Lane line recognition method, terminal device and computer readable storage medium

Abstract

The application is suitable for the field of radar technology, and provides a lane line identification method, a terminal device and a computer readable storage medium, which comprises the following steps: determining a target analysis point cloud line segment according to a current point cloud; segmenting the target analysis point cloud line segment according to a preset segmentation rule; and identifying whether the current point cloud is a lane line point cloud according to a point cloud distance characteristic value of the target analysis point cloud line segment, a point cloud distance characteristic value of each segment point cloud data, a point cloud reflectivity characteristic value of each segment point cloud data and a distance characteristic value of point cloud of an adjacent vertical channel of the target analysis line segment. The target analysis point cloud line segment is determined according to the current point cloud, each segment point cloud distance characteristic value and reflectivity characteristic value are determined through segmentation, and then whether the current point cloud is a lane line point cloud is identified based on the above, so that the identification precision of the lane line point cloud can be effectively improved.

Description

Technical Field

[0001] This application belongs to the field of radar technology, and in particular relates to a lane line recognition method, terminal equipment and computer-readable storage medium. Background Technology

[0002] Due to its advantages such as high resolution, high sensitivity, strong anti-interference ability, and unaffected by dark conditions, lidar is often used in fields such as autonomous driving, logistics vehicles, robots, and intelligent public transportation.

[0003] In the field of autonomous driving, when vehicles are driving on the road, they need to identify lane markings on the ground to distinguish lanes in order to plan the vehicle's route in advance. Because lane markings are often covered with a highly reflective material, they are currently identified by the reflectivity of point cloud data obtained from LiDAR scanning. However, when the road surface is used for a long time and is neglected, both the road surface and the lane markings will wear down, leading to inaccurate lane marking identification. Summary of the Invention

[0004] This application provides a lane line recognition method, a terminal device, and a computer-readable storage medium to solve the problem that when a laser is directly irradiated onto a high-reflectivity target object, the received signal is oversaturated and cannot effectively recover the true echo waveform, resulting in a large deviation in the measurement results.

[0005] In a first aspect, embodiments of this application provide a lane line recognition method, including: Determine the target point cloud line segment based on the current point cloud; The target analysis point cloud segments are segmented according to preset segmentation rules; Calculate the point cloud distance feature value of the target analysis point cloud line segment based on the distance value of the target analysis point cloud line segment, and calculate the point cloud distance feature value of each segment of point cloud data based on the distance value of each segment, and calculate the point cloud reflectance feature value of each segment of point cloud data based on the reflectance value of each segment. Calculate the distance feature values ​​of the point cloud of adjacent vertical channels to the target analysis line segment; The system identifies whether the current point cloud is a lane line point cloud based on the point cloud distance feature value of the target analysis point cloud line segment, the point cloud distance feature value of each point cloud data segment, the point cloud reflectance feature value of each point cloud data segment, and the distance feature value of the point cloud of the adjacent vertical channel to the target analysis line segment.

[0006] In one implementation of the first aspect, identifying whether the current point cloud is a lane line point cloud based on the point cloud distance feature value of the target analysis point cloud line segment, the point cloud distance feature value of each point cloud data segment, the point cloud reflectance feature value of each point cloud data segment, and the distance feature value of the point cloud of the adjacent vertical channel to the target analysis line segment includes: The identification conditions are set based on the point cloud distance feature value of the target analysis point cloud line segment, the point cloud distance feature value of each point cloud data segment, the point cloud reflectance feature value of each point cloud data segment, and the distance feature value of the point cloud of the adjacent vertical channel to the target analysis line segment. If the identification conditions are met, the current point cloud is determined to be a lane line point cloud; otherwise, the current point cloud is determined not to be a lane line point cloud.

[0007] In one implementation of the first aspect, the identification conditions include: The average distance between the target analysis point cloud line segments is greater than the first preset threshold; The ratio of the average reflectance of the point cloud in the second segment to the average reflectance of the point cloud in the first segment, and the ratio of the average reflectance of the point cloud in the second segment to the average reflectance of the point cloud in the third segment, are both greater than the second preset threshold. The reflectance variance of the first segment, the reflectance variance of the second segment, and the reflectance variance of the third segment are all less than the third preset threshold. The current vertical angle of the point cloud is negative; The absolute value of the difference between the average distance of the point cloud of adjacent vertical channels and the average distance of the line segment of the target analysis point cloud is greater than the fourth preset threshold.

[0008] In one implementation of the first aspect, after identifying whether the current point cloud is a lane line point cloud based on the point cloud distance feature value of the target analysis point cloud line segment, the point cloud distance feature value of each point cloud data segment, the point cloud reflectance feature value of each point cloud data segment, and the distance feature value of the point cloud of the adjacent vertical channel to the target analysis line segment, the method further includes: Enhance the reflectivity of lane line point clouds.

[0009] In one implementation of the first aspect, the enhancement of the reflectivity of the lane line point cloud includes: Enhancement coefficients are used to increase the reflectivity of lane line point clouds.

[0010] In one implementation of the first aspect, after enhancing the reflectivity of the lane line point cloud based on the enhancement coefficient, it also includes: If the reflectivity of the enhanced lane line point cloud exceeds the maximum reflectivity limit, then the reflectivity of the enhanced lane line point cloud will be set to the maximum reflectivity limit.

[0011] Secondly, embodiments of this application provide a terminal device, including: The target determination unit is used to determine the target analysis point cloud segment based on the current point cloud. The segmentation unit is used to segment the target analysis point cloud line segments according to preset segmentation rules; The first calculation unit is used to calculate the point cloud distance feature value of the target analysis point cloud line segment based on the distance value of the target analysis point cloud line segment, and to calculate the point cloud distance feature value of each segment of point cloud data based on the distance value of each segment, and to calculate the point cloud reflectance feature value of each segment of point cloud data based on the reflectance value of each segment. The second calculation unit is used to calculate the distance feature value of the point cloud of the adjacent vertical channel to the target analysis line segment; The identification unit is used to identify whether the current point cloud is a lane line point cloud based on the point cloud distance feature value of the target analysis point cloud line segment, the point cloud distance feature value of each point cloud data segment, the point cloud reflectance feature value of each point cloud data segment, and the distance feature value of the point cloud of the adjacent vertical channel to the target analysis line segment.

[0012] Thirdly, embodiments of this application provide a terminal device, the terminal device including a processor, a memory, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the lane line recognition method as described in the first aspect or any optional method of the first aspect.

[0013] Fourthly, embodiments of this application provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements the lane line recognition method as described in the first aspect or any alternative method of the first aspect.

[0014] Fifthly, embodiments of this application provide a computer program product that, when run on a terminal device, causes the terminal device to execute the lane line recognition method described in the first aspect or any optional method of the first aspect.

[0015] The beneficial effects of the embodiments in this application compared with the prior art are: Implementing the lane line recognition method, terminal device, computer-readable storage medium, and computer program product provided in this application has the following beneficial effects: The lane line recognition method provided in this application determines the target analysis point cloud segment through the current point cloud, then determines the distance feature value and reflectivity feature value of each segment of the point cloud, and then identifies whether the current point cloud is a lane line point cloud based on this, which can effectively improve the recognition accuracy of lane line point clouds. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a schematic diagram of the working scenario of a lidar. Figure 2 This is a schematic flowchart illustrating a lane line recognition method provided in an embodiment of this application; Figure 3 This is a schematic diagram illustrating an application scenario of the lane line recognition method provided in the embodiments of this application; Figure 4 This is a schematic diagram of the angle of the lane line relative to the center of the lidar in an embodiment of this application; Figure 5 This is a segmentation diagram provided in an embodiment of the present application for segmenting the target analysis point cloud into line segments; Figure 6 This is a schematic diagram illustrating the implementation process of another lane line recognition method provided in this application embodiment; Figure 7 This is a schematic diagram of the structure of a terminal device provided in an embodiment of this application; Figure 8 This is a schematic diagram of the structure of a terminal device provided in another embodiment of this application; Figure 9 This is a schematic diagram of the structure of a computer-readable storage medium provided in an embodiment of this application. Detailed Implementation

[0018] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0019] It should be understood that the term "and / or" as used in this application specification and the appended claims refers to any combination of one or more of the associated listed items, as well as all possible combinations, and includes such combinations. Furthermore, in the description of this application specification and the appended claims, the terms "first," "second," "third," etc., are used only for distinguishing descriptions and should not be construed as indicating or implying relative importance.

[0020] It should also be understood that references to "one embodiment" or "some embodiments" in this specification mean that one or more embodiments of this application include the specific features, structures, or characteristics described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0021] LiDAR (Light Detection and Ranging) is an automatic remote sensing device that uses a laser as its emission source and employs photoelectric detection technology for detection. A LiDAR system can include a transmitter, a receiver, a scanning control system, and a data processing system. The working principle of LiDAR is as follows: the transmitter emits a detection laser towards the target object. After the laser hits the target object, it is reflected back, forming an echo. The receiver receives this echo and processes it to obtain information such as the target object's distance, size, speed, and reflectivity. Different electro-optic sensors in a LiDAR system emit laser beams at different vertical angles in the air. The echo signals returned after the laser beams detect an object are received by the photoelectric sensors and converted into electrical signals. After analog-to-digital conversion, distance measurement is achieved. The scanning control system controls the LiDAR's motor to rotate to the next horizontal angle. The entire transmitter and receiver repeat this process, obtaining distance and reflectivity values ​​at different horizontal angles from the same vertical angle. This generates three-dimensional distance information of the target in space, facilitating further sensing and processing to determine the target's shape, size, and type.

[0022] For example, please see Figure 1 , Figure 1 A block diagram of a lidar is shown, such as... Figure 1As shown, the aforementioned lidar may include a control and processing unit, a transmitter, a receiver, a transmitting lens, and a receiving lens. The control and processing unit controls the lidar to operate according to a certain transmission and reception sequence, and processes the received data to obtain the target distance result. The transmitter is typically an array of multiple semiconductor lasers, which emit lasers according to a certain timing sequence under the drive of the control and processing unit. When a target is present in the direction of emission, the target reflects the laser and returns an echo. The returned echo passes through the receiving lens to the receiver, which may specifically be a receiving photoelectric sensor. The receiving photoelectric sensor converts the received light signal (i.e., the echo) into an electrical signal, amplifies it, and performs analog-to-digital conversion to obtain a digital signal corresponding to the electrical signal. After subsequent digital processing, the target distance result and reflectivity result are obtained.

[0023] In the field of autonomous driving, when vehicles are driving on roads, they need to identify lane markings to distinguish lanes in order to plan their routes in advance. Currently, lane markings are typically identified using the reflectivity of point cloud data obtained from LiDAR scanning. Reflectivity represents the optical reflectivity of an object, with values ​​ranging from 0 to 255. A lower value indicates lower reflectivity, and a higher value indicates higher reflectivity. To improve lane marking accuracy, lane markings are usually set to white or yellow and covered with a high-reflectivity material, resulting in a reflectivity between 30 and 60, while the road surface is typically black with a reflectivity of around 10. However, over time, both the road surface and lane markings wear down, reducing the difference in reflectivity. For example, the reflectivity of a worn-out road surface might be around 8, while the reflectivity of worn-out lane markings might be between 13 and 15. In such cases, the difference between the reflectivity of lane markings and the road surface in the LiDAR point cloud scan becomes very small, leading to inaccurate lane marking identification.

[0024] The lane line recognition method provided in the embodiments of this application will be described in detail below: Please see Figure 2 , Figure 2 This is a schematic flowchart illustrating a lane line recognition method provided in this application embodiment. It should be noted that the executing entity of the lane line recognition method provided in this application embodiment can be a LiDAR, specifically the control and processing unit inside the LiDAR, or a terminal device communicatively connected to the LiDAR. The aforementioned terminal device can be a mobile terminal such as a smartphone, tablet, or wearable device, or it can be a computer, cloud server, radar-assisted computer, or other devices in various application scenarios. It should be noted that the following explanation uses a LiDAR as the executing entity as an example: like Figure 2As shown, the lane line recognition method provided in this application embodiment may include S11~S14, which are detailed below: S11: Determine the target point cloud segment based on the current point cloud.

[0025] Please see Figure 3 , Figure 3 A schematic diagram illustrating an application scenario of the lane line recognition method provided in this application embodiment is shown. For example... Figure 3 As shown, a lidar for acquiring point cloud data can be installed on the top of the vehicle. The lidar emits lasers toward the road surface, and after reflection from the road surface and lane lines, the lidar can acquire point cloud data, which includes road surface point clouds and lane line point clouds.

[0026] In this embodiment, point cloud data reflected from the road surface and lane lines is first acquired using a lidar system. This point cloud data includes distance and reflectivity values. Assume the point cloud data pairs containing distance and reflectivity values ​​are Dis(m,n) and Ref(m,n).

[0027] Where Dis(m,n) represents the distance value, Ref(m,n) represents the reflectivity, m represents the point cloud index at different vertical angles, the corresponding vertical angle is AngV(m), and n represents the point cloud index at different horizontal angles, the corresponding horizontal angle is AngH(n).

[0028] It should be noted that the aforementioned vertical and horizontal angles are determined during the LiDAR scanning process. When scanning the road surface, the LiDAR can be controlled to scan at a certain vertical angle. At that vertical angle, the LiDAR can then be controlled to rotate horizontally. As an example, the LiDAR can rotate 360° horizontally. Therefore, it can be controlled to rotate 5° at a time (this angle value can be set according to the required acquisition accuracy; it is only an example and not a limitation). This allows the distance and reflectivity values ​​at different horizontal angles from the same vertical angle to be obtained. By adjusting the vertical angle and repeating the above operation, the distance and reflectivity values ​​at different vertical and horizontal angles can be obtained.

[0029] In this embodiment of the application, by taking a segment of point cloud data on the left and right sides of the current point cloud as the center, the target analysis point cloud line segment can be obtained.

[0030] In practical applications, assuming the distance value of the current point cloud data is Dis(m,n) and the reflectance value of the current point cloud data is Ref(m,n), then the distance value of the target analysis point cloud line segment can be represented as: Dis(m,nk:n+k), and the reflectance value of the target analysis point cloud line segment can be represented as: Ref(m,nk:n+k). That is, the distance value and reflectance value are each taken from 2k+1 points.

[0031] In one embodiment of this application, S11 may include the following steps: The number of lane line point clouds is determined by obtaining the width of the lane line and the angle of the lane line relative to the center of the lidar. It is understandable that the lane width value can be obtained from standard lane width data, or it can be obtained by extracting lane feature point data from the detection data to obtain possible lane width values. As an optional method, the center angle of the lane line relative to the lidar can be determined based on the distance information of lane feature points on the left and right sides of the radar in the detection data.

[0032] The target analysis point cloud segment is determined based on the current point cloud, the number of lane line point clouds, and the preset multiplier.

[0033] In practical applications, the value of k can be determined based on the width of the lane line, a preset multiple, and the angle of the lane line relative to the center of the lidar.

[0034] Please see Figure 4 , Figure 4 A schematic diagram illustrating the angle of the lane line relative to the center of the lidar in an embodiment of this application is shown. Figure 4 As shown, assuming the width between the two lane lines is Llane, and the distance between the lane line and the center of the LiDAR is Dis(m,n), the angle the lane line is relative to the center of the LiDAR can be expressed as: thetalane=2*atan(1 / 2Llane / Dis(m,n)).

[0035] Based on this, the number of point clouds for the lane lines can be determined, denoted as PointN: PointN = thetalane / AngR; Where AngR is the horizontal resolution of the lidar, which can be set to 0.2.

[0036] To ensure coverage of lane lines, a certain margin is allowed in the target analysis point cloud segments. Therefore, the theoretical number of lane line point clouds is multiplied by a preset system factor, and then rounded down, i.e.: PointN=round(Mr*thetalane / AngR); Mr is a preset coefficient, which can be 1.2.

[0037] For ease of calculation, the number of lane line point clouds can also be set to an odd number pointing upwards, i.e.: PointN=2*(ceil(PointN+1) / 2)-1.

[0038] The number of point cloud segments in the target analysis point cloud is a preset multiple of the number of point cloud segments in the lane lines, that is: PointA = N * PointN; Where N is a preset multiple, specifically 3, meaning the number of point cloud segments in the target analysis point cloud is 3 times the number of point cloud segments in the lane lines.

[0039] Based on this, the value of K can be determined as: K = (PointA-1) / 2. That is, for point clouds with the same vertical angle m, taking the distance and reflectivity values ​​of the current point cloud as the center, take K points forward horizontally, and then take K points backward horizontally. It should be noted that taking K points forward means taking K points obtained before the lidar acquired the current point cloud, and taking K points backward means taking K points obtained after the lidar acquired the current point cloud.

[0040] S12: Divide the target analysis point cloud line segments according to the preset segmentation rules.

[0041] In this embodiment, the aforementioned preset segmentation rule can be determined based on the application scenario. For example, the target analysis point cloud line segments can be divided into N segments; where N is a preset multiple in S11. Of course, the aforementioned preset segmentation rule can also be to divide the target analysis point cloud line segments into N segments according to different point cloud number requirements, etc., and is not limited here.

[0042] Please see Figure 5 , Figure 5 This illustration shows a segmentation diagram of dividing a target analysis point cloud into segments, according to an embodiment of this application. For example... Figure 5 As shown, the example is to divide the target analysis point cloud into 3 segments according to the preset segmentation rule: After obtaining the target analysis point cloud line segments, the distance value of the target analysis point cloud line segments is denoted as DisG(u) = Dis(m,nk:n+k), and the reflectance value is denoted as RefG(u) = Ref(m,nk:n+k). Among them, the value of u ranges from 0 to 3PointN-1.

[0043] The distance value is divided into three segments, denoted as follows: First segment: DisG00(0:PointN-1) = DisG(0:PointN-1); Second section: DisG01(0:PointN-1) = DisG(PointN:2*PointN-1); Third segment: DisG02(0:PointN-1) = DisG(2*PointN:3*PointN-1).

[0044] The reflectance value is divided into three segments, denoted as follows: First segment: RefG00(0:PointN-1) = RefG(0:PointN-1); Second section: RefG01(0:PointN-1) = RefG(PointN:2*PointN-1); Third segment: RefG02(0:PointN-1) = RefG(2*PointN:3*PointN-1).

[0045] S13: Calculate the point cloud distance feature value of the target analysis point cloud line segment based on the distance value of the target analysis point cloud line segment, and calculate the point cloud distance feature value of each segment of point cloud data based on the distance value of each segment, and calculate the point cloud reflectance feature value of each segment of point cloud data based on the reflectance value of each segment.

[0046] In this embodiment of the application, the point cloud distance feature value of the target analysis point cloud line segment is first calculated based on the distance value of the target analysis point cloud line segment. The point cloud distance feature value of the target analysis point cloud line segment includes the mean distance and the variance of the target analysis point cloud line segment.

[0047] The mean distance of the point cloud line segments in the target analysis is: M ; The variance of the distance between the target analysis point cloud line segments is: Sigma .

[0048] In the embodiments of this application, using Figure 5 The example shown illustrates dividing the target analysis point cloud into three segments. The point cloud distance feature value of the first segment is calculated based on the distance value of the first segment; the point cloud distance feature value of the second segment is calculated based on the distance value of the second segment; and the point cloud distance feature value of the third segment is calculated based on the distance value of the third segment. These point cloud distance feature values ​​also include the mean distance and the variance of the distance.

[0049] The mean distance of the first segment of the point cloud is: M ; The distance variance of the first point cloud segment is: Sigma .

[0050] Similarly, the mean distance M of the second segment of the point cloud was calculated. The distance variance Sigma of the second point cloud segment The mean distance M of the third segment of the point cloud The distance variance Sigma of the second point cloud segment .

[0051] In this embodiment, the point cloud reflectance characteristic value of the first segment is calculated based on the reflectance value of the first segment, the point cloud reflectance characteristic value of the second segment is calculated based on the reflectance value of the second segment, and the point cloud reflectance characteristic value of the third segment is calculated based on the reflectance value of the third segment. The aforementioned point cloud reflectance characteristic value includes the mean reflectance and the variance of the reflectance.

[0052] The mean reflectance of the first segment of the point cloud is: M ; The variance of reflectance of the first point cloud segment is: SigmaRef .

[0053] Similarly, the mean reflectance M of the second segment of the point cloud was calculated. The variance of reflectance of the second point cloud (Sigma) The mean reflectance M of the third segment of the point cloud The variance of reflectance of the second point cloud (Sigma) .

[0054] S14: Calculate the distance feature value of the point cloud of the adjacent vertical channel to the target analysis line segment.

[0055] In this embodiment, the point cloud of the adjacent vertical channel to the target analysis line segment refers to the point cloud data of adjacent vertical channels with different vertical angles but the same horizontal angle, including the point cloud of the previous vertical channel and the point cloud of the next vertical channel. The distance feature value mentioned above may include the average distance.

[0056] Specifically, the average distance between point clouds in adjacent vertical channels is calculated as follows: M ; M .

[0057] S15: Identify whether the current point cloud is a lane line point cloud based on the point cloud distance feature value of the target analysis point cloud line segment, the point cloud distance feature value of each point cloud data segment, the point cloud reflectance feature value of each point cloud data segment, and the distance feature value of the point cloud of the adjacent vertical channel to the target analysis line segment.

[0058] In practical applications, recognition conditions are set based on the point cloud distance feature value of the target analysis point cloud line segment, the point cloud distance feature value of each point cloud data segment, the point cloud reflectance feature value of each point cloud data segment, and the distance feature value of the point cloud of the adjacent vertical channel to the target analysis line segment. If the recognition conditions are met, the current point cloud is determined to be a lane line point cloud; otherwise, the current point cloud is determined not to be a lane line point cloud.

[0059] In practical applications, the recognition conditions set above based on the point cloud distance feature values ​​of the target analysis point cloud line segments, the point cloud distance feature values ​​of each point cloud data segment, the point cloud reflectance feature values ​​of each point cloud data segment, and the distance feature values ​​of the point clouds in adjacent vertical channels to the target analysis line segments can include the following conditions: Condition 1: The average distance between line segments in the target analysis point cloud is greater than a first preset threshold. The first preset threshold can be set based on the actual scenario and is not limited here. For example, the first preset threshold can be set to 5 centimeters.

[0060] Condition 2: The ratio of the average reflectance of the point cloud in the second segment to the average reflectance of the point cloud in the first segment, and the ratio of the average reflectance of the point cloud in the second segment to the average reflectance of the point cloud in the third segment, are both greater than a second preset threshold. This second preset threshold can also be set based on the actual scenario and is not limited here. For example, the second preset threshold can be set to any value between 1.3 and 1.5.

[0061] Condition 3: The reflectance variances of the first segment, the second segment, and the third segment are all less than a third preset threshold. This third preset threshold can also be set based on the actual scenario and is not limited here. For example, the third preset threshold can be set to 5.

[0062] Condition 4: The vertical angle of the current point cloud is negative.

[0063] Condition 5: The absolute value of the difference between the average distance of the point clouds of adjacent vertical channels and the average distance of the line segments of the target analysis point cloud is greater than the fourth preset threshold. The third preset threshold mentioned above can also be set based on the actual scenario, and is not limited here. For example, the third preset threshold can be set to 20 centimeters.

[0064] When the above 5 conditions are met, the current point cloud can be determined to be the point cloud on the lane line (i.e., the lane line point cloud).

[0065] As can be seen from the above, the lane line recognition method provided in this application determines the target analysis point cloud segment through the current point cloud, then determines the distance feature value and reflectivity feature value of each segment of the point cloud, and then identifies whether the current point cloud is a lane line point cloud based on this, which can effectively improve the recognition accuracy of lane line point clouds.

[0066] Please see Figure 6 , Figure 6 A schematic diagram illustrating the implementation flow of a lane line recognition method according to another embodiment of this application is shown. Figure 6 As shown, unlike the previous embodiment, the lane line recognition method provided in this application embodiment further includes the following steps: S16: Enhance the reflectivity of lane line point clouds.

[0067] In this embodiment of the application, in order to facilitate the identification of lane line point clouds by applications such as autonomous driving, the reflectivity of the current point cloud can be enhanced after it is determined that the current point cloud is a lane line point cloud.

[0068] Specifically, the reflectivity of lane line point clouds can be enhanced based on enhancement coefficients. That is: RefGEnh(K) = RefG(K) * KRat; Where RefGEnh(K) is the reflectivity of the enhanced lane line point cloud, RefG(K) is the reflectivity of the lane line point cloud obtained from the received point cloud, and KRat is the enhancement coefficient.

[0069] In this embodiment, the enhancement coefficient can be determined based on the ratio of the average reflectance of the point cloud in the second segment to the average reflectance of the point cloud in the first segment, and the ratio of the average reflectance of the point cloud in the second segment to the average reflectance of the point cloud in the third segment. Specifically, it can be the maximum, minimum, or average of the ratios of the second and third segment reflectance values. KRat = max(Ratio0100, Ratio0102), or; KRat = min(Ratio0100, Ratio0102), or; KRat=1 / 2(Ratio0100+Ratio0102).

[0070] In this embodiment of the application, to prevent the reflectivity of the enhanced point cloud from exceeding the maximum reflectivity limit, the reflectivity of the enhanced point cloud can be limited based on the maximum reflectivity limit. That is, if the reflectivity of the enhanced lane line point cloud exceeds the maximum reflectivity limit, the reflectivity of the enhanced lane line point cloud is set to the maximum reflectivity limit.

[0071] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0072] Based on the lane line recognition method provided in the above embodiments, the present invention further provides embodiments of terminal devices that implement the above method embodiments.

[0073] Please see Figure 7 , Figure 7This is a schematic diagram of the structure of a terminal device provided in an embodiment of this application. In this embodiment, the terminal device includes units used for performing... Figure 2 The steps in the corresponding embodiments. Please refer to the details. Figure 2 as well as Figure 2 The relevant descriptions in the corresponding embodiments are shown below. For ease of explanation, only the parts relevant to this embodiment are shown. Figure 7 As shown, the terminal device 70 includes: a target determination unit 71, a segmentation unit 72, a first calculation unit 73, a second calculation unit 74, and an identification unit 75. Wherein: The target determination unit 71 is used to determine the target analysis point cloud segment based on the current point cloud.

[0074] Segmentation unit 72 is used to segment the target analysis point cloud line segments according to preset segmentation rules.

[0075] The first calculation unit 73 is used to calculate the point cloud distance feature value of the target analysis point cloud line segment based on the distance value of the target analysis point cloud line segment, and to calculate the point cloud distance feature value of each segment of point cloud data based on the distance value of each segment, and to calculate the point cloud reflectance feature value of each segment of point cloud data based on the reflectance value of each segment.

[0076] The second calculation unit 74 is used to calculate the distance feature value of the point cloud of the adjacent vertical channel to the target analysis line segment.

[0077] The identification unit 75 is used to identify whether the current point cloud is a lane line point cloud based on the point cloud distance feature value of the target analysis point cloud line segment, the point cloud distance feature value of each point cloud data segment, the point cloud reflectance feature value of each point cloud data segment, and the distance feature value of the point cloud of the adjacent vertical channel to the target analysis line segment.

[0078] In one implementation of this application embodiment, the target determination unit 71 may include a number determination unit and a line segment determination unit, wherein: The number determination unit is used to obtain the width between two lane lines and the angle of the lane line relative to the center of the lidar to determine the number of lane line point clouds; The line segment determination unit is used to determine the target analysis point cloud line segments based on the current point cloud, the number of lane line point clouds, and a preset multiplier.

[0079] In one implementation of this application, the identification unit includes a condition setting unit and a determination unit.

[0080] The condition setting unit is used to set recognition conditions based on the point cloud distance feature value of the target analysis point cloud line segment, the point cloud distance feature value of each point cloud data segment, the point cloud reflectance feature value of each point cloud data segment, and the distance feature value of the point cloud of the adjacent vertical channel to the target analysis line segment. The determination unit is used to determine that the current point cloud is a lane line point cloud if the recognition conditions are met, otherwise it determines that the current point cloud is not a lane line point cloud.

[0081] In one implementation of this application, the terminal device further includes an enhancement unit.

[0082] The enhancement unit is used to enhance the reflectivity of lane line point clouds.

[0083] Specifically, it enhances the reflectivity of lane line point clouds based on enhancement coefficients.

[0084] It should be noted that the information interaction and execution process between the above-mentioned units are based on the same concept as the method embodiments of this application. Their specific functions and technical effects can be referred to the method embodiments section, and will not be repeated here.

[0085] Figure 8 This is a schematic diagram of the structure of a terminal device provided in another embodiment of this application. For example... Figure 8 As shown, the terminal device 8 provided in this embodiment includes: a processor 80, a memory 81, and a computer program 82 stored in the memory 81 and executable on the processor 80, such as an image segmentation program. When the processor 80 executes the computer program 82, it implements the steps in the various lane line recognition method embodiments described above, for example... Figure 2 S11~S15 are shown. Alternatively, when the processor 80 executes the computer program 82, it implements the functions of each module / unit in the above-described terminal device embodiments, for example... Figure 7 The functions of units 71-75 shown.

[0086] For example, the computer program 82 can be divided into one or more modules / units, which are stored in the memory 81 and executed by the processor 80 to complete this application. The one or more modules / units can be a series of computer program instruction segments capable of performing specific functions, which describe the execution process of the computer program 82 in the terminal device 8. For example, the computer program 82 can be divided into various units; please refer to the specific functions of each unit. Figure 7 The relevant descriptions in the corresponding embodiments are not repeated here.

[0087] The terminal device may include, but is not limited to, a processor 80 and a memory 81. Those skilled in the art will understand that... Figure 8 This is merely an example of terminal device 8 and does not constitute a limitation on terminal device 8. It may include more or fewer components than shown, or combine certain components, or different components. For example, the terminal device may also include input / output devices, network access devices, buses, etc.

[0088] The processor 80 may be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), 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.

[0089] The memory 81 can be an internal storage unit of the terminal device 8, such as a hard disk or memory of the terminal device 8. The memory 81 can also be an external storage device of the terminal device 8, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped on the terminal device 8. Furthermore, the memory 81 can include both internal and external storage units of the terminal device 8. The memory 81 is used to store the computer program and other programs and data required by the terminal device. The memory 81 can also be used to temporarily store data that has been output or will be output.

[0090] This application also provides a computer-readable storage medium. See also... Figure 9 , Figure 9 This is a schematic diagram of the structure of a computer-readable storage medium provided in an embodiment of this application, such as... Figure 9 As shown, a computer-readable storage medium 90 stores a computer program 91, which, when executed by a processor, can implement the lane line recognition method described above.

[0091] This application provides a computer program product that, when run on a terminal device, enables the terminal device to implement the lane line recognition method described above.

[0092] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the terminal device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this application. The specific working process of the units and modules in the above system can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0093] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, refer to the relevant descriptions of other embodiments.

[0094] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0095] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A lane line recognition method, characterized in that, include: Taking the current point cloud as the center, take a segment of point cloud data on the left and right sides at a horizontal angle to obtain the target analysis point cloud line segment; The target analysis point cloud segments are segmented according to preset segmentation rules; Calculate the point cloud distance feature value of the target analysis point cloud line segment based on the distance value of the target analysis point cloud line segment, and calculate the point cloud distance feature value of each segment of point cloud data based on the distance value of each segment, and calculate the point cloud reflectance feature value of each segment of point cloud data based on the reflectance value of each segment. Calculate the distance feature values ​​of the point clouds of adjacent vertical channels to the target analysis line segment; wherein, the point clouds of adjacent vertical channels refer to the point clouds of adjacent vertical channels with different vertical angles but the same horizontal angle; the distance feature values ​​include the average distance. The current point cloud is identified as a lane line point cloud based on the point cloud distance feature value of the target analysis point cloud line segment, the point cloud distance feature value of each point cloud data segment, the point cloud reflectance feature value of each point cloud data segment, and the distance feature value of the point cloud of the adjacent vertical channel to the target analysis line segment.

2. The lane line recognition method according to claim 1, characterized in that, The step of identifying whether the current point cloud is a lane line point cloud based on the point cloud distance feature value of the target analysis point cloud line segment, the point cloud distance feature value of each point cloud data segment, the point cloud reflectance feature value of each point cloud data segment, and the distance feature value of the point cloud of the adjacent vertical channel to the target analysis line segment includes: The identification conditions are set based on the point cloud distance feature value of the target analysis point cloud line segment, the point cloud distance feature value of each point cloud data segment, the point cloud reflectance feature value of each point cloud data segment, and the distance feature value of the point cloud of the adjacent vertical channel to the target analysis line segment. If the identification conditions are met, the current point cloud is determined to be a lane line point cloud; otherwise, the current point cloud is determined not to be a lane line point cloud.

3. The lane line recognition method according to claim 2, characterized in that, The identification conditions include: The average distance between the target analysis point cloud line segments is greater than the first preset threshold; The ratio of the average reflectance of the point cloud in the second segment to the average reflectance of the point cloud in the first segment, and the ratio of the average reflectance of the point cloud in the second segment to the average reflectance of the point cloud in the third segment, are both greater than the second preset threshold. The reflectance variance of the first segment, the reflectance variance of the second segment, and the reflectance variance of the third segment are all less than the third preset threshold. The current vertical angle of the point cloud is negative; The absolute value of the difference between the average distance of the point cloud of adjacent vertical channels and the average distance of the line segment of the target analysis point cloud is greater than the fourth preset threshold.

4. The lane line recognition method according to claim 1, characterized in that, After identifying whether the current point cloud is a lane line point cloud based on the point cloud distance feature values ​​of the target analysis point cloud segment, the point cloud distance feature values ​​of each point cloud data segment, the point cloud reflectance feature values ​​of each point cloud data segment, and the distance feature values ​​of the point clouds of adjacent vertical channels to the target analysis point segment, the process also includes: Enhance the reflectivity of lane line point clouds.

5. The lane line recognition method according to claim 4, characterized in that, The enhanced reflectivity of the lane line point cloud includes: Enhancement coefficients are used to increase the reflectivity of lane line point clouds.

6. The lane line recognition method according to claim 5, characterized in that, After enhancing the reflectivity of the lane line point cloud based on the enhancement factor, it also includes: If the reflectivity of the enhanced lane line point cloud exceeds the maximum reflectivity limit, then the reflectivity of the enhanced lane line point cloud will be set to the maximum reflectivity limit.

7. A terminal device, characterized in that, include: The target determination unit is used to take a segment of point cloud data on the left and right sides of the current point cloud as the center, and obtain the target analysis point cloud line segment; The segmentation unit is used to segment the target analysis point cloud line segments according to preset segmentation rules; The first calculation unit is used to calculate the point cloud distance feature value of the target analysis point cloud line segment based on the distance value of the target analysis point cloud line segment, and to calculate the point cloud distance feature value of each segment of point cloud data based on the distance value of each segment, and to calculate the point cloud reflectance feature value of each segment of point cloud data based on the reflectance value of each segment. The second calculation unit is used to calculate the distance feature values ​​of the point clouds of adjacent vertical channels to the target analysis line segment; wherein, the point clouds of adjacent vertical channels refer to the point clouds of adjacent vertical channels with different vertical angles but the same horizontal angle; the distance feature values ​​include the average distance. The identification unit is used to identify whether the current point cloud is a lane line point cloud based on the point cloud distance feature value of the target analysis point cloud line segment, the point cloud distance feature value of each point cloud data segment, the point cloud reflectance feature value of each point cloud data segment, and the distance feature value of the point cloud of the adjacent vertical channel of the target analysis line segment.

8. A terminal device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer-readable instructions, it implements the lane line recognition method as described in any one of claims 1 to 6.

9. A computer-readable storage medium storing a computer program, characterized in that, When the computer-readable instructions are executed by a processor, they implement the lane line recognition method as described in any one of claims 1 to 6.

Images