the maximum distance between the control point and the intersection point is taken, the control point k2 of the normal line L2 is farthest from the intersection point q2, then the control point k2 is moved outward from the convex hull in the direction of the normal line by a fixed step, thereby obtaining a new control point k2'. Then, the new control point k2' replaces the original control point k2 in the fitting process, and the control points k1, k2', k3, etc. are used to re-iterate the fitted curve T1 to obtain a fitted curve T2, whereinFigure 11- Figure 12 This is a schematic diagram illustrating an embodiment of the change in the fitted curve during the B-spline fitting process provided in this application. It can be seen that the fitted curve T2 obtained after fitting the adjusted control points is closer to the center line F1 of the first lane than the fitted curve T1 before the control point adjustment. This process is then repeated continuously. Figure 13 The process involves continuously selecting the control point whose normal to the center line F1 of the first lane has the largest distance from the currently obtained fitted curve. After moving the control point, a new control point is obtained to replace the original control point, and the iterative fitting is repeated. For example... Figure 14 This is a schematic diagram of another embodiment of the change in the fitted curve during the B-spline fitting process provided in this application. Normals are drawn from control points k1, k2, and k3 on the first lane centerline F1, and the intersection points of these normals and the fitted curve T2 are obtained. Then, among all the drawn normals, the point with the largest distance between the control point and the intersection point is selected. The control point of normal L3 has the largest distance to the intersection point. Next, control point k3 is moved a fixed step size outwards from the convex hull in the direction of the normal, thus obtaining a new control point k3'. Then, the new control point k3' replaces k2 in the original control point, and the fitted curve T2 is iteratively fitted again using control points k1, k2', k3', etc., to obtain the fitted curve T3. This process continues until, after n fitting iterations, the distance between the obtained fitted curve Tn and the first lane centerline F1 satisfies certain conditions, for example… Figure 15 The diagram shows the center line of the second lane obtained by B-spline fitting provided in this application. It can be seen that the fitting curve Tn obtained after multiple iterations of fitting is closer to the center line of the first lane F1. Therefore, the fitting curve Tn obtained at this time can be used as the center line of the second lane F2 after smoothing.

[0064] In some embodiments, the fitted curve obtained during the fitting process of the first curve F1 can be expressed by the following formula: Among them, P i The control point k2 selected for calculating curve C(u) is then moved to obtain control point k2', which is P. i +v, where v is the step size, which can be a unit in the coordinate system of the first curve F1, used to represent the distance the control point moves. The fitted curve obtained after fitting based on the new control points is expressed by the formula D(u)=C(u)+N. i,p (u)v.

[0065] In some embodiments, the step length moved each time when adjusting the control points can be a fixed value or can be adjusted. In some embodiments, the step length is related to the category of the interval being fitted, for example, the step length of the fourth category interval can be greater than the step length of the third category interval when fitting, so as to reduce the number of iterations required when fitting the fourth category interval to a certain extent.

[0066] Exemplarily, Figure 15 The B-spline fitting result after adjusting the control points of the curve provided in the present application is shown in the schematic diagram, wherein a U-turn curve is taken as an example, it can be seen that the fitting curve F2 represented by a solid line after fitting is very close to the first curve F1 represented by a dashed line at the U-turn, and the error is small, when the fitting curve F2 is used as a driving path, the vehicle can be made to drive closer to the center line of the lane. In addition, ​ The Q1 part of the fitting curve corresponds to the second category Δθ of the calculated heading angle on d51-d52, and the Q2 part corresponds to the fourth category Δθ of the calculated heading angle on d53-d54

[0067] S104: performing splicing processing on the curve after the fitting processing to obtain a second lane center line.

[0068] Finally, after performing B-spline fitting on each segment of the first lane center line in S103, the fitted curves are spliced to obtain a complete curve, which is recorded as a second lane center line. Since each segment of the curve is fitted in S103, and the starting point and the ending point of the curve remain unchanged during the fitting process, the second lane center line obtained after splicing can be equivalent to the first lane center line after smoothing processing, and the second lane center line can be used as a trajectory for subsequent vehicle driving. Since the second lane center line is smoothed, the problems such as sudden change of curvature in the first curve are eliminated, and the smoothness and curvature continuity of the second lane center line are ensured.

[0069] In summary, in the embodiment of the present application, the first lane center line obtained by interpolating the discrete lane center line data is segmented and processed, and the curves of different segments are respectively smoothed by B-spline fitting, and the fitting curves of multiple segments are spliced to obtain the second lane center line. Thus, the entire first lane center line does not need to be fitted, and for the curves corresponding to each type of different interval in the first lane center line, the fitting times and the sampling point interval used in B-spline fitting are different. When fitting the curves corresponding to the 90-degree turning and U-turn of the vehicle, the fitting times can be increased and the sampling point interval can be increased, while for the curves corresponding to the straight driving and small turning of the vehicle, the fitting times can be reduced and the sampling point interval can be reduced. When fitting the first lane center line, the calculation amount of processing the entire curve is reduced, and the steps in the process of smoothing the first lane center line by B-spline fitting are reduced, thereby improving the speed and efficiency of fitting the first lane center line of the lane center line. The obtained second lane center line can also ensure smoothness and improve the accuracy of the fitted curve, and provides a more accurate basis for the subsequent vehicle to take the second lane center line as the driving trajectory.

[0070] In addition, the present application also measures the curvature change of the first lane center line F1 in the range of the window by establishing a sliding window, and improves the calculation efficiency by segmenting the curve through the change of the heading angle in a manner more consistent with the actual driving conditions of the vehicle. Furthermore, the present application also iteratively adjusts the position of the control point during B-spline fitting of the curve, which can further improve the fitting accuracy during B-spline fitting, so that the second lane center line obtained by fitting is closer to the actual discrete lane center line, and the fitting accuracy is further improved.

[0071] In the foregoing embodiments, the lane center line smoothing method provided by the embodiments of the present application is introduced. In order to realize each function in the method provided by the embodiments of the present application, the electronic device as an execution subject can include a hardware structure and / or a software module to realize each function in the form of a hardware structure, a software module, or a hardware structure plus a software module. Whether a certain function in each function is executed in the form of a hardware structure, a software module, or a hardware structure plus a software module depends on the specific application of the technical solution and the design constraint conditions.

[0072] For example, the application also provides a lane center line smoothing device, comprising: an interpolation module, a classification module, a fitting module and a splicing module. The interpolation module is configured to perform quadratic interpolation processing on a plurality of discrete data of a lane center line to be smoothed to obtain a first lane center line. The classification module is configured to divide the first lane center line into a plurality of classified intervals according to the curvature of the first lane center line. The fitting module is configured to perform fitting processing on the plurality of classified intervals respectively to obtain a plurality of fitting curves. The splicing module is configured to perform splicing processing on the plurality of fitting curves to obtain a second lane center line. Specifically, the specific principles and implementation manners of the above steps performed by each module in the lane center line smoothing device can refer to the lane center line smoothing method in the foregoing embodiments of the application, and will not be described herein again.

[0073] It should be noted that the division of each module of the above device is only a logical functional division, and all or part of the modules can be integrated into one physical entity, or can be physically separated. These modules can all be implemented in the form of software called by a processing element; all can be implemented in the form of hardware; some modules can be implemented in the form of software called by a processing element, and some modules can be implemented in the form of hardware. A separate processing element can be provided, or the processing element can be integrated in a chip of the above device, and in addition, the functions of the above determination modules can be stored in the form of program code in the memory of the above device, and called and executed by a processing element of the above device. The implementation of other modules is similar. In addition, all or part of these modules can be integrated together, or can be independently implemented. The processing element described herein can be an integrated circuit having a signal processing capability. In the implementation process, each step of the above method or each of the above modules can be completed by an integrated logic circuit of hardware or an instruction in the form of software in the processing element.

[0074] For example, the above modules can be one or more integrated circuits configured to implement the above method, such as one or more application specific integrated circuits (ASICs), or one or more digital signal processors (DSPs), or one or more field programmable gate arrays (FPGAs), etc. For another example, when a certain module above is implemented in the form of scheduling program code by a processing element, the processing element can be a general-purpose processor, such as a central processing unit (CPU) or other processor that can call program code. For another example, these modules can be integrated together to implement in the form of a system on a chip (SOC).

[0075] In the embodiments described above, all or some of the steps can be implemented by hardware, software, firmware or any combination thereof. When implemented by software, all or some of the steps can be implemented in the form of one or more computer programs. The computer programs include one or more computer instructions. When the computer programs are loaded into and executed by a computer, all or some of the steps described in the embodiments of the present application are performed. The computer can be a general purpose computer, a special purpose computer, a computer network, or other programmable apparatuses. The computer instructions can be stored in a computer readable storage medium or transmitted from one computer readable storage medium to another computer readable storage medium, for example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center through wired (such as coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (such as infrared, wireless, microwave, etc.) manner. The computer readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server, data center, etc. integrated with one or more available media. The available media can be a magnetic medium (for example, floppy disk, hard disk, magnetic tape), an optical medium (for example, DVD), or a semiconductor medium (for example, solid state disk (SSD)) and the like.

[0076] The present application also provides an electronic device, comprising: a processor and a memory; wherein the memory stores a computer program, and when the processor executes the computer program, the processor can be used to execute the point cloud data smoothing method according to any one of the preceding embodiments of the present application.

[0077] The present application also provides a computer readable storage medium, which stores a computer program, and when the computer program is executed, the computer program can be used to execute the point cloud data smoothing method according to any one of the preceding embodiments of the present application.

[0078] The present application also provides a chip for executing instructions, and the chip is used to execute the point cloud data smoothing method performed by the electronic device according to any one of the preceding embodiments of the present application.

[0079] Those skilled in the art can understand that all or part of the steps of the methods described above can be completed by program instruction related hardware. The foregoing program can be stored in a computer readable storage medium. When the program is executed, the steps of the methods described above are executed; and the foregoing storage medium includes ROM, RAM, magnetic disk or optical disk and various media that can store program codes.

[0080] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, but not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for part or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for smoothing lane centerlines, characterized in that, include: The first lane centerline is obtained by performing a second interpolation on multiple discrete data of the smoothed lane centerline. Based on the curvature of the first lane centerline, the first lane centerline is divided into multiple categories of intervals; The intervals of the multiple categories are fitted separately to obtain multiple fitting curves, wherein different fitting parameters are used when fitting the intervals of each category. The multiple fitted curves are spliced ​​together to obtain the center line of the second lane; The division of the first lane centerline into multiple categories of intervals based on its curvature includes: A sliding window of a preset size is set at preset intervals along the center line of the first lane; The first lane centerline is classified according to the absolute value Δθ of the difference in heading angles of the first lane centerlines within all sliding windows on the first lane centerline.

2. The method according to claim 1, characterized in that, The center line of the first lane is divided into several categories of sections, including: The first type of interval: the interval on the Δθ change curve corresponding to the sliding window where the peak is less than the first threshold; wherein, the first type of interval corresponds to the straight-moving state of the vehicle; The second type of interval: the interval on the Δθ change curve corresponding to the waveform where the peak is greater than the first threshold and less than the second threshold; wherein, the second type of interval corresponds to the non-90-degree turning state of the vehicle; The third type of interval: the interval on the Δθ change curve corresponding to the waveform where the peak is greater than the first threshold and less than the third threshold; wherein, the third type of interval corresponds to the 90-degree turn state of the vehicle; The fourth type of interval: the interval on the Δθ change curve corresponding to the waveform where the peak is greater than the third threshold; wherein, the fourth type of interval corresponds to the U-turn state of the vehicle.

3. The method according to claim 2, characterized in that, The classification of the first lane centerline based on the absolute value Δθ of the difference in heading angles among all sliding windows along the first lane centerline includes: Based on the first threshold, a first type of interval is determined on the center line of the first lane; Identify multiple local maxima points on the centerline of the first lane, outside the first type of interval; The turning position on the center line of the first lane is determined based on the half-peak width corresponding to the multiple local maxima points. Based on the second threshold and the turning position, the second type of interval on the center line of the first lane is determined; Based on the third threshold and the turning position, the third type of interval and the fourth type of interval on the center line of the first lane are determined.

4. The method according to claim 3, characterized in that, Also includes: When the length of the first type of interval is greater than a preset threshold, the first interval on one side of the first type of interval and the endpoints adjacent to the first type of interval are separated. When the length of the first type of interval is less than a preset threshold, the minimum point of the distance between the half-peak in the third interval on one side of the first type of interval and the preset threshold in the direction of the first type of interval is taken as the endpoint of the third interval and the first type of interval.

5. The method according to any one of claims 1-4, characterized in that, The sliding window is rectangular in shape, with a side length twice the length of the vehicle; The preset distance is 0.1 to 0.5 times the length of the vehicle.

6. The method according to any one of claims 2-4, characterized in that, The fitting parameters include: the number of fitting iterations and the sampling point spacing; Among them, the number of times the first type of interval is fitted ≤ the number of times the second type of interval is fitted ≤ the number of times the third type of interval is fitted ≤ the number of times the fourth type of interval is fitted; The sampling point spacing of the first type of interval is less than the sampling point spacing of the second type of interval, which is less than the sampling point spacing of the third type of interval and less than the sampling point spacing of the fourth type of interval.

7. The method according to claim 6, characterized in that, By fitting an interval, a fitted curve for that interval is obtained, including: Determine the first control point of the first fitted curve obtained by curve fitting within the interval; The first control point is adjusted according to a preset rule to obtain the second control point; The first fitted curve is fitted using the second control point to obtain the second fitted curve; After iterating the fitting curve multiple times using the above fitting process, the fitting curve obtained from the last fitting process is taken as the fitting curve for the interval.

8. The method according to claim 7, characterized in that, Determining the first control point of the first fitted curve obtained by curve fitting within the interval includes: Determine the intersection points of the normals of multiple control points of the first fitted curve with the center line of the first lane; The control point with the largest distance from the plurality of control points to their corresponding intersection points is designated as the first control point; The step of adjusting the first control point according to a preset rule to obtain the second control point includes: The first control point is moved a predetermined distance outward from the bending direction of the curve within the interval to obtain the second control point.

9. A lane centerline smoothing device, characterized in that, include: The interpolation module is used to perform secondary interpolation on multiple discrete data of the lane centerline to be smoothed, so as to obtain the first lane centerline. The classification module is used to divide the center line of the first lane into multiple categories of intervals according to the curvature of the center line of the first lane; The fitting module is used to perform fitting processing on the intervals of the multiple categories respectively to obtain multiple fitting curves, wherein the fitting parameters used when fitting the intervals of each category are different. The splicing module is used to splice the multiple fitted curves to obtain the center line of the second lane; The classification module is specifically used to set a sliding window of a preset size at preset intervals on the center line of the first lane; and to classify the center line of the first lane according to the absolute value △θ of the difference in the heading angle of the center line of the first lane within all the sliding windows on the center line of the first lane.

Images