Vehicle positioning method, device, computer equipment and storage medium

Abstract

A vehicle positioning method includes: obtaining a lane line map, a first spatial position of a vehicle in a driving area at a previous moment, and a scene image of the driving area collected at a current moment; the scene image of the driving area includes a scene lane line segment; determining a second spatial position of the scene lane line segment in the driving area by performing parallax detection on the scene image of the driving area; determining a plurality of position prediction points within a preset distance of the first spatial position, and screening out standard lane line segments corresponding to each position prediction point from the lane line map; determining the distance difference between each standard lane line segment and the scene lane line segment, and determining the probability value of each position prediction point belonging to the vehicle position at the current moment according to the distance difference; setting a third spatial position of a position prediction point in the driving area with a probability value higher than a preset threshold as the spatial position of the vehicle at the current moment. The use of this method can reduce the cost of unmanned driving.

Description

Technical Field

[0001] The present application relates to a vehicle positioning method, device, computer equipment and storage medium. Background Art

[0002] With the development of science and technology, vehicle autonomous driving technology based on GNSS (Global Navigation Satellite System) has become a research hotspot. GNSS is a commonly used global positioning method, which can provide vehicle positioning results without cumulative errors. However, in cities with tall buildings or under bridges and other places where the signal is unstable, the positioning accuracy of GNSS will be greatly affected. At this time, the GNSS positioning results cannot be used for the positioning of autonomous vehicles.

[0003] When GNSS signals are unstable, such as in tunnels, self-driving vehicles usually use a high-precision IMU (Inertial measurement unit) in combination with lidar for accurate positioning. However, high-precision IMUs are relatively expensive and difficult to meet the needs of large-scale production. Summary of the invention

[0004] According to various embodiments disclosed in the present application, a vehicle positioning method is provided.

[0005] A vehicle positioning method includes:

[0006] Acquire a lane line map, a first spatial position of the vehicle in the driving area at a previous moment, and a scene image of the driving area collected at a current moment; the scene image of the driving area includes a scene lane line segment;

[0007] Determine a second spatial position of the on-site lane line segment in the driving area by performing parallax detection on the on-site image of the driving area;

[0008] Determine a plurality of position prediction points within a preset distance of the first spatial position, and select standard lane line segments corresponding to each position prediction point from the lane line map;

[0009] Determine the distance difference between each standard lane line segment and the on-site lane line segment, and determine the probability value of each position prediction point belonging to the vehicle position at the current moment based on the distance difference;

[0010] A third spatial position of a position prediction point having a probability value higher than a preset threshold in the driving area is set as the spatial position of the vehicle at the current moment.

[0011] A vehicle positioning device comprises:

[0012] A live lane line determination module is used to obtain a lane line map, a first spatial position of a vehicle in a driving area at a previous moment, and a live image of the driving area collected at a current moment; the live image of the driving area includes live lane line segments; and determine a second spatial position of the live lane line segments in the driving area by performing parallax detection on the live image of the driving area;

[0013] a standard lane line determination module, used to determine a plurality of position prediction points located within a preset distance from the first spatial position, and to select standard lane line segments corresponding to each position prediction point from the lane line map;

[0014] The positioning module is used to determine the distance difference between each standard lane line segment and the on-site lane line segment, and determine the probability value of each position prediction point belonging to the vehicle's position at the current moment based on the distance difference; a position prediction point with a probability value higher than a preset threshold value in the driving area is set as the third spatial position of the vehicle at the current moment.

[0015] A computer device includes a memory and one or more processors, wherein the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processors, the one or more processors perform the following steps:

[0016] Acquire a lane line map, a first spatial position of the vehicle in the driving area at a previous moment, and a scene image of the driving area collected at a current moment; the scene image of the driving area includes a scene lane line segment;

[0017] Determine a second spatial position of the on-site lane line segment in the driving area by performing parallax detection on the on-site image of the driving area;

[0018] Determine a plurality of position prediction points within a preset distance of the first spatial position, and select standard lane line segments corresponding to each position prediction point from the lane line map;

[0019] Determine the distance difference between each standard lane line segment and the on-site lane line segment, and determine the probability value of each position prediction point belonging to the vehicle position at the current moment based on the distance difference;

[0020] A third spatial position of a position prediction point having a probability value higher than a preset threshold in the driving area is set as the spatial position of the vehicle at the current moment.

[0021] One or more computer storage media storing computer readable instructions, which, when executed by one or more processors, cause the one or more processors to perform the following steps:

[0022] Acquire a lane line map, a first spatial position of the vehicle in the driving area at a previous moment, and a scene image of the driving area collected at a current moment; the scene image of the driving area includes a scene lane line segment;

[0023] Determine a second spatial position of the on-site lane line segment in the driving area by performing parallax detection on the on-site image of the driving area;

[0024] Determine a plurality of position prediction points within a preset distance of the first spatial position, and select standard lane line segments corresponding to each position prediction point from the lane line map;

[0025] Determine the distance difference between each standard lane line segment and the on-site lane line segment, and determine the probability value of each position prediction point belonging to the vehicle position at the current moment based on the distance difference;

[0026] A third spatial position of a position prediction point having a probability value higher than a preset threshold in the driving area is set as the spatial position of the vehicle at the current moment.

[0027] The details of one or more embodiments of the present application are set forth in the following drawings and description. Other features and advantages of the present application will become apparent from the description, drawings, and claims. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0029] Figure 1 A diagram showing an application scenario of a vehicle positioning method according to one or more embodiments;

[0030] Figure 2 is a flow chart of a vehicle positioning method according to one or more embodiments;

[0031] Figure 3 is a schematic diagram of determining a distance difference according to one or more embodiments;

[0032] Figure 4 A schematic diagram of a process of generating a lane map according to one or more embodiments;

[0033] Figure 5 A block diagram of a vehicle positioning device according to one or more embodiments;

[0034] Figure 6 is a block diagram of a vehicle positioning device in another embodiment;

[0035] Figure 7 is a block diagram of a computer device according to one or more embodiments. DETAILED DESCRIPTION

[0036] In order to make the technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0037] The vehicle positioning method provided in this application can be applied to Figure 1 In the application environment shown. Among them, the unmanned vehicle 102 has a built-in vehicle positioning module for vehicle positioning, or the vehicle positioning module is integrated into the server 104, so that the vehicle positioning method can be completed in the terminal 102, or in the server 104, or through the cooperation of the terminal 102 and the server 104. Among them, the terminal 102 and the server 104 communicate through the network. The terminal 102 can be, but is not limited to, various personal computers, laptops, smart phones, tablet computers and portable wearable devices, and the server 104 can be implemented by an independent server or a server cluster composed of multiple servers.

[0038] In one embodiment, Figure 2 As shown, a vehicle positioning method is provided, which is applied to Figure 1 The server in the example is used to illustrate the following steps:

[0039] S202, obtaining a lane line map, a first spatial position of a vehicle in a driving area at a previous moment, and a scene image of the driving area collected at a current moment; the scene image of the driving area includes scene lane line segments.

[0040] Among them, the lane line map refers to map data including the spatial position information of the lane line. By querying the lane line map, the current autonomous driving vehicle can know its own spatial position information. Before obtaining the lane line map, the map sampling vehicle can collect field data in the driving area through laser radar and image acquisition equipment, so as to generate a lane line map based on the field data. The driving area refers to the area where the vehicle is driving. A driving area can be divided into multiple segmented areas, so that the server can generate a lane line map corresponding to each segmented area, and combine the lane line maps corresponding to each segmented area to obtain a lane line map corresponding to the driving area. The driving area scene image is a scene image collected by the image acquisition device of the vehicle driving in the driving area; the scene lane line segment is the lane line segment displayed in the driving area scene image.

[0041] Specifically, when the vehicle just enters the driving area, for example, when it just enters the tunnel entrance, in order to locate the vehicle, the autonomous driving vehicle will obtain the lane line map corresponding to the current driving area, and based on the image acquisition device mounted on the vehicle, collect the scene image of the driving area at the current moment. Among them, the driving area can be a virtual geographical boundary enclosed by a virtual fence in a target field area such as a tunnel or a road surface. For example, the driving area can be a virtual geographical boundary enclosed by the entrance and exit of the tunnel. When the autonomous driving vehicle just enters the driving area, for example, when it just arrives at the tunnel entrance, the autonomous driving vehicle directly uses the spatial position of the tunnel entrance as its own spatial position; when the automatically set vehicle drives in the driving area for a period of time, the autonomous driving vehicle obtains its first spatial position in the driving area at the previous moment, and determines the spatial position at the current moment based on the first spatial position. For example, when the autonomous driving vehicle just arrives at the tunnel entrance at the previous moment, the first spatial position is the spatial position of the tunnel entrance. Among them, the interval between the previous moment and the current moment can be freely set according to demand. Among them, the spatial position in the driving area is the position coordinate of the object in the world coordinate system established with the driving area as the reference. This world coordinate system is a three-dimensional coordinate system established with the position of the map sampling vehicle at the entrance of the driving area as the origin. For example, this world coordinate system can be a three-dimensional coordinate system established with the center point of the image acquisition device mounted on the map sampling vehicle as the origin.

[0042] In one embodiment, since there is still a GNSS signal at the entrance of the driving area, the server can determine the current driving area based on the GNSS signal at the entrance of the driving area, and obtain a lane line map corresponding to the current driving area.

[0043] In one embodiment, the server may obtain a complete lane line map composed of multiple driving areas, and filter out a lane line map corresponding to the current driving area from the complete lane line map according to the current driving area.

[0044] S204, determining a second spatial position of a lane line segment in the driving area by performing parallax detection on the scene image of the driving area.

[0045] Specifically, the image acquisition device mounted on the autonomous driving vehicle can obtain multiple images of the driving area, so that the server can perform parallax detection on the multiple images of the driving area, and determine the position information of the lane line segment relative to the autonomous driving vehicle based on the parallax detection results, and obtain the second spatial position of the lane line segment in the driving area based on the position information of the lane line segment relative to the autonomous driving vehicle and the internal and external parameters of the image acquisition device. Among them, the position information of the lane line segment relative to the autonomous driving vehicle is the position information of the lane line segment in the coordinate system established with the midpoint of the image acquisition device as the origin. The internal parameters of the image acquisition device can specifically be the internal parameter matrix of the camera.

[0046] In one embodiment, a scene image of a driving area includes a first image and a second image; by performing parallax detection on the scene image of the driving area, a second spatial position of a scene lane line segment in the driving area is determined, including: performing pixel matching on the first image and the second image, and determining the parallax between the first image and the second image based on the matching result; determining the depth information of the driving area based on the parallax; extracting the scene lane line information from the scene image of the driving area; determining the second spatial position of the scene lane line segment in the driving area based on the scene lane line information and the depth information of the driving area.

[0047] Specifically, the image acquisition device mounted on the autonomous driving vehicle may be a binocular camera; the first image and the second image may be a left image and a right image, respectively, wherein the left image is an image frame acquired by the left camera of the binocular camera; and the right image is an image frame acquired by the right camera of the binocular camera. When it is necessary to identify the lane line segments in the driving area, the first image and the second image may be acquired through the binocular camera.

[0048] The server obtains the first image and traverses each pixel in the first image. The server obtains the first pixel in the current traversal order in the first image and determines the pixel coordinates of the first pixel in the current traversal order in the first image. The server determines the second candidate pixel in the second image that has the same pixel coordinates as the first pixel in the current traversal order, and uses the second candidate pixel as a reference to search for multiple second candidate pixels within a preset distance in the second image according to a preset pixel step. For example, when the pixel coordinates of the first pixel in the current traversal order are (2,3), the server uses (2,3) as a reference to search for multiple second candidate pixels within a 1-pixel step in eight directions, including up, down, left, right, upper left, upper right, lower left, and lower right. Among them, the pixel coordinates are the coordinate values ​​of the pixel in the pixel coordinate system. The pixel coordinate system is a coordinate system established with the upper left corner of the image as the origin.

[0049] The server determines the grayscale value of the first pixel point in the current traversal order and the grayscale values ​​of multiple second candidate pixels, and determines the grayscale difference between each second candidate pixel point and the first pixel point in the current traversal order based on the grayscale value of the first pixel point in the current traversal order and the grayscale values ​​of each second candidate pixel point. The server selects a second target pixel point with the smallest grayscale difference from the multiple second candidate pixels, and uses the second target pixel point as a pixel point that matches the current traversal order.

[0050] In one embodiment, when there are multiple second target pixels with the smallest grayscale difference in the second image, the server selects multiple pixels in the first image that are located near the first pixel in the current traversal order, and forms an area through the first pixel in the current traversal order and multiple pixels located near the first pixel in the current traversal order. In the matching process, the matching of pixel points is converted into matching of pixel areas and pixel areas, so that the server can still determine the depth of the on-site lane line segment relative to the automated driving vehicle at the current moment based on the matched area.

[0051] Since the first image and the second image are acquired based on binocular camera acquisition, it is only necessary to search for the second pixel point that matches the first pixel point in the current traversal order within a preset range, which greatly reduces the matching efficiency of pixel point matching.

[0052] S206, determining a plurality of position prediction points within a preset distance of the first spatial position, and selecting standard lane line segments corresponding to the respective position prediction points from the lane line map.

[0053] Specifically, due to the limitations of vehicle dynamics, the vehicle will not produce particularly large movements between two moments. Therefore, based on the movement limitations of the vehicle and the position of the vehicle at the previous moment, a rough guess of the vehicle's position at the current moment can be obtained. The server uses the first spatial position as the origin to determine multiple position prediction points within a preset distance. For example, the server uses the first spatial position as the center of the circle, the preset distance as the radius, and a preset step length to determine multiple position prediction points located within the circle. Since the preset step length and the first spatial position are known, the server can determine the third spatial position of each position prediction point in the driving area based on the preset step length and the first spatial position.

[0054] Furthermore, the server traverses each position prediction point, and based on the third spatial position of the position prediction point in the current traversal order, selects the standard lane line segment closest to the third spatial position of the position prediction point in the current traversal order from the lane line map, and sets the standard lane line segment closest to the third spatial position of the position prediction point in the current traversal order as the standard lane line segment corresponding to the position prediction point in the current traversal order. The length of the standard lane line segment can be freely set according to demand, for example, the length of the standard lane line segment can be set to 5 meters.

[0055] In one embodiment, the standard lane lines in the lane line map may be pre-segmented to obtain a plurality of standard lane line segments. The server determines the spatial position information of each lane point in at least one standard lane line segment, and determines the distance difference between each lane point and the position prediction point of the current traversal order based on the spatial point position information and the third spatial position information of the position prediction point of the current traversal order, and determines the standard lane line segment closest to the position prediction point of the current traversal order based on the distance difference.

[0056] S208, determining the distance difference between each standard lane line segment and the on-site lane line segment, and determining the probability value of each position prediction point belonging to the vehicle position at the current moment based on the distance difference.

[0057] S210: A third spatial position of a position prediction point having a probability value higher than a preset threshold within the driving area is set as the spatial position of the vehicle at the current moment.

[0058] Specifically, when the standard lane line segment corresponding to each position prediction point is obtained, the server determines the distance difference between each standard lane line segment and the on-site lane line segment, and determines the probability value of each position prediction point belonging to the vehicle's position at the current moment based on the distance difference. For example, the distance difference is divided by the length of the on-site lane line segment to obtain the corresponding probability value. When the probability value of each position prediction point belonging to the vehicle's position at the current moment is obtained, the server uses a position prediction point with a probability value higher than a preset threshold as the target position prediction point, and obtains the third spatial position of the target position prediction point in the driving area, and sets the third spatial position of the target position prediction point as the spatial position of the vehicle at the current moment.

[0059] In one embodiment, the distance difference between each standard lane line segment and the on-site lane line segment is determined, and based on the distance difference, the probability value of each position prediction point belonging to the vehicle's position at the current moment is determined, including: traversing the standard lane line segments; determining the distance difference between the standard lane line segments in the current traversal order and the on-site lane line segments based on the position information of the target parts in the standard lane line segments in the current traversal order and the position information of the target parts in the on-site lane line segments; determining the probability value of the position prediction point corresponding to the standard lane line segments in the current traversal order belonging to the vehicle's position at the current moment based on the distance difference.

[0060] Specifically, the server determines the position information of the target part in the on-site lane line segment in the driving area according to the second spatial position of the on-site lane line segment in the driving area. The server traverses each standard lane line segment corresponding to the position prediction point and determines the standard lane line segment of the current traversal order. The server determines the position information of the target part of the standard lane line segment of the current traversal order in the driving area according to the second spatial position of the standard lane line segment of the current traversal order in the driving area. The server subtracts the position information of the target part in the on-site lane line segment in the driving area from the position information of the target part in the on-site lane line segment of the current traversal order in the driving area to obtain a position difference, and uses the absolute value of the position difference as the distance difference between the standard lane line segment of the current traversal order and the on-site lane line segment. Further, the server determines the probability value of the position prediction point corresponding to the standard lane line segment of the current traversal order belonging to the vehicle position at the current moment according to the distance difference. Among them, the target part can be the middle of the lane line segment, that is, the middle position of the lane line segment.

[0061] In one embodiment, the server may segment the on-site lane line segments and the standard lane line segments, and determine the distance difference between the standard lane line and the on-site lane line based on the position information of the target part of the segmented segments.

[0062] For example, reference Figure 3 , when the on-site lane line segment is the left on-site lane line segment Ls on the left and the right on-site lane line segment Rs on the right, and the standard lane line segments in the current traversal order are the left standard lane line segment Lc on the left and the right standard lane line segment Rc on the right, the server divides Ls, Rs, Lc and Rc equally according to the lengths of the segments to obtain Ls1, Ls2…Lsn; Rs1, Rs2…Rsn; Lc1, Lc2…Lcn, Rc1, Rc2…Rcn. The server obtains the position information of the middle positions of Ls1, Ls2…Lsn; Rs1, Rs2…Rsn; Lc1, Lc2…Lcn, Rc1, Rc2…Rcn respectively, and determines the left average distance between the left on-site lane line segment Ls and the left standard lane line segment Lc on the left, and the right average distance between the right on-site lane line segment Rs and the right standard lane line segment Rc on the right based on the position information of the middle position, and determines the distance difference between the on-site lane line segment and the standard lane line segment in the current traversal order according to the left average distance and the right average distance. Figure 3 Schematic diagram of distance difference determination in one embodiment. In one embodiment, the server determines the distance difference in the first segmented area based on the position information of the middle position of Ls1 and the corresponding position information of the middle position of Lc1, determines the distance difference in the second segmented area based on the position information of the middle position of Ls2 and the corresponding position information of the middle position of Lc2, and so on, to obtain the distance difference between each segmented area, thereby determining the distance difference between the on-site lane line segment and the standard lane line segment in the current traversal order according to the distance difference between each segmented area.

[0063] By segmenting the lane line segments, the distance difference determined based on the segmented lane line segments is made more accurate, thereby improving the accuracy of vehicle positioning.

[0064] In one embodiment, the server determines the first spatial position of the vehicle in the driving area at the previous moment, and can estimate the position prediction point of the vehicle at the current moment based on the first spatial position, so that the current standard lane line segment closest to the current position prediction point can be filtered out from the lane line map; by obtaining the standard lane line segment closest to the position prediction point, the distance between multiple standard lane line segments and the on-site lane line segment can be determined, and the standard lane line segment closest to the target standard lane line segment is used as the target standard lane line segment, the position prediction point closest to the target standard lane line segment is used as the target position prediction point, and the third spatial position of the target position prediction point in the driving area is used as the spatial position of the vehicle at the current moment.

[0065] In the above vehicle positioning method, by acquiring the on-site image of the driving area collected at the current moment, the spatial position information of the on-site lane line segment in the driving area can be determined according to the on-site image of the driving area; by determining the first spatial position of the vehicle in the driving area at the previous moment, multiple position prediction points of the vehicle at the current moment can be estimated according to the first spatial position; by determining multiple position prediction points and acquiring the lane line map, the standard lane line segments corresponding to each position prediction point can be screened out from the lane line map, so that the probability value of each position prediction point belonging to the position of the vehicle at the current moment can be determined according to the distance difference between each standard lane line segment and the on-site lane line segment, and then the spatial position of the vehicle at the current moment can be determined according to the probability value. Since in the process of missing GNSS signals, the position information at the current moment can be determined only according to the lane line map and the on-site image of the driving area collected by the image acquisition device, compared with the traditional positioning that requires IMU, this embodiment can greatly reduce the production cost and meet large-scale production use.

[0066] In one embodiment, a scene image of a driving area includes a first image and a second image; by performing parallax detection on the scene image of the driving area, a second spatial position of a scene lane line segment in the driving area is determined, including: performing pixel matching on the first image and the second image, and determining the parallax between the first image and the second image based on the matching result; determining the depth information of the driving area based on the parallax; extracting the scene lane line information from the scene image of the driving area; determining the second spatial position of the scene lane line segment in the driving area based on the scene lane line information and the depth information of the driving area.

[0067] Specifically, the server performs pixel matching on the first pixel in the first image and the second pixel in the second image to determine the second pixel that matches each first pixel. The server traverses each first pixel and determines the first pixel coordinates of the first pixel in the current traversal order in the first image and the second pixel coordinates of the matching second pixel in the second image, and determines the depth information of the object point in the driving area corresponding to the first pixel in the current traversal order from the vehicle at the current moment based on the first pixel coordinates and the second pixel coordinates.

[0068] For example, the server uses the optical center of the image acquisition device mounted on the autonomous vehicle at the current moment as the origin to establish the camera coordinate system, and uses the vehicle's forward direction as the Z axis, the direction perpendicular to the image acquisition device as the Y axis, and the left-to-right direction as the X axis. The server determines the pixel coordinate difference disparity based on the first pixel coordinate of the first pixel point in the current traversal order and the corresponding second pixel coordinate, and uses the formula The Z value of the object point corresponding to the first pixel in the camera coordinate system is obtained, that is, the depth information of the first pixel is obtained.

[0069] The server integrates the depth information of each first pixel point to obtain the depth information of the driving area. Among them, Baseline is the horizontal difference between the optical centers of the left and right cameras; FocalLength is the focal length of the camera.

[0070] Furthermore, the server extracts the semantic information of the on-site lane line segment by a semantic segmentation method, determines the position area of ​​the on-site lane line segment in the on-site image of the driving area according to the semantic information, and determines the on-site lane line information according to the position area. The on-site lane line information is information that characterizes the lane line in the driving area, for example, the on-site lane line information can be the pixel coordinates of the on-site lane line segment in the on-site image of the driving area, or the edge of the on-site lane line segment in the on-site image of the driving area. Furthermore, the server determines the second spatial position of the on-site lane line segment in the driving area according to the on-site lane line information and the depth information of the driving area.

[0071] In one of the embodiments, edge information of the on-site lane line segment may be extracted by an edge extraction method, and the pixel coordinates of the on-site lane line segment in the on-site image of the driving area may be determined based on the edge information.

[0072] In the above embodiment, the depth information of the driving area is predicted by a binocular camera, so that the second spatial position of the on-site lane line segment in the driving area can be determined based on the depth information. Compared with the traditional IMU-based spatial position determination, this embodiment can greatly reduce the cost required for autonomous driving.

[0073] In one embodiment, the on-site lane line information includes the pixel coordinates of the on-site lane line segments in the on-site image of the driving area; extracting the on-site lane line information in the on-site image of the driving area includes: extracting the on-site lane line features in the on-site image of the driving area through a pre-trained lane line recognition model; determining a lane line probability map based on the on-site lane line features; determining the pixel coordinates of the on-site lane line segments in the on-site image of the driving area based on the lane line probability map.

[0074] Specifically, when obtaining lane line segments, the lane line features can be first extracted from the scene image of the driving area through the pre-trained lane line recognition model to obtain the scene lane line features, and then a lane line probability map is generated based on the obtained scene lane line features. Finally, the scene lane line segments are extracted from the non-zero pixel points of the lane line probability map, and the pixel coordinates of the scene lane line in the scene image of the driving area are determined. The lane line probability map is used to characterize the probability that the pixel points in the scene image of the driving area belong to the scene lane line. Among them, the lane line recognition model is a machine learning model used to identify the lane lines in the image.

[0075] In one embodiment, after obtaining the on-site lane line features, the Hough transform algorithm may be used on the on-site lane line features to extract straight line segments on the on-site image of the driving area as lane line segments.

[0076] In the above embodiment, since the lane line recognition model is pre-trained, the lane lines in the scene image of the driving area can be accurately identified by the lane line recognition model, thereby improving the accuracy of vehicle positioning.

[0077] In one embodiment, the second spatial position of the on-site lane line segment in the driving area is determined based on the on-site lane line information and the depth information of the driving area, including: filtering out the depth information of the on-site lane line segment from the depth information of the driving area based on the pixel coordinates of the on-site lane line segment in the on-site image of the driving area; determining the second spatial position of the on-site lane line segment based on the pixel coordinates of the on-site lane line segment in the on-site image of the driving area and the depth information of the on-site lane line segment.

[0078] Specifically, when the depth information of each pixel in the scene image of the driving area is obtained, the server filters out the depth information of the scene lane line segment from the depth information of each pixel in the scene image of the driving area according to the pixel coordinates of the scene lane line segment in the scene image of the driving area. The server converts the pixel points located in the pixel coordinate system to the image coordinate system according to the pixel coordinates of each pixel in the scene image of the driving area, and converts the pixel points located in the image coordinate system to the camera coordinate system according to the depth information of the scene lane line segment, thereby obtaining the position information of the object point corresponding to the pixel point in the camera coordinate system. Among them, the image coordinate system is a coordinate system established with the midpoint of the image as the origin. For example, when determining the pixel coordinates of the scene lane line segment in the first image and the depth information Z value of each first pixel, the server combines the pixel coordinates and the depth information to obtain the coordinates of each first pixel constituting the scene lane line segment in the camera coordinate system.

[0079] Furthermore, the server determines the angle of the autonomous driving vehicle or image acquisition device in the world coordinate system, and the displacement of the autonomous driving vehicle or image acquisition device in the world coordinate system, and based on the formula pw=rwc*pc+twc, converts each pixel point in the camera coordinate system to the world coordinate system, thereby obtaining the second spatial position of the lane line segment in the driving area. Among them, pw is the point in the world coordinate system; rwc is the angle of the autonomous driving vehicle or image acquisition device in the world coordinate system; pc is the point in the camera coordinate system; twc is the displacement of the autonomous driving vehicle or image acquisition device in the world coordinate system.

[0080] In one of the embodiments, the extrinsic parameter matrix of the image acquisition device at the current moment can be obtained based on PnP (pespective-n-point), Homograpy Matrix, and other algorithms, so as to determine rwc and twc according to the extrinsic parameter matrix.

[0081] In this embodiment, by determining the second spatial position of the on-site lane line segment within the driving area, the position information of the autonomous driving vehicle can be subsequently determined based on the second spatial position.

[0082] In one embodiment, before obtaining the lane line map, the process further includes generating a lane line map, and the step of generating the lane line map includes:

[0083] S402: Segment the driving area to obtain at least one segmented area.

[0084] S404, acquiring a third image and a laser point cloud collected for the current segmented area.

[0085] S406, filtering out lane line point clouds projected onto the standard lane line segments in the third image from the laser point cloud.

[0086] S408: Determine a fourth spatial position of a standard lane line segment in the current segment area within the driving area according to the position information of the lane line point cloud in the current segment area.

[0087] S410, obtaining a lane map by combining the fourth spatial position of each standard lane line segment in the driving area.

[0088] Specifically, in order to better determine the location information of the lane lines in the driving area, the driving area can be segmented to obtain at least one segmented area. For example, when the length of the driving area is 1000 meters, the driving area can be divided into 1000 segmented areas according to a step length of 1 meter. The server obtains the third image and the laser point cloud collected for the current segmented area; wherein the third image is collected by an image acquisition device mounted on a map acquisition vehicle, and the third image contains lane line segments. It is easy to understand that the server can generate a corresponding lane line map for the lane lines in the driving area in advance, and the lane lines at this time are standard lane line segments; when the unmanned vehicle needs to be positioned based on the lane line map, the scene image of the driving area can be captured based on the binocular camera, and the lane line segments in the scene image of the driving area are called scene lane line segments. wherein the laser point cloud is collected by a laser radar mounted on a map acquisition vehicle. The server identifies the standard lane line points in the third image, and projects the laser point cloud collected at the same time for the current segmented area into the third image to obtain a radar projection map. Based on the radar projection map, the server selects the lane line point cloud projected onto the standard lane line segment from the laser point cloud. The server determines the fourth spatial position of each point constituting the standard lane line segment in the driving area by combining the pixel coordinates of the standard lane line segment in the third image and the depth information of the lane point cloud. When the server obtains the fourth spatial position of the standard lane line segment corresponding to each segmented area, the server combines the fourth spatial position of each standard lane line segment to obtain a lane line map.

[0089] In this embodiment, the lane line map can be obtained based on the third image and the laser point cloud, which is convenient and quick. By pre-generating the lane line map, positioning can be performed based on the pre-generated lane line map later.

[0090] In one embodiment, the third image is based on an image frame captured by an image acquisition device for the current segmented area; the laser point cloud is based on point cloud data captured by a laser radar for the current segmented area; the lane line point cloud projected onto the standard lane line segment in the third image is screened out from the laser point cloud, including: determining the laser coordinate system of the laser radar and the camera coordinate system of the image acquisition device; converting the laser point cloud in the laser coordinate system and the third image in the camera coordinate system into the same coordinate system; projecting the laser point cloud onto the third image in the same coordinate system to obtain a radar projection image; and determining the lane line point cloud projected onto the standard lane line segment based on the radar projection image.

[0091] Among them, the laser coordinate system is a coordinate system established with the center of the laser radar installed on the map collection vehicle as the origin; the camera coordinate system is a coordinate system established with the center of the image acquisition device installed on the map collection vehicle as the origin.

[0092] Specifically, the server can predetermine the relative rotation matrix and relative displacement vector between the laser coordinate system and the camera coordinate system, and transform the laser point cloud in the laser coordinate system and the third image in the camera coordinate system into the same coordinate system according to the relative rotation matrix and the relative displacement vector, for example, transform the laser point cloud into the camera coordinate system. The server projects the laser point cloud in the same coordinate system into the third image to obtain a radar projection map. The server identifies the standard lane line segments in the radar projection map, and based on the identified standard lane line segments, filters out the lane line point clouds projected onto the standard lane line segments from the laser point cloud.

[0093] In this embodiment, by converting the depth point cloud image acquired based on the laser coordinate system and the third image acquired based on the camera coordinate system into the same coordinate system, each point in the laser point cloud can accurately find the corresponding image pixel point in the third image, thereby achieving precise projection.

[0094] In one embodiment, a lane line point cloud includes multiple lane line points; based on the position information of the lane line point cloud in the current segmented area, determining the fourth spatial position of the standard lane line segment in the current segmented area in the driving area, including: based on the position information of the lane line point cloud in the current segmented area, determining the position information of each lane line point in the driving area; based on the position information of each lane line point in the driving area, performing curve fitting on multiple lane line points to obtain a lane line function corresponding to the current segmented area; based on the lane line function, determining the fourth spatial position of the standard lane line segment in the current segmented area in the driving area.

[0095] Specifically, the server determines the spatial coordinates of each lane line point constituting the standard lane line segment in the radar projection image in the camera coordinate system in combination with the pixel coordinates of each pixel point in the third image and the coordinates of each point in the laser point cloud in the camera coordinate system. The server determines the rotation angle and displacement of the map collection vehicle relative to the entrance position of the driving area at the current moment, and according to the rotation angle and displacement, converts the spatial coordinates of each point constituting the standard lane line segment in the camera coordinate system to the world coordinate system determined by the entrance position of the driving area, thereby obtaining the fourth spatial position of each lane line point constituting the standard lane line segment in the driving area. The server performs curve fitting on the standard lane line segment based on the fourth spatial position of each lane line point constituting the standard lane line segment in the driving area, obtains the lane line function corresponding to the current segmented area, and uses the lane line function segment obtained by fitting as the standard lane line segment, thereby obtaining the fourth spatial position of the standard lane line segment in the current segmented area in the driving area.

[0096] For example, since the lane line segment is a curve or a straight line in the segmented area, the lane line equation of the lane line segment can be pre-assumed. The lane line equation in the world coordinate system in the driving area can be expressed as x=x0+ax+bx 2 ,y=y0+cy+dy 2 ,z=z0+ez+fz 2 . Among them, a, b, c, d, e, and f are all lane line equation coefficients; x0, y0, and z0 are all constants. For the lane lines in the image, we can use the formula The laser radar point cloud is projected onto the pixel plane. Among them, u, v are the pixel coordinates of the laser point in the laser point cloud projected onto the third image, and x, y, z are the coordinates of the laser point in the laser point cloud in the camera coordinate system. The server selects the point cloud portion close to the lane line position observed in the image to obtain the lane line point cloud. The server combines the lane line point cloud and the lane line equation, and solves the lane line equation coefficient under the least squares angle through SVD to obtain the lane line equation, that is, the lane line function.

[0097] In one of the embodiments, a GNSS is also installed on the map collection vehicle, and the server can determine the rotation angle and displacement of the map collection vehicle relative to the entrance of the driving area at the current moment based on the GNSS signal, thereby determining the rotation angle and displacement of the image collection device relative to the entrance of the driving area based on the rotation angle and displacement of the map collection vehicle relative to the entrance of the driving area, and based on the rotation angle and displacement of the image collection device relative to the entrance of the driving area, convert the point in the camera coordinate system into the world coordinate system.

[0098] In the above embodiment, by performing curve fitting on each lane line point, a corresponding lane line function can be obtained, so that a lane line map can be obtained by only storing the lane line function in each segmented area.

[0099] It should be understood that although Figure 2 , 4 The steps in the flowchart are shown in sequence as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Moreover, Figure 2 , 4 At least part of the steps may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least part of the sub-steps or stages of other steps.

[0100] In one embodiment, Figure 5 As shown, a vehicle positioning device 500 is provided, comprising: a live lane line determination module 502, a standard lane line determination module 504 and a positioning module 506, wherein:

[0101] The on-site lane line determination module 502 is used to obtain a lane line map, the first spatial position of the vehicle in the driving area at the previous moment, and a scene image of the driving area collected at the current moment; the scene image of the driving area includes on-site lane line segments; by performing parallax detection on the scene image of the driving area, the second spatial position of the on-site lane line segments in the driving area is determined.

[0102] The standard lane line determination module 504 is used to determine a plurality of position prediction points within a preset distance from the first spatial position, and select the standard lane line segments corresponding to each position prediction point from the lane line map.

[0103] The positioning module 506 is used to determine the distance difference between each standard lane line segment and the on-site lane line segment, and determine the probability value of each position prediction point belonging to the vehicle's position at the current moment based on the distance difference; and set the third spatial position of a position prediction point with a probability value higher than a preset threshold in the driving area as the spatial position of the vehicle at the current moment.

[0104] In one embodiment, Figure 6 As shown, the scene image of the driving area includes a first image and a second image; the scene lane line determination module 502 also includes a disparity module 5021, which is used to perform pixel matching on the first image and the second image, and determine the disparity between the first image and the second image based on the matching result; determine the depth information of the driving area based on the disparity; extract the scene lane line information in the scene image of the driving area; determine the second spatial position of the scene lane line segment in the driving area based on the scene lane line information and the depth information of the driving area.

[0105] In one embodiment, the disparity module 5021 is also used to traverse at least one first pixel point in the first image; determine at least one second candidate pixel point located in the second image according to the pixel coordinates of the first pixel point in the current traversal order in the first image; determine the grayscale difference between each second candidate pixel point and the first pixel point in the current traversal order according to the grayscale value of the first pixel point in the current traversal order and the grayscale values ​​of each second candidate pixel point; and set the second candidate pixel point with the smallest grayscale difference as the second pixel point that matches the first pixel point in the current traversal order.

[0106] In one embodiment, the on-site lane line information includes the pixel coordinates of the on-site lane line segments in the on-site image of the driving area; the disparity module 5021 is also used to extract the on-site lane line features in the on-site image of the driving area through a pre-trained lane line recognition model; generate a lane line probability map based on the on-site lane line features; and determine the pixel coordinates of the on-site lane line segments in the on-site image of the driving area based on the lane line probability map.

[0107] In one embodiment, the disparity module 5021 is also used to filter out the depth information of the live lane line segment from the depth information of the driving area based on the pixel coordinates of the live lane line segment in the live image of the driving area; determine the second spatial position of the live lane line segment based on the pixel coordinates of the live lane line segment in the live image of the driving area and the depth information of the live lane line segment.

[0108] In one embodiment, the positioning module 506 is also used to traverse the standard lane line segments; determine the distance difference between the standard lane line segments in the current traversal order and the on-site lane line segments based on the position information of the target parts in the standard lane line segments in the current traversal order and the position information of the target parts in the on-site lane line segments; and determine the probability value of the position prediction point corresponding to the standard lane line segments in the current traversal order belonging to the vehicle's position at the current moment based on the distance difference.

[0109] In one embodiment, the vehicle positioning device 500 also includes a map generation module 508, which is used to segment the driving area to obtain at least one segmented area; obtain a third image and a laser point cloud collected for the current segmented area; filter out the lane line point cloud on the standard lane line segment projected on the third image from the laser point cloud; determine the fourth spatial position of the standard lane line segment in the current segmented area in the driving area based on the position information of the lane line point cloud in the current segmented area; and obtain a lane line map in combination with the fourth spatial position of each standard lane line segment in the driving area.

[0110] In one embodiment, the map generation module 508 also includes a projection module 5081, which is used to determine the laser coordinate system of the laser radar and the camera coordinate system of the image acquisition device; convert the laser point cloud located in the laser coordinate system and the third image located in the camera coordinate system into the same coordinate system; project the laser point cloud onto the third image in the same coordinate system to obtain a radar projection image; and determine the lane line point cloud projected onto the standard lane line segment based on the radar projection image.

[0111] In one embodiment, the map generation module 508 is also used to determine the position information of each lane line point in the driving area based on the position information of the lane line point cloud in the current segmented area; perform curve fitting on multiple lane line points based on the position information of each lane line point in the driving area to obtain a lane line function corresponding to the current segmented area; and determine the fourth spatial position of the standard lane line segment in the current segmented area within the driving area based on the lane line function.

[0112] The specific definition of the vehicle positioning device can be found in the definition of the vehicle positioning method above, which will not be repeated here. Each module in the above-mentioned vehicle positioning device can be implemented in whole or in part by software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0113] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 7 As shown. The computer device includes a processor, a memory, a network interface and a database connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, computer-readable instructions and a database. The internal memory provides an environment for the operation of the operating system and computer-readable instructions in the non-volatile storage medium. The database of the computer device is used to store vehicle positioning data. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer-readable instructions are executed by the processor, a vehicle positioning method is implemented.

[0114] Those skilled in the art will understand that the structure shown in FIG. Y is merely a block diagram of a partial structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different arrangement of components.

[0115] In one embodiment, a computer device is further provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above method embodiments when executing the computer program.

[0116] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0117] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing related hardware through computer-readable instructions, and the computer-readable instructions can be stored in a non-volatile computer-readable storage medium. When the computer-readable instructions are executed, they may include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0118] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0119] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. A vehicle positioning method, characterized in that: include: Acquire a lane line map, a first spatial position of the vehicle in the driving area at a previous moment, and a scene image of the driving area collected at a current moment; the scene image of the driving area includes a scene lane line segment; Determine a second spatial position of the on-site lane line segment in the driving area by performing parallax detection on the on-site image of the driving area, and determine position information of a target part in the on-site lane line segment in the driving area according to the second spatial position; Determine a plurality of position prediction points within a preset distance of the first spatial position, and select standard lane line segments corresponding to each position prediction point from the lane line map; Traversing the standard lane line segments; Subtract the position information of the target part in the standard lane line segment in the current traversal order from the position information of the target part in the on-site lane line segment in the driving area to obtain a position difference; The absolute value of the position difference is used as the distance difference between the standard lane line segment and the on-site lane line segment in the current traversal order; Determine, based on the distance difference, a probability value that the position prediction point corresponding to the standard lane line segment in the current traversal order belongs to the position of the vehicle at the current moment; A third spatial position of a position prediction point having a probability value higher than a preset threshold in the driving area is set as the spatial position of the vehicle at the current moment.

2. The method according to claim 1, characterized in that The driving area scene image includes a first image and a second image; and determining the second spatial position of the scene lane line segment in the driving area by performing parallax detection on the driving area scene image includes: Performing pixel matching on the first image and the second image, and determining a disparity between the first image and the second image according to a matching result; Determining depth information of the driving area according to the parallax; Extracting on-site lane line information from the on-site image of the driving area; According to the on-site lane line information and the depth information of the driving area, a second spatial position of the on-site lane line segment in the driving area is determined.

3. The method according to claim 2, characterized in that The performing pixel matching on the first image and the second image comprises: Traversing at least one first pixel point in the first image; Determine at least one second candidate pixel point located in the second image according to the pixel coordinates of the first pixel point in the current traversal order in the first image; Determine, according to the grayscale value of the first pixel in the current traversal order and the grayscale values ​​of each second candidate pixel, a grayscale difference between each second candidate pixel and the first pixel in the current traversal order; The second candidate pixel point with the smallest grayscale difference is set as the second pixel point that matches the first pixel point in the current traversal order.

4. The method according to claim 2, characterized in that: The on-site lane line information includes pixel coordinates of the on-site lane line segments in the on-site image of the driving area; and extracting the on-site lane line information in the on-site image of the driving area includes: Extracting on-site lane line features in the on-site image of the driving area through a pre-trained lane line recognition model; Generate a lane line probability map according to the on-site lane line features; According to the lane line probability map, the pixel coordinates of the on-site lane line segment in the on-site image of the driving area are determined.

5. The method according to claim 4, characterized in that The determining, according to the on-site lane line information and the depth information of the driving area, a second spatial position of the on-site lane line segment in the driving area comprises: Filtering the depth information of the on-site lane line segment from the depth information of the driving area according to the pixel coordinates of the on-site lane line segment in the on-site image of the driving area; The second spatial position of the on-site lane line segment is determined according to the pixel coordinates of the on-site lane line segment in the on-site image of the driving area and the depth information of the on-site lane line segment.

6. The method according to claim 1, characterized in that The method of obtaining the lane line map also includes: Segmenting the driving area to obtain at least one segmented area; Acquire a third image and a laser point cloud collected for the current segmented area; Filtering out a lane line point cloud projected onto a standard lane line segment in the third image from the laser point cloud; Determine a fourth spatial position of a standard lane line segment in the current segment area in the driving area according to the position information of the lane line point cloud in the current segment area; A lane map is obtained by combining the fourth spatial position of each standard lane line segment in the driving area.

7. The method according to claim 6, characterized in that The third image is an image frame acquired by an image acquisition device for the current segmented area; the laser point cloud is point cloud data acquired by a laser radar for the current segmented area; The step of selecting a lane line point cloud projected onto a standard lane line segment in the third image from the laser point cloud comprises: Determine the laser coordinate system of the laser radar and the camera coordinate system of the image acquisition device; Converting the laser point cloud in the laser coordinate system and the third image in the camera coordinate system into the same coordinate system; Projecting the laser point cloud onto a third image in the same coordinate system to obtain a radar projection image; According to the radar projection image, a lane line point cloud projected onto the standard lane line segment is determined.

8. The method according to claim 6, characterized in that The lane line point cloud includes a plurality of lane line points; and determining the fourth spatial position of the standard lane line segment in the current segment area in the driving area according to the position information of the lane line point cloud in the current segment area includes: Determine the position information of each lane line point in the driving area according to the position information of the lane line point cloud in the current segmented area; According to the position information of each lane line point in the driving area, curve fitting is performed on the multiple lane line points to obtain a lane line function corresponding to the current segmented area; According to the lane line function, a fourth spatial position of a standard lane line segment in a current segmented area in the driving area is determined.

9. A vehicle positioning device, comprising: A live lane line determination module is used to obtain a lane line map, a first spatial position of a vehicle in a driving area at a previous moment, and a live image of the driving area collected at a current moment; the live image of the driving area includes live lane line segments; and determine a second spatial position of the live lane line segments in the driving area by performing parallax detection on the live image of the driving area; a standard lane line determination module, used to determine a plurality of position prediction points located within a preset distance from the first spatial position, and to select standard lane line segments corresponding to each position prediction point from the lane line map; A positioning module, used to determine the position information of the target part in the on-site lane line segment in the driving area according to the second spatial position; Traversing the standard lane line segments; The position information of the target part in the standard lane line segment of the current traversal order is subtracted from the position information of the target part in the on-site lane line segment within the driving area to obtain a position difference; the absolute value of the position difference is used as the distance difference between the standard lane line segment of the current traversal order and the on-site lane line segment; based on the distance difference, the probability value of the position prediction point corresponding to the standard lane line segment of the current traversal order belongs to the position of the vehicle at the current moment is determined; the third spatial position of a position prediction point in the driving area with a probability value higher than a preset threshold is set as the spatial position of the vehicle at the current moment.

10. A computer device comprising a memory and one or more processors, wherein the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the one or more processors, the one or more processors perform the following steps: Acquire a lane line map, a first spatial position of the vehicle in the driving area at a previous moment, and a scene image of the driving area collected at a current moment; the scene image of the driving area includes a scene lane line segment; Determine a second spatial position of the on-site lane line segment in the driving area by performing parallax detection on the on-site image of the driving area, and determine position information of a target part in the on-site lane line segment in the driving area according to the second spatial position; Determine a plurality of position prediction points within a preset distance of the first spatial position, and select standard lane line segments corresponding to each position prediction point from the lane line map; Traversing the standard lane line segments; Subtract the position information of the target part in the standard lane line segment in the current traversal order from the position information of the target part in the on-site lane line segment in the driving area to obtain a position difference; The absolute value of the position difference is used as the distance difference between the standard lane line segment and the on-site lane line segment in the current traversal order; Determine, based on the distance difference, a probability value that the position prediction point corresponding to the standard lane line segment in the current traversal order belongs to the position of the vehicle at the current moment; A third spatial position of a position prediction point having a probability value higher than a preset threshold in the driving area is set as the spatial position of the vehicle at the current moment.

11. The computer device according to claim 10, characterized in that The driving area scene image includes a first image and a second image; when the processor executes the computer-readable instructions, the processor further performs the following steps: Performing pixel matching on the first image and the second image, and determining a disparity between the first image and the second image according to a matching result; Determining depth information of the driving area according to the parallax; Extracting on-site lane line information from the on-site image of the driving area; According to the on-site lane line information and the depth information of the driving area, a second spatial position of the on-site lane line segment in the driving area is determined.

12. The computer device according to claim 11, characterized in that When the processor executes the computer readable instructions, the processor further performs the following steps: Traversing at least one first pixel point in the first image; Determine at least one second candidate pixel point located in the second image according to the pixel coordinates of the first pixel point in the current traversal order in the first image; Determine, according to the grayscale value of the first pixel in the current traversal order and the grayscale values ​​of each second candidate pixel, a grayscale difference between each second candidate pixel and the first pixel in the current traversal order; The second candidate pixel point with the smallest grayscale difference is set as the second pixel point that matches the first pixel point in the current traversal order.

13. The computer device according to claim 11, characterized in that The on-site lane line information includes pixel coordinates of the on-site lane line segments in the on-site image of the driving area; when the processor executes the computer-readable instructions, the processor further performs the following steps: Extracting on-site lane line features in the on-site image of the driving area through a pre-trained lane line recognition model; Generate a lane line probability map according to the on-site lane line features; According to the lane line probability map, the pixel coordinates of the on-site lane line segment in the on-site image of the driving area are determined.

14. The computer device according to claim 13, characterized in that When the processor executes the computer readable instructions, the processor further performs the following steps: Filtering the depth information of the on-site lane line segment from the depth information of the driving area according to the pixel coordinates of the on-site lane line segment in the on-site image of the driving area; The second spatial position of the on-site lane line segment is determined according to the pixel coordinates of the on-site lane line segment in the on-site image of the driving area and the depth information of the on-site lane line segment.

15. One or more computer storage media storing computer readable instructions, which when executed by one or more processors cause the one or more processors to perform the following steps: Acquire a lane line map, a first spatial position of the vehicle in the driving area at a previous moment, and a scene image of the driving area collected at a current moment; the scene image of the driving area includes a scene lane line segment; Determine a second spatial position of the on-site lane line segment in the driving area by performing parallax detection on the on-site image of the driving area, and determine position information of a target part in the on-site lane line segment in the driving area according to the second spatial position; Determine a plurality of position prediction points within a preset distance of the first spatial position, and select standard lane line segments corresponding to each position prediction point from the lane line map; Traversing the standard lane line segments; Subtract the position information of the target part in the standard lane line segment in the current traversal order from the position information of the target part in the on-site lane line segment in the driving area to obtain a position difference; The absolute value of the position difference is used as the distance difference between the standard lane line segment and the on-site lane line segment in the current traversal order; Determine, based on the distance difference, a probability value that the position prediction point corresponding to the standard lane line segment in the current traversal order belongs to the position of the vehicle at the current moment; A third spatial position of a position prediction point having a probability value higher than a preset threshold in the driving area is set as the spatial position of the vehicle at the current moment.

16. The storage medium according to claim 15, characterized in that The driving area scene image includes a first image and a second image; when the computer readable instructions are executed by the processor, the following steps are further performed: Performing pixel matching on the first image and the second image, and determining a disparity between the first image and the second image according to a matching result; Determining depth information of the driving area according to the parallax; Extracting on-site lane line information from the on-site image of the driving area; According to the on-site lane line information and the depth information of the driving area, a second spatial position of the on-site lane line segment in the driving area is determined.

17. The storage medium according to claim 16, characterized in that The driving area scene image includes a first image and a second image; when the computer readable instructions are executed by the processor, the following steps are further performed: Traversing at least one first pixel point in the first image; Determine at least one second candidate pixel point located in the second image according to the pixel coordinates of the first pixel point in the current traversal order in the first image; Determine, according to the grayscale value of the first pixel in the current traversal order and the grayscale values ​​of each second candidate pixel, a grayscale difference between each second candidate pixel and the first pixel in the current traversal order; The second candidate pixel point with the smallest grayscale difference is set as the second pixel point that matches the first pixel point in the current traversal order.

18. The storage medium according to claim 16, characterized in that The on-site lane line information includes pixel coordinates of the on-site lane line segments in the on-site image of the driving area; when the computer-readable instructions are executed by the processor, the following steps are further performed: Extracting on-site lane line features in the on-site image of the driving area through a pre-trained lane line recognition model; Generate a lane line probability map according to the on-site lane line features; According to the lane line probability map, the pixel coordinates of the on-site lane line segment in the on-site image of the driving area are determined.

19. The storage medium according to claim 18, characterized in that When the computer readable instructions are executed by the processor, the following steps are further performed: Filtering the depth information of the on-site lane line segment from the depth information of the driving area according to the pixel coordinates of the on-site lane line segment in the on-site image of the driving area; The second spatial position of the on-site lane line segment is determined according to the pixel coordinates of the on-site lane line segment in the on-site image of the driving area and the depth information of the on-site lane line segment.

Images