Lane line detecting and tracking method and device

Abstract

The embodiment of the invention provides a lane line detection tracking method and device, and relates to the technical field of image processing. The method comprises the following steps: acquiring an aerial view image of the surrounding environment of a vehicle at the current moment; detecting the aerial view image to obtain a visual point sequence corresponding to a plurality of lane lines in the aerial view image; on the basis of the position relation of the visual points corresponding to the multiple lane lines, abnormal visual points in the visual points corresponding to the multiple lane lines are removed to generate an initial lane line set; after the historical lane line and the initial lane line set are unified to the same coordinate system, obtaining a to-be-tracked lane line from the initial lane line set based on the position information and / or linear relationship between the historical lane line and each lane line in the initial lane line set; and fusing the to-be-tracked lane line with the corresponding historical lane line to obtain a target lane line. The embodiment of the invention is used for improving the lane line detection precision and avoiding the problem that the output result is inaccurate during lane line tracking becomes an urgent need to be solved.

Description

Technical Field

[0001] The present disclosure relates to the field of image processing technology, and in particular to a lane line detection and tracking method and device. Background Art

[0002] With the rapid development of autonomous driving technology, lane line detection and tracking has become a basic part of vehicle assisted driving and unmanned driving. Accurately detecting and identifying lane lines is an important prerequisite for realizing autonomous driving.

[0003] At present, when detecting and tracking lane lines based on bird's-eye view images, the vehicle's surrounding environment information is first obtained through sensors and other devices, and then the bird's-eye view (BEV) technology is used to generate lane lines in vector maps or other images; however, in actual road environments, there are a large number of complex situations, such as lane lines being blocked by other obstacles, lane lines being worn, and changes in ambient brightness, which may cause sudden changes in the position, length, shape and other elements of the same lane line, resulting in inaccurate output of lane lines in maps or images, resulting in a poor user experience. Therefore, how to improve the accuracy of lane line detection and avoid inaccurate results generated during lane line detection and tracking has become an urgent problem to be solved. Summary of the invention

[0004] In view of this, an embodiment of the present application provides a lane line detection and tracking method and device for improving the lane line detection accuracy and avoiding the inaccurate output results during lane line tracking, which becomes a problem that needs to be solved urgently.

[0005] In a first aspect, an embodiment of the present application provides a lane line detection and tracking method, comprising:

[0006] Obtain a bird's-eye view image of the vehicle's surroundings at the current moment;

[0007] Detecting the bird's-eye view image to obtain a visual point sequence corresponding to a plurality of lane lines in the bird's-eye view image;

[0008] Based on the positional relationship of the visual points corresponding to the multiple lane lines, abnormal visual points among the visual points corresponding to the multiple lane lines are removed to generate an initial lane line set;

[0009] After unifying the historical lane lines and the initial lane line set into the same coordinate system, and based on the position information and / or line type relationship of the historical lane lines, obtaining the lane lines to be tracked from the initial lane line set; wherein the historical lane lines are lane lines corresponding to the lane lines to be tracked obtained at historical moments;

[0010] The lane line to be tracked is merged with the corresponding historical lane line to obtain the target lane line.

[0011] As an optional implementation of the embodiment of the present application, the acquiring the lane line to be tracked from the initial lane line set based on the position information and line type relationship between the historical lane line and each lane line in the initial lane line set includes:

[0012] Acquire a lane line set to be tracked from the initial lane line set; the lane line set to be tracked includes a preset number of lane lines in the initial lane line set that are closest to the historical lane lines;

[0013] Based on the lane line parameter equation corresponding to the historical lane line, obtaining the target lane line parameter;

[0014] For each lane line in the set of lane lines to be tracked, obtaining a lane line parameter corresponding to the lane line;

[0015] The difference between the lane line parameters corresponding to each lane line in the lane line set to be tracked and the target lane line parameters is obtained, and the lane line with the smallest difference is taken as the lane line to be tracked.

[0016] As an optional implementation of the embodiment of the present application, the acquiring the lane line to be tracked from the initial lane line set based on the position information of the historical lane line and each lane line in the initial lane line set comprises:

[0017] Obtaining position information of each lane line in the initial lane line set;

[0018] Based on the position information of each lane line in the initial lane line set, the lane line closest to the historical lane line is obtained as the lane line to be tracked.

[0019] As an optional implementation of the embodiment of the present application, the acquiring the lane line to be tracked from the initial lane line set based on the line type relationship between the historical lane line and each lane line in the initial lane line set includes:

[0020] Based on the lane line parameter equation corresponding to the historical lane line, obtaining the target lane line parameter;

[0021] For each lane line in the initial lane line set, obtaining a lane line parameter corresponding to the lane line;

[0022] The difference between the lane line parameters corresponding to each lane line in the initial lane line set and the target lane line parameters is obtained, and the lane line with the smallest difference is used as the lane line to be tracked.

[0023] As an optional implementation of the embodiment of the present application, based on the positional relationship of the visual points corresponding to the multiple lane lines, removing abnormal visual points from the visual points corresponding to the multiple lane lines to generate an initial lane line set includes:

[0024] For each visual point sequence corresponding to the multiple lane lines, obtaining the radius of a circle formed by all three adjacent visual points in the visual point sequence corresponding to the lane line;

[0025] If the radius of a circle formed by any three adjacent visual points is less than or equal to a preset value, the intermediate visual point of any three adjacent visual points is removed; the intermediate visual point is determined based on the detection order of the three adjacent visual points;

[0026] and / or;

[0027] For each of the multiple lane lines, a visual point sequence corresponding to the lane line is used to obtain an angle formed by three adjacent visual points in the visual point sequence corresponding to the lane line.

[0028] If the included angle formed by any three adjacent visual points is less than or equal to a preset angle, the intermediate visual point among any three adjacent visual points is eliminated; the intermediate visual point is determined based on the detection order of the three adjacent visual points.

[0029] As an optional implementation of the embodiment of the present application, based on the positional relationship of the visual points corresponding to the multiple lane lines, the abnormal visual points among the visual points corresponding to the multiple lane lines are removed to generate an initial lane line set; including:

[0030] For each lane line of the multiple lane lines, when the visual points in the visual point sequence corresponding to the lane line are continuous abnormal visual points and the number is equal to a preset value, the last abnormal visual point is retained.

[0031] As an optional implementation of the embodiment of the present application, before fusing the lane line to be tracked with the corresponding historical lane line to obtain the target lane line, the method further includes:

[0032] Obtaining the distance between the lane line to be tracked and the historical lane line;

[0033] If the distance is less than or equal to the threshold, the verification is passed, and the lane line to be tracked is fused with the corresponding historical lane line to obtain the target lane line.

[0034] As an optional implementation of the embodiment of the present application, before fusing and outputting the lane line to be tracked with the corresponding historical lane line, the method further includes:

[0035] After the lane line to be tracked passes the verification, the tracking times are increased by one;

[0036] After the lane line to be tracked fails to pass the verification, the number of times of not tracking is increased by one;

[0037] When the number of tracking times of the lane line to be tracked is greater than or equal to the first preset number, and the position information is beyond the preset distance behind the position information of the current vehicle, the lane line to be tracked is removed;

[0038] When the tracking times of the lane line to be tracked are less than the first preset times, and the non-tracking times are greater than the second preset times, the lane line to be tracked is removed.

[0039] In a second aspect, an embodiment of the present application provides a lane line detection and tracking device, comprising:

[0040] An acquisition unit, used to acquire a bird's-eye view image of the vehicle's surroundings at the current moment;

[0041] A detection unit, configured to detect the bird's-eye view image and obtain a visual point sequence corresponding to a plurality of lane lines in the bird's-eye view image;

[0042] A removing unit, configured to remove abnormal visual points from the visual points corresponding to the multiple lane lines based on the positional relationship of the visual points corresponding to the multiple lane lines, so as to generate an initial lane line set;

[0043] A processing unit, configured to unify the historical lane lines and the initial lane line set into the same coordinate system, and then obtain the lane line to be tracked from the initial lane line set based on the position information and / or line type relationship between the historical lane lines and each lane line in the initial lane line set; wherein the historical lane line is a lane line corresponding to the lane line to be tracked obtained at a historical moment;

[0044] A fusion unit is used to fuse the lane line to be tracked with the corresponding historical lane line to obtain a target lane line.

[0045] As an optional implementation of the embodiment of the present application, the processing unit is specifically used to obtain a lane line set to be tracked from the initial lane line set; the lane line set to be tracked includes a preset number of lane lines in the initial lane line set that are closest to the historical lane lines; based on the lane line parameter equation corresponding to the historical lane line, the target lane line parameters are obtained; for each lane line in the lane line set to be tracked, the lane line parameters corresponding to the lane line are obtained; the difference between the lane line parameters corresponding to each lane line in the lane line set to be tracked and the target lane line parameters is obtained, and the lane line with the smallest difference is taken as the lane line to be tracked.

[0046] As an optional implementation of the embodiment of the present application, the acquisition unit is specifically used to obtain the position information of each lane line in the initial lane line set; based on the position information of each lane line in the initial lane line set, obtain the lane line closest to the historical lane line as the lane line to be tracked.

[0047] As an optional implementation of the embodiment of the present application, the acquisition unit is specifically used to acquire the target lane line parameters based on the lane line parameter equation corresponding to the historical lane line; for each lane line in the initial lane line set, acquire the lane line parameters corresponding to the lane line; acquire the difference between the lane line parameters corresponding to each lane line in the initial lane line set and the target lane line parameters, and take the lane line with the smallest difference as the lane line to be tracked.

[0048] As an optional implementation manner of the embodiment of the present application, the elimination unit is specifically used to obtain, for each lane line in the multiple lane lines, a circular radius composed of all three adjacent visual points in the visual point sequence corresponding to the lane line; if the circular radius composed of any three adjacent visual points is less than or equal to a preset value, then the middle visual point of any three adjacent visual points is eliminated; the middle visual point is determined based on the detection order of the three adjacent visual points; and / or; for each lane line in the multiple lane lines, obtain the angle composed of three adjacent visual points in the visual point sequence corresponding to the lane line; if the angle composed of any three adjacent visual points is less than or equal to a preset angle, then the middle visual point of any three adjacent visual points is eliminated; the middle visual point is determined based on the detection order of the three adjacent visual points.

[0049] As an optional implementation of an embodiment of the present application, the elimination unit is specifically used to retain the last abnormal visual point for each lane line among the multiple lane lines when the visual points in the visual point sequence corresponding to the lane line are continuous abnormal visual points and the number is equal to a preset value.

[0050] As an optional implementation of the embodiment of the present application, before the lane line to be tracked is merged with the corresponding historical lane line to obtain the target lane line, the lane line detection and tracking device also includes: a verification unit, which is specifically used to obtain the distance between the lane line to be tracked and the historical lane line; if the distance is less than or equal to a threshold, the verification is passed, and the lane line to be tracked is merged with the corresponding historical lane line to obtain the target lane line.

[0051] As an optional implementation of the embodiment of the present application, before the lane line to be tracked is merged with the corresponding historical lane line and outputted, the elimination unit is also used to increase the tracking times by one after the lane line to be tracked passes the verification; increase the non-tracking times by one after the lane line to be tracked fails the verification; when the tracking times of the lane line to be tracked is greater than or equal to a first preset times, and the position information is beyond a preset distance behind the position information of the current vehicle, the lane line to be tracked is eliminated; when the tracking times of the lane line to be tracked is less than the first preset times, and the non-tracking times are greater than the second preset times, the lane line to be tracked is eliminated.

[0052] The lane line detection and tracking method provided in the embodiment of the present application is specifically as follows: obtaining a bird's-eye view image of the vehicle's surrounding environment at the current moment; detecting the bird's-eye view image to obtain a sequence of visual points corresponding to multiple lane lines in the bird's-eye view image; based on the positional relationship of the visual points corresponding to the multiple lane lines, removing abnormal visual points from the visual points corresponding to the multiple lane lines to generate an initial lane line set; after unifying the historical lane lines and the initial lane line set into the same coordinate system, obtaining the lane line to be tracked from the initial lane line set based on the position information and / or line type relationship between the historical lane lines and each lane line in the initial lane line set; wherein the historical lane line is a lane line corresponding to the lane line to be tracked obtained at a historical moment; and the lane line to be tracked is fused with the corresponding historical lane line to obtain the target lane line. The embodiment of the present application obtains the positional relationship of visual points in the visual point sequence corresponding to the multiple lane lines, and then eliminates abnormal visual points among the visual points corresponding to the multiple lane lines, thereby improving the accuracy of lane line detection; at the same time, combined with the position information and / or line type relationship of the historical lane line, the lane line to be tracked corresponding to the historical lane line is obtained, and finally the historical lane line is merged with the lane line to be tracked to obtain the target lane line, further ensuring that the lane line to be tracked is the lane line of the historical lane line at the current moment, thereby improving the accuracy of the results output during lane line tracking. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] The accompanying drawings herein are incorporated in and constitute a part of the specification, illustrate embodiments consistent with the present application, and together with the description, are used to explain the principles of the present application.

[0054] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings required in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0055] Figure 1One of the step flow charts of the lane line detection and tracking method provided in an embodiment of the present application;

[0056] Figure 2 One of the schematic diagrams of the lane line detection and tracking method provided in the embodiment of the present application;

[0057] Figure 3 A second flowchart of the lane line detection and tracking method provided in an embodiment of the present application;

[0058] Figure 4 A second schematic diagram of the lane line detection and tracking method provided in an embodiment of the present application;

[0059] Figure 5 A third schematic diagram of the lane line detection and tracking method provided in an embodiment of the present application;

[0060] Figure 6 The third step flow chart of the lane line detection and tracking method provided in the embodiment of the present application;

[0061] Figure 7 A fourth schematic diagram of the lane line detection and tracking method provided in an embodiment of the present application;

[0062] Figure 8 A schematic diagram of the structure of a lane line detection and tracking device provided in an embodiment of the present application;

[0063] Fig. 9 A schematic diagram of the hardware structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0064] In order to more clearly understand the above-mentioned objectives, features and advantages of the present disclosure, the scheme of the present disclosure will be further described below. It should be noted that the embodiments of the present disclosure and the features in the embodiments can be combined with each other without conflict.

[0065] In the following description, many specific details are set forth to facilitate a full understanding of the present disclosure, but the present disclosure may also be implemented in other ways different from those described herein; it is obvious that the embodiments in the specification are only part of the embodiments of the present disclosure, rather than all of the embodiments.

[0066] In the embodiments of the present application, words such as "exemplary" or "for example" are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a concrete way. In addition, in the description of the embodiments of the present application, unless otherwise specified, the meaning of "multiple" refers to two or more.

[0067] It should be noted that, in this article, the term "comprises" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or device including the element.

[0068] Based on the above content, the embodiment of the present application provides a lane line detection and tracking method, referring to Figure 1 As shown, the lane line detection and tracking method includes the following steps S101-S105:

[0069] S101. Obtain a bird's-eye view image of the vehicle's surroundings at the current moment.

[0070] In some embodiments, the method for obtaining a bird's-eye view image of the vehicle's surroundings at the current moment may be: by collecting the current vehicle's surrounding environment information through multiple sensors arranged around the vehicle body, and then obtaining a bird's-eye view image based on the current surrounding environment information after algorithm calculation and coordinate conversion, etc., wherein the bird's-eye view image (Bird's Eye View, BEV) is a perspective or coordinate system (3D space) used to describe the perceived real world. In an autonomous driving vehicle, sensors include cameras, millimeter-wave radars, laser radars, etc. The traditional image space perception method in the autonomous driving system is to analyze and calculate the data collected by different sensors such as radars and cameras on the car separately, and then merge the analysis results of each channel into a unified spatial coordinate system for planning the vehicle's driving trajectory.

[0071] S102: Detect the bird's-eye view image to obtain a visual point sequence corresponding to a plurality of lane lines in the bird's-eye view image.

[0072] In some embodiments, by detecting the lane lines in the bird's-eye view, a visual point coordinate sequence corresponding to the plurality of lane lines can be obtained. Figure 2 As shown, through detection Figure 2 The lane line in A is obtained as follows Figure 2The visual point sequence shown in B; in the actual detection process, since the images around the vehicle are collected during the operation of the vehicle, it will be affected by factors such as vehicle engine vibration, vehicle shaking or ambient lighting, lane line wear, obstacle occlusion, etc., resulting in the problem of uneven visual point sequence. If the lane line is still tracked and fused according to the uneven visual point sequence, a large error will be generated with the actually displayed lane line. Therefore, the embodiment of the present application removes abnormal points in the visual point sequence to make the visual point sequence area smooth to obtain accurate lane lines. The specific method refers to the following step S103.

[0073] S103: based on the positional relationship between the visual points corresponding to the multiple lane lines, remove abnormal visual points from the visual points corresponding to the multiple lane lines to generate an initial lane line set.

[0074] In some embodiments, reference Figure 2 As shown in C, by removing the abnormal points in the visual point sequence corresponding to multiple lane lines, we can obtain Figure 2 The relatively smooth visual point sequence corresponding to the multiple lane lines shown in C.

[0075] Furthermore, after removing abnormal points in the visual point sequence corresponding to the multiple lane lines, the corresponding lane lines can be obtained through the visual point sequence to generate the initial lane line set.

[0076] S104: After unifying the historical lane lines and the initial lane line set into the same coordinate system, based on the position information and / or line type relationship between the historical lane lines and each lane line in the initial lane line set, obtain the lane line to be tracked from the initial lane line set.

[0077] Wherein, the historical lane line is a lane line corresponding to the lane line to be tracked obtained at a historical moment;

[0078] In some embodiments, first, it is necessary to obtain the historical lane lines (the historical lane lines can be obtained from the stored data, such as the historical lane lines can be the lane lines obtained by the vehicle at multiple consecutive moments before the current moment), and then unify the historical lane lines and the initial lane line set into the same coordinate system to facilitate subsequent calculation and processing; specifically, the historical lane lines and the lane lines in the initial lane line set can be unified into the world coordinate system; then, based on the position information and / or line type relationship between the historical lane lines and each lane line in the initial lane line set, the lane lines to be tracked are obtained from the initial lane line set. This is to facilitate further screening in the initial lane lines, and obtain the lane line with the closest distance and the closest slope to the historical lane as the lane line to be tracked.

[0079] S105: Fusing the lane line to be tracked with the corresponding historical lane line to obtain a target lane line.

[0080] In some embodiments, after obtaining the lane line to be tracked, the lane line with tracking needs to be verified, that is, verifying that the lane line to be tracked is the state of the historical lane line at the historical moment at the current moment. After the lane line to be tracked is verified, the lane line to be tracked is merged with the corresponding historical lane line and output. Specifically, the lane line to be tracked is first merged with the historical lane line, and the weight parameters of the trajectory points on the lane line to be tracked and the trajectory points on the historical lane line can be calculated respectively by the distance between the trajectory points on the lane line to be tracked and the trajectory points on the historical lane line and the current position information of the vehicle. And based on the weight parameters, the lane line to be tracked and the historical lane line are weightedly merged to obtain the final lane line to be output.

[0081] Optionally, before outputting the final lane line to be output, it is also necessary to analyze the final lane line to be output based on the cubic B-spline interpolation method to obtain the fitting equation corresponding to the final lane line to be output, and the fitting equation is to optimize the final lane line to be output, wherein the cubic B-spline interpolation method is a commonly used numerical analysis method, which can fit a smooth continuous function curve through a given set of scattered data. The basic idea is to approximate the data in a small interval with a low-order polynomial, and use the connection conditions of the joints of these polynomials to ensure the smoothness of the entire curve. It is widely used in data approximation and interpolation problems in engineering, science, mathematics and other fields.

[0082] Optionally, before outputting the final lane line to be output, the final lane line to be output may be secondarily smoothed based on an OSQP solver, where the OSQP solver is a numerical optimization software for solving QP problems and can be used to smooth curves.

[0083] In some embodiments, the lane line to be tracked can be applied to environment information display (Environment Information Display, EID) and control of autonomous driving vehicles, etc.

[0084] The lane line detection and tracking method provided in the embodiment of the present application is specifically as follows: obtaining a bird's-eye view image of the vehicle's surrounding environment at the current moment; detecting the bird's-eye view image to obtain a sequence of visual points corresponding to multiple lane lines in the bird's-eye view image; based on the positional relationship of the visual points corresponding to the multiple lane lines, removing abnormal visual points from the visual points corresponding to the multiple lane lines to generate an initial lane line set; after unifying the historical lane lines and the initial lane line set into the same coordinate system, obtaining the lane line to be tracked from the initial lane line set based on the position information and / or line type relationship between the historical lane lines and each lane line in the initial lane line set; wherein the historical lane line is a lane line corresponding to the lane line to be tracked obtained at a historical moment; and the lane line to be tracked is fused with the corresponding historical lane line to obtain the target lane line. The embodiment of the present application obtains the positional relationship of visual points in the visual point sequence corresponding to the multiple lane lines, and then eliminates abnormal visual points among the visual points corresponding to the multiple lane lines, thereby improving the accuracy of lane line detection; at the same time, combined with the position information and / or line type relationship of the historical lane line, the lane line to be tracked corresponding to the historical lane line is obtained, and finally the historical lane line is merged with the lane line to be tracked to obtain the target lane line, further ensuring that the lane line to be tracked is the lane line of the historical lane line at the current moment, thereby improving the accuracy of the results output during lane line tracking.

[0085] As an extension and refinement of the above embodiment, the present application embodiment provides a lane line detection and tracking method, referring to Figure 3 As shown, the lane line detection and tracking method includes the following steps S301 to S304:

[0086] S301, obtaining a bird's-eye view image of the vehicle's surroundings at the current moment.

[0087] S302: Detect the bird's-eye view image to obtain a visual point sequence corresponding to a plurality of lane lines in the bird's-eye view image.

[0088] In some embodiments, the method for detecting the bird's-eye view image and obtaining a visual point sequence corresponding to a plurality of lane lines in the bird's-eye view image in step S102 may be as follows:

[0089] The bird's-eye view image is input into a lane line detection and tracking model to obtain a visual point sequence corresponding to a plurality of lane lines in the bird's-eye view image.

[0090] Among them, the lane line detection and tracking model is a model obtained by training a machine model based on sample data, and the sample data includes multiple sample bird's-eye view images and visual point sequences corresponding to multiple lane lines generated by each sample bird's-eye view image.

[0091] In some embodiments, the lane line detection and tracking model can be a globally associated lane line detection and tracking network model (Guided Aggregation Net, GANet) based on visual points to detect lane lines, where each visual point is directly regressed to the starting point of the lane line instead of expanding point by point. Specifically, the association of a visual point with its lane line is achieved by globally predicting its offset to the starting point of the corresponding lane, which is independent of each other and can be performed in parallel to greatly improve efficiency. In addition, a lane-aware feature aggregator (LFA) is further proposed, which adaptively captures the local correlation between adjacent visual points to supplement the local information of the global association.

[0092] S303: based on the positional relationship between the visual points corresponding to the multiple lane lines, remove abnormal visual points from the visual points corresponding to the multiple lane lines to generate an initial lane line set.

[0093] In some embodiments, in the above step S103, based on the positional relationship of the visual points corresponding to the multiple lane lines, the method of removing abnormal visual points from the visual points corresponding to the multiple lane lines can refer to the following method:

[0094] Method 1:

[0095] Step 1: for each lane line in the plurality of lane lines, a visual point sequence corresponding to the lane line is obtained, wherein the radius of a circle formed by all three adjacent visual points in the visual point sequence corresponding to the lane line is obtained.

[0096] Step 2: If the radius of a circle formed by any three adjacent visual points is less than or equal to a preset value, the middle visual point of any three adjacent visual points is removed.

[0097] The intermediate visual point is determined based on the detection order of the three adjacent visual points.

[0098] In some embodiments, the radius of a circle formed by three adjacent points in the visual point sequence is obtained, referring to Figure 4As shown, the circle O1 composed of three adjacent points a, b, c obtains its radius R1; the circle O2 composed of three adjacent points d, e, f obtains its radius R2; it can be found that the radius of the circle composed of three adjacent points a, b, c is much smaller than the radius of the circle composed of three adjacent points d, e, f. The reason is that the points d, e, f are smoother than the points a, b, c. Therefore, the angle formed by the three adjacent points can reflect the degree of smoothness between the three points. Then a preset value can be set. By comparing the relationship between the radius of the circle composed of the three adjacent points and the preset value, it can be known whether the middle point of the three adjacent points meets the conditions. If not, the middle visual point of the three visual points will be eliminated. It should be noted that the visual point sequence in the embodiment of the present application has a corresponding detection order. Between three adjacent points, the corresponding detection order can be determined according to the detection order, and then the middle point is determined; exemplarily, when the three adjacent points can be arranged in the detection order as: a, b, c, that is, the middle point is b; wherein, the detection order can be pre-set, such as determining the detection order according to a specific direction (the driving direction of the vehicle, or the direction of the line connecting the head and tail points in the visual point sequence, etc.), the three adjacent points can be arranged according to the front and back relationship in the specific direction, and then the middle point is determined.

[0099] Specifically, a circle formed by three adjacent points can be obtained by a three-point circle determination algorithm, and the radius corresponding to the circle can be further obtained. The three-point circle determination algorithm is implemented based on the three-point circle determination theorem, which states that three non-collinear points can uniquely determine a circle.

[0100] Method 2:

[0101] Step 3: For each lane line in the plurality of lane lines, a visual point sequence corresponding to the lane line is obtained, wherein an angle formed by three adjacent visual points in the visual point sequence corresponding to the lane line is obtained.

[0102] Step 4: If the included angle formed by any three adjacent visual points is less than or equal to a preset angle, the middle visual point of any three adjacent visual points is removed.

[0103] The intermediate visual point is determined based on the detection order of the three adjacent visual points.

[0104] In some embodiments, the angle formed by three adjacent points in the visual point sequence is obtained, referring to Figure 5As shown, the angle abc formed by the three adjacent points a, b, and c is used to obtain the size of the angle abc; the size of the angle def formed by the three adjacent points d, e, and f is used to obtain the size of the angle def; it can be found that the angle formed by the three adjacent points a, b, and c is much smaller than the angle formed by the three adjacent points d, e, and f. The reason is that the points d, e, and f are smoother than the points a, b, and c. Therefore, the radius of the circle formed by the three adjacent points can reflect the degree of smoothness between the three points. Then a preset angle can be set. By comparing the size relationship between the angle formed by the three adjacent points and the preset angle, it can be known whether the middle point of the three adjacent points meets the conditions. If not, the middle visual point of the three visual points will be eliminated. It should be noted that the visual point sequence in the embodiment of the present application has a corresponding detection order. Between the three adjacent points, the corresponding detection order can be determined according to the detection order, and then the middle point is determined; wherein the detection order can be pre-set.

[0105] It should be noted that the lane line detection and tracking method also includes: for each lane line among the multiple lane lines, when the visual points in the visual point sequence corresponding to the lane line are continuous abnormal visual points and the number is equal to a preset value, the last abnormal visual point is retained.

[0106] In some embodiments, the preset value can be 10, that is, when 10 consecutive points are all abnormal key points, the 10th abnormal visual point is retained and used as a reference. If all abnormal visual points are deleted, once a large number of consecutive visual points are deleted, then when generating lane lines, there are too many intermediate visual points, and the error of the described lane lines will be greater, and the accuracy cannot be guaranteed. Therefore, in order to ensure the reference of the visual points, when the continuous points are eliminated and the number reaches the preset value, the last visual point is retained.

[0107] S304: After unifying the historical lane lines and the initial lane line set into the same coordinate system, based on the position information and / or line type relationship between the historical lane lines and each lane line in the initial lane line set, obtain the lane line to be tracked from the initial lane line set.

[0108] The historical lane line is a lane line obtained at a historical moment and corresponding to the lane line to be tracked.

[0109] Optionally, in the embodiment of the present application, the acquiring of the lane line to be tracked from the initial lane line set may be based on the position information and line type relationship between the historical lane line and each lane line in the initial lane line set, and the lane line to be tracked is acquired from the initial lane line set, specifically including the following steps 11 to 14:

[0110] Step 11: Acquire a lane line set to be tracked from the initial lane line set.

[0111] The lane line set to be tracked includes a preset number of lane lines in the initial lane line set that are closest to the historical lane lines.

[0112] In some embodiments, the target lane line of the previous frame is the historical lane line of the current frame (i.e., the current moment), and the historical lane line and the initial lane line set need to be unified into the same coordinate system, and then based on the position information of the historical lane line of the current frame, a preset number of lane lines closest to the historical lane line are obtained from the initial lane line set to form the lane line set to be tracked. Figure 6 As shown, it is a schematic diagram of the lane line set to be tracked, including: historical lane line AB, three lane lines to be tracked CD, EF, GH. Specifically, the distance between the lane line to be tracked and the historical lane line can be obtained by comparing the distance between the head and tail endpoints of the historical lane line and the head and tail endpoints of the lane line to be tracked, or by comparing the vertical distances from the head and tail endpoints of the historical lane line to the lane line to be tracked.

[0113] Step 12: Based on the lane line parameter equation corresponding to the historical lane line, obtain the target lane line parameters.

[0114] In some embodiments, reference Figure 7 As shown, the target lane line parameters are obtained based on the lane line parameter equation, and the formula of the lane line parameter equation is obtained as follows:

[0115] y=C 0 +C 1 x+C 2 x 2 +C 3 x 3

[0116] Among them, C 0 is the horizontal coordinate value of the intersection of the extended line of the lane line and the x-axis, C 1 is the slope of the lane line, C 2 represents the curvature of the lane line and is positive if it bends to the right, C 3 Represents the rate of change of lane curvature. When it is positive, it means that the radius of curvature is gradually decreasing. When it is negative, it means that the radius of curvature is gradually increasing.

[0117] It should be noted that in the embodiment of the present application, the main purpose is to obtain the C in the lane line equation 0 and C 1 , the historical lane line can be compared with the C of each lane line in the lane line set to be tracked 0 and C 1, comprehensively compare and obtain the lane line in the lane line set to be tracked that is closest to the historical lane line as the lane line to be tracked.

[0118] Step 13: For each lane line in the set of lane lines to be tracked, obtain a lane line parameter corresponding to the lane line.

[0119] In the embodiment of the present application, the lane line parameters corresponding to each lane line in the lane line set to be tracked are also obtained based on the lane line parameter equation.

[0120] Specifically, when the lane line parameter in the initial lane line set is C 0 When the C of the historical lane line is obtained first 0 , i.e., the target lane line parameter; then obtain the target lane line parameter and the C of each lane line in the lane line set to be tracked 0 , that is, the lane line parameters, the smaller the difference between the two, the closer the two are to the x-axis intersection position information.

[0121] When the lane line parameter is compared to C 1 When the C of the historical lane line is obtained first 1 , i.e., the target lane line parameter; then obtain the target lane line parameter and the C of each lane line in the lane line set to be tracked 1 , that is, the lane line parameters, the smaller the difference between the two, the closer their slopes are.

[0122] Step 14: Obtain the difference between the lane line parameters corresponding to each lane line in the lane line set to be tracked and the target lane line parameters, and take the lane line with the smallest difference as the lane line to be tracked.

[0123] In some embodiments, the comprehensive comparison C 0 and C 1 Size, obtain a lane line with a small difference in x-axis intersection position information and a similar slope to the historical lane line from the lane line set to be tracked as the lane line to be tracked, so as to ensure that the lane line to be tracked is the state of the historical lane line in the current frame.

[0124] Optionally, in the embodiment of the present application, the acquiring the lane line to be tracked from the initial lane line set may be: acquiring the lane line to be tracked from the initial lane line set based on the position information of the historical lane line and each lane line in the initial lane line set. The acquiring the lane line to be tracked from the initial lane line set may specifically include the following steps 21-22:

[0125] Step 21: Obtain the position information of each lane line in the initial lane line set.

[0126] In some embodiments, it is also necessary to unify the historical lane lines and the initial lane line set into the same coordinate system, and then obtain the position information of each lane line in the initial lane line set.

[0127] Step 22: Based on the position information of each lane line in the initial lane line set, obtain the lane line closest to the historical lane line as the lane line to be tracked.

[0128] In some embodiments, it is also possible to obtain the lane line closest to the historical lane line as the lane line to be tracked based only on the position information of each lane line in the initial lane line set. Compared with judging by the lane line distance and line type relationship, obtaining the lane line to be tracked from the initial lane line set based only on the lane line distance is more efficient and will also occupy less computing resources.

[0129] Optionally, in the embodiment of the present application, the acquiring the lane line to be tracked from the initial lane line set may also be: acquiring the lane line to be tracked from the initial lane line set based on the line type relationship between the historical lane line and each lane line in the initial lane line set; specifically including the following steps 31-32:

[0130] Step 31: Obtain target lane line parameters based on the lane line parameter equation corresponding to the historical lane line.

[0131] In some embodiments, it is also necessary to unify the historical lane lines and the initial lane line set into the same coordinate system, and then obtain the target lane line parameters based on the lane line parameter equation corresponding to the historical lane lines.

[0132] Specifically, for obtaining the target lane line parameters, the method described in step 12 above may be referred to, and details will not be repeated here.

[0133] Step 32: For each lane line in the initial lane line set, obtain a lane line parameter corresponding to the lane line.

[0134] Specifically, for each lane line in the initial lane line set, the specific method for obtaining the lane line parameters corresponding to the lane line can refer to the above step 13, which will not be repeated here.

[0135] Step 33: Obtain the difference between the lane line parameters corresponding to each lane line in the initial lane line set and the target lane line parameters, and take the lane line with the smallest difference as the lane line to be tracked.

[0136] In some embodiments, after obtaining the lane line corresponding to each lane line in the initial lane line set, 0 and C1 Afterwards, the C of the historical lane line 0 and C 1 A one-to-one comparison is performed, and the lane line with the smallest difference with the historical lane line is selected from the initial lane line set as the lane line to be tracked.

[0137] In some embodiments, it is also possible to obtain the lane line in the initial set with the smallest lane line parameter difference with the historical lane line based only on the line type relationship between each lane line in the initial lane line set and the historical lane line as the lane line to be tracked. Compared with judging by both lane line distance and line type relationship, obtaining the lane line to be tracked from the initial lane set based only on the line type relationship of the lane line is more efficient and will also occupy less computing resources.

[0138] S305: Obtain the distance between the lane line to be tracked and the historical lane line.

[0139] In some embodiments, after determining the lane line to be tracked, it is also necessary to verify the lane line to be tracked to ensure that the lane line to be tracked is the lane line of the historical lane line at the current moment, so as to further improve the accuracy of the output target lane line. The specific verification method is to judge the distance between the lane line to be tracked and the historical lane line and the threshold; if it is less than or equal to the threshold, it means that the lane line to be tracked is relatively close to the historical lane line, and the probability that the lane line to be tracked is the lane line of the historical lane line at the current moment is very high, and then the subsequent fusion output step is performed; if it is greater than the threshold, it means that the lane line to be tracked is relatively far from the historical lane line, and the probability that the lane line to be tracked is the lane line of the historical lane line at the current moment is very small, and then the subsequent fusion output step is not performed.

[0140] S306: If the distance is less than or equal to the threshold, the verification is passed, and the lane line to be tracked is merged with the corresponding historical lane line to obtain the target lane line.

[0141] In some embodiments, the lane line to be tracked is continuously verified, and only when the verification passes, the lane line to be tracked is merged with the corresponding historical lane line to obtain the target lane line.

[0142] In an embodiment of the present application, the lane line detection and tracking method further includes: after the lane line to be tracked passes verification, the number of tracking times is increased by one; after the lane line to be tracked fails verification, the number of non-tracking times is increased by one.

[0143] Specifically, setting the tracking times and the untracking times for the lane line to be tracked is used to analyze whether the state of the target tracking in the current frame is stable. If it is unstable, it will be removed. If it is stable, it will be retained and output for display. The specific implementation method includes the following:

[0144] When the number of tracking times of the lane line to be tracked is greater than or equal to the first preset number, and the position information is beyond the preset distance behind the position information of the current vehicle, the lane line to be tracked is removed.

[0145] Exemplarily, when the lane line to be tracked is tracked more than or equal to 5 times in the current frame and before, and its position information is 20 meters behind the current vehicle's position information, it indicates that the lane line to be tracked may no longer be near the current vehicle's position information and is removed.

[0146] When the tracking times of the lane line to be tracked are less than the first preset times, and the non-tracking times are greater than the second preset times, the lane line to be tracked is removed.

[0147] Exemplarily, when the lane line to be tracked is tracked less than 5 times in the current frame and before, and the number of untracked times is greater than 30 times, it indicates that the number of untracked times of the lane line to be tracked is too high and extremely unstable, and there may be errors in the detection, so it is eliminated.

[0148] It should be noted that after the lane line to be tracked passes the verification, the lane line to be tracked is saved in a grid of a preset length.

[0149] The embodiment of the present application can also filter out the reverse lane line, boundary lane line, and intrusion lane line in the opposite direction of the target lane line through the parameters in the lane line parameter equation corresponding to the lane line mentioned in the above embodiment, and remove them. The embodiment of the present application improves the stability and accuracy of lane line tracking by removing unstable lane lines that do not meet current actual needs.

[0150] In some embodiments, the verified lane line to be tracked is saved so that in the next frame, the historical lane line can be obtained to obtain the set of lane lines to be tracked that are closest to the historical lane line, and the lane line to be tracked is saved based on a preset grid length. The grid length can be set so that when the lane line to be tracked is subsequently output, the historical lane line corresponding to the lane line to be tracked will be displayed; for example, the grid length can be set to -30 to 80 meters, that is, the lane lines displayed subsequently are within 30 meters behind the current vehicle position information and within 80 meters in front of the current vehicle position information.

[0151] Based on the same inventive concept, as an implementation of the above method, an embodiment of the present invention further provides a lane line detection and tracking device, which corresponds to the above method embodiment. For ease of reading, this embodiment will no longer repeat the details of the above method embodiment one by one, but it should be clear that the lane line detection and tracking device in this embodiment can correspond to all the contents in the above method embodiment.

[0152] The present application provides a lane line detection and tracking device. Figure 8 is a schematic diagram of the structure of the lane line detection and tracking device, as shown in Figure 8 As shown, the lane line detection and tracking device 800 includes:

[0153] The acquisition unit 801 is used to obtain a bird's-eye view image of the vehicle's surroundings at the current moment;

[0154] A detection unit 802 is used to detect the bird's-eye view image and obtain a visual point sequence corresponding to a plurality of lane lines in the bird's-eye view image;

[0155] A removing unit 803 is used to remove abnormal visual points from the visual points corresponding to the multiple lane lines based on the positional relationship of the visual points corresponding to the multiple lane lines, so as to generate an initial lane line set;

[0156] An acquisition unit 804 is used to acquire a lane line to be tracked from the initial lane line set;

[0157] The verification unit 805 is used to verify the lane line to be tracked, and after the lane line to be tracked passes the verification, merge the lane line to be tracked with the historical lane line and output the merged lane line.

[0158] As an optional implementation of the embodiment of the present application, the acquisition unit 804 is specifically used to acquire the lane line set to be tracked from the initial lane line set; the lane line set to be tracked includes a preset number of lane lines in the initial lane line set that are closest to the historical lane lines; the historical lane lines are lane lines that have been detected in the previous frame; and the lane lines to be tracked are acquired from the lane line set to be tracked.

[0159] As an optional implementation of the embodiment of the present application, the acquisition unit 804 is specifically used to obtain the target lane line parameters based on the lane line parameter equation corresponding to the historical lane line; for each lane line in the lane line set to be tracked, obtain the lane line parameters corresponding to the lane line; obtain the difference between the lane line parameters corresponding to each lane line in the lane line set to be tracked and the target lane line parameters, and take the lane line with the smallest difference as the lane line to be tracked.

[0160] As an optional implementation of an embodiment of the present application, the detection unit 802 is specifically used to input the bird's-eye view image into a lane line detection and tracking model to obtain a visual point sequence corresponding to multiple lane lines in the bird's-eye view image; wherein the lane line detection and tracking model is a model obtained by training a machine model based on sample data, and the sample data includes multiple sample bird's-eye view images and visual point sequences corresponding to multiple lane lines generated by each sample bird's-eye view image.

[0161] As an optional implementation of the embodiment of the present application, the elimination unit 803 is specifically used to obtain, for each lane line in the multiple lane lines, a circular radius consisting of all three adjacent visual points in the visual point sequence corresponding to the lane line; if the circular radius consisting of any three adjacent visual points is less than or equal to a preset value, the middle visual point of any three adjacent visual points is eliminated; and / or; for each lane line in the multiple lane lines, obtain the angle consisting of three adjacent visual points in the visual point sequence corresponding to the lane line; if the angle consisting of any three adjacent visual points is less than or equal to a preset angle, the middle visual point of any three adjacent visual points is eliminated.

[0162] As an optional implementation of an embodiment of the present application, the elimination unit 803 is also used to retain the last abnormal visual point for each lane line among the multiple lane lines when the visual points in the visual point sequence corresponding to the lane line are continuous abnormal visual points and the number is equal to a preset value.

[0163] As an optional implementation of the embodiment of the present application, the verification unit 805 is specifically used to obtain the distance between the lane line to be tracked and the historical lane line; if the distance is less than or equal to a threshold, the verification passes; if the distance is greater than the threshold, the verification fails.

[0164] As an optional implementation of the embodiment of the present application, the verification unit 805 is further used to increase the number of tracking times by one after the lane line to be tracked passes the verification; and increase the number of non-tracking times by one after the lane line to be tracked fails the verification.

[0165] As an optional implementation of the embodiment of the present application, the elimination unit 803 is also used to eliminate the lane line to be tracked when the tracking number of the lane line to be tracked is greater than or equal to a first preset number, and the position information is beyond a preset distance behind the position information of the current vehicle; when the tracking number of the lane line to be tracked is less than the first preset number, and the number of untracked times is greater than a second preset number, the lane line to be tracked is eliminated.

[0166] As an optional implementation of the embodiment of the present application, the lane line detection and tracking device also includes a storage unit: after the lane line to be tracked passes the verification, the lane line to be tracked is saved in a grid of a preset length.

[0167] Based on the same inventive concept, an embodiment of the present application also provides an electronic device. Fig. 9 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application, such as Fig. 9 As shown, the electronic device provided in this embodiment includes: a memory 901 and a processor 902, wherein the memory 901 is used to store a computer program; and the processor 902 is used to execute the lane line detection and tracking method provided in the above embodiment when executing the computer program.

[0168] Based on the same inventive concept, an embodiment of the present application also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the computing device implements the lane line detection and tracking method provided in the above embodiment.

[0169] Based on the same inventive concept, an embodiment of the present application further provides a vehicle, which includes the lane line detection and tracking device provided by the above embodiment or the electronic device provided by the above embodiment.

[0170] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media that include computer-usable program code.

[0171] The processor may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0172] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0173] Computer readable media include permanent and non-permanent, removable and non-removable storage media. Storage media can be implemented by any method or technology to store information, and the information can be computer-readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0174] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A lane line detection and tracking method, It is characterized in that include: Obtain a bird's-eye view image of the vehicle's surroundings at the current moment; Detecting the bird's-eye view image to obtain a visual point sequence corresponding to a plurality of lane lines in the bird's-eye view image; Based on the positional relationship of the visual points corresponding to the multiple lane lines, abnormal visual points among the visual points corresponding to the multiple lane lines are removed to generate an initial lane line set; After unifying the historical lane lines and the initial lane line set into the same coordinate system, based on the position information and / or line type relationship between the historical lane lines and each lane line in the initial lane line set, the lane line to be tracked is obtained from the initial lane line set; wherein the historical lane line is a lane line corresponding to the lane line to be tracked obtained at a historical moment; The lane line to be tracked is merged with the corresponding historical lane line to obtain the target lane line.

2. The method according to claim 1, It is characterized in that The acquiring the lane line to be tracked from the initial lane line set based on the position information and line type relationship between the historical lane line and each lane line in the initial lane line set comprises: Acquire a lane line set to be tracked from the initial lane line set; the lane line set to be tracked includes a preset number of lane lines in the initial lane line set that are closest to the historical lane lines; Based on the lane line parameter equation corresponding to the historical lane line, obtaining the target lane line parameter; For each lane line in the set of lane lines to be tracked, obtaining a lane line parameter corresponding to the lane line; The difference between the lane line parameters corresponding to each lane line in the lane line set to be tracked and the target lane line parameters is obtained, and the lane line with the smallest difference is taken as the lane line to be tracked.

3. The method according to claim 1, It is characterized in that The method of acquiring a lane line to be tracked from the initial lane line set based on the position information of the historical lane line and each lane line in the initial lane line set comprises: Obtaining position information of each lane line in the initial lane line set; Based on the position information of each lane line in the initial lane line set, the lane line closest to the historical lane line is obtained as the lane line to be tracked.

4. The method according to claim 1, It is characterized in that The acquiring the lane line to be tracked from the initial lane line set based on the line type relationship between the historical lane line and each lane line in the initial lane line set comprises: Based on the lane line parameter equation corresponding to the historical lane line, obtaining the target lane line parameter; For each lane line in the initial lane line set, obtaining a lane line parameter corresponding to the lane line; The difference between the lane line parameters corresponding to each lane line in the initial lane line set and the target lane line parameters is obtained, and the lane line with the smallest difference is used as the lane line to be tracked.

5. The method according to claim 1, It is characterized in that The removing abnormal visual points from the visual points corresponding to the multiple lane lines based on the positional relationship of the visual points corresponding to the multiple lane lines to generate an initial lane line set includes: For each visual point sequence corresponding to the multiple lane lines, obtaining the radius of a circle formed by all three adjacent visual points in the visual point sequence corresponding to the lane line; If the radius of a circle formed by any three adjacent visual points is less than or equal to a preset value, the intermediate visual point of any three adjacent visual points is removed; the intermediate visual point is determined based on the detection order of the three adjacent visual points; and / or; For each of the multiple lane lines, a visual point sequence corresponding to the lane line is used to obtain an angle formed by three adjacent visual points in the visual point sequence corresponding to the lane line. If the included angle formed by any three adjacent visual points is less than or equal to a preset angle, the intermediate visual point among any three adjacent visual points is eliminated; the intermediate visual point is determined based on the detection order of the three adjacent visual points.

6. The method according to claim 1, It is characterized in that The method of removing abnormal visual points from the visual points corresponding to the multiple lane lines based on the positional relationship of the visual points corresponding to the multiple lane lines to generate an initial lane line set comprises: For each lane line of the multiple lane lines, when the visual points in the visual point sequence corresponding to the lane line are continuous abnormal visual points and the number is equal to a preset value, the last abnormal visual point is retained.

7. The method according to claim 1, It is characterized in that Before fusing the lane line to be tracked with the corresponding historical lane line to obtain the target lane line, the method further includes: Obtaining the distance between the lane line to be tracked and the historical lane line; If the distance is less than or equal to the threshold, the verification is passed, and the lane line to be tracked is fused with the corresponding historical lane line to obtain the target lane line.

8. The method according to claim 7, It is characterized in that Before fusing and outputting the lane line to be tracked with the corresponding historical lane line, the method further includes: After the lane line to be tracked passes the verification, the tracking times are increased by one; After the lane line to be tracked fails to pass the verification, the number of times of not tracking is increased by one; When the number of tracking times of the lane line to be tracked is greater than or equal to the first preset number, and the position information is beyond the preset distance behind the position information of the current vehicle, the lane line to be tracked is removed; When the tracking times of the lane line to be tracked are less than the first preset times, and the non-tracking times are greater than the second preset times, the lane line to be tracked is removed.

9. A lane line detection and tracking device, It is characterized in that include: An acquisition unit, used to acquire a bird's-eye view image of the vehicle's surroundings at the current moment; A detection unit, configured to detect the bird's-eye view image and obtain a visual point sequence corresponding to a plurality of lane lines in the bird's-eye view image; A removing unit, configured to remove abnormal visual points from the visual points corresponding to the multiple lane lines based on the positional relationship of the visual points corresponding to the multiple lane lines, so as to generate an initial lane line set; A processing unit, configured to unify the historical lane lines and the initial lane line set into the same coordinate system, and then obtain the lane line to be tracked from the initial lane line set based on the position information and / or line type relationship between the historical lane lines and each lane line in the initial lane line set; wherein the historical lane line is a lane line corresponding to the lane line to be tracked obtained at a historical moment; A fusion unit is used to fuse the lane line to be tracked with the corresponding historical lane line to obtain a target lane line.

10. An electronic device, It is characterized in that include: A memory and a processor, wherein the memory is used to store a computer program; and the processor is used to enable the electronic device to implement the lane line detection and tracking method described in any one of claims 1-8 when executing the computer program.

11. A computer-readable storage medium, It is characterized in that The computer-readable storage medium stores a computer program, and when the computer program is executed by a computing device, the computing device implements the lane line detection and tracking method according to any one of claims 1 to 8.

12. A vehicle, It is characterized in that include: The lane line detection and tracking device as described in claim 9 or the electronic device as described in claim 10.

Images