Method for determining a drivable area of a vehicle and vehicle

Abstract

The embodiment of the application provides a drivable area determination method and vehicle, and the method comprises the following steps: acquiring a traffic channel line set of a traffic channel in which the vehicle travels; determining a first traffic channel line subset based on attribute information of the traffic channel lines in the traffic channel line set; determining a second traffic channel line subset based on lateral information of the traffic channel lines in the first traffic channel line subset; determining a drivable boundary of the vehicle and two end points of the traffic channel lines in the second traffic channel line subset based on the second traffic channel line subset; and determining a drivable area by the drivable boundary and an area surrounded by straight lines of the two end points along the lateral direction. The application solves the technical problem of low accuracy of the drivable area determination of the vehicle.

Description

Technical Field

[0001] This application relates to the field of vehicles, and more specifically, to a method for determining the drivable area of ​​a vehicle and a vehicle. Background Technology

[0002] In related technologies, visual semantic segmentation or LiDAR point cloud clustering methods are used to determine the drivable area of ​​a vehicle. The aforementioned visual semantic segmentation method performs pixel-level classification on vehicle images to obtain a mask of the drivable area, thereby determining the drivable region. However, these methods are easily affected by lighting conditions, shadows, and rain or snow, leading to distortion of the drivable area and reducing the accuracy of drivable region determination.

[0003] The aforementioned lidar point cloud clustering method determines the drivable area by projecting the vehicle's 3D point cloud onto a bird's-eye view and then performing clustering to extract roadside or obstacle boundaries. However, this method is prone to failure on open roads without physical roadside barriers, resulting in gaps in the drivable area and reducing the accuracy of vehicle drivability determination.

[0004] In summary, the methods described above still suffer from the technical problem of low accuracy in determining the vehicle's drivable area. Currently, there is no satisfactory solution to this problem. Summary of the Invention

[0005] This application provides a method for determining the drivable area of ​​a vehicle and a vehicle, so as to at least solve the technical problem of low accuracy in determining the drivable area of ​​a vehicle.

[0006] According to one aspect of the embodiments of this application, a method for determining the drivable area of ​​a vehicle is provided. The method may include: acquiring a set of traffic lane lines in a traffic corridor where the vehicle is traveling; determining a first subset of traffic lane lines based on attribute information of the traffic lane lines in the set, wherein the attribute information represents the state of the traffic lane lines in a time dimension and / or spatial dimension, and the first subset of traffic lane lines satisfies the time dimension and / or spatial dimension requirements for determining the drivable area; determining a second subset of traffic lane lines based on lateral information of the traffic lane lines in the first subset of traffic lane lines, wherein the lateral information represents the positional relationship between multiple traffic lane lines in the lateral direction of the traffic corridor, and the second subset of traffic lane lines satisfies the positional relationship requirements for determining the drivable area; determining the drivable boundary of the vehicle and the two endpoints of the traffic lane lines in the second subset of traffic lane lines based on the second subset of traffic lane lines; and defining the area enclosed by the drivable boundary and the two endpoints along lateral straight lines as the drivable area.

[0007] Furthermore, based on the attribute information of the traffic corridor lines in the traffic corridor line set, a first subset of traffic corridor lines is determined, including: determining traffic corridor lines whose attribute information satisfies time-domain conditions as elements of the first subset of traffic corridor lines, wherein the time-domain conditions are used to represent the accuracy requirements that the traffic corridor lines need to meet in the time dimension; and / or, determining traffic corridor lines whose attribute information satisfies spatial conditions as elements of the first subset of traffic corridor lines, wherein the spatial conditions are used to represent the accuracy requirements that the traffic corridor lines need to meet in the spatial dimension.

[0008] Further, the attribute information includes the identification information of traffic corridor lines. Traffic corridor lines whose attribute information satisfies the time-domain conditions are grouped together and determined as a first subset of traffic corridor lines, including at least one of the following: In response to traffic corridor lines with the same identification information, if the duration of consecutive occurrences is greater than or equal to a duration threshold, the identification information is determined to satisfy the time-domain conditions, and the traffic corridor lines with the identification information are determined as elements of the first subset of traffic corridor lines; In response to the occurrence of a traffic corridor line with target identification information, the identification information is determined to satisfy the time-domain conditions, and the traffic corridor lines with the target identification information are determined as elements of the first subset of traffic corridor lines, wherein the target identification information is identification information that has not appeared in historical moments before the current moment; In response to the identification information being in a non-lost state, the identification information is determined to satisfy the time-domain conditions, and the traffic corridor lines with the identification information are determined as elements of the first subset of traffic corridor lines. The elements in the set; and / or, the attribute information includes first endpoint information and second endpoint information of the traffic corridor line, the first endpoint information being used to indicate the location of the first endpoint of the traffic corridor line, the second endpoint information being used to indicate the location of the second endpoint of the traffic corridor line, and the distance between the second endpoint and the vehicle being greater than the distance between the first endpoint and the vehicle, the traffic corridor lines whose attribute information satisfies the spatial conditions are determined as elements in the first traffic corridor line subset, including: in response to the first endpoint information of the traffic corridor line being less than or equal to a first attention distance, and the second endpoint information being greater than or equal to a second attention distance, determining that the first endpoint information and the second endpoint information satisfy the spatial conditions, wherein the first attention distance is less than the second attention distance; the traffic corridor lines whose first endpoint information and the second endpoint information satisfy the spatial conditions are determined as elements in the first traffic corridor line subset.

[0009] Furthermore, the second traffic corridor subset includes the first traffic corridor and the second traffic corridor. Based on the lateral information of the traffic corridors in the first traffic corridor subset, the second traffic corridor subset is determined, including: identifying the traffic corridor with the largest lateral information as the first traffic corridor, and identifying the traffic corridor with the smallest lateral information as the second traffic corridor; and / or, based on the second traffic corridor subset, the drivable boundary of the vehicle is determined, including: determining the target longitudinal length of the traffic corridors in the second traffic corridor subset; determining the reference line set corresponding to the second traffic corridor subset based on the target longitudinal length; and determining the drivable boundary based on the reference line set.

[0010] Furthermore, the attribute information includes the first endpoint information, the second endpoint information, and the type information of the traffic corridor line. The first endpoint information indicates the location of the first endpoint of the traffic corridor line, the second endpoint information indicates the location of the second endpoint of the traffic corridor line, and the distance between the second endpoint and the vehicle is greater than the distance between the first endpoint and the vehicle. The type information indicates whether there are physical obstacles in the traffic corridor line. Determining the target longitudinal length of the traffic corridor lines in the second traffic corridor line subset includes: determining the initial longitudinal length of the traffic corridor lines in the second traffic corridor line subset based on the difference between the first endpoint information and the second endpoint information; and weighting the initial longitudinal length based on the type information to obtain the target longitudinal length.

[0011] Further, based on the target longitudinal length, a reference line set corresponding to the second traffic corridor line subset is determined, including: identifying the traffic corridor line with the largest target longitudinal length in the second traffic corridor line subset as an element in the reference line set; and / or, based on the reference line set, a drivable boundary is determined, including: obtaining the equation intercepts corresponding to the traffic corridor lines in the second traffic corridor line subset; smoothing the equation intercepts to obtain smoothed equation intercepts, and determining the smoothed equation intercepts as the reference offset of the second traffic corridor line subset; replacing the equation intercepts of the reference line equations of the reference lines in the reference line set with the reference offsets, and determining the replaced reference line equations as the boundary equations of the drivable boundary.

[0012] Furthermore, the drivable boundary includes a first drivable boundary and a second drivable boundary, and the endpoints include a first endpoint and a second endpoint. The distance between the second endpoint and the vehicle is greater than the distance between the first endpoint and the vehicle. The area enclosed by the drivable boundary and the two endpoints along transverse straight lines is defined as the drivable area. This includes defining the area enclosed by the boundary equation of the first drivable boundary, the boundary equation of the second drivable boundary, the equation of the first endpoint along the transverse straight line, and the equation of the second endpoint along the transverse straight line as the drivable area.

[0013] Furthermore, the method also includes: sampling the traffic channel lines in the first traffic channel line subset along the longitudinal direction of the traffic channel to obtain multiple sampling points on the traffic channel lines; and determining the lateral information of the traffic channel lines based on the position information of the multiple sampling points on the traffic channel lines.

[0014] Furthermore, the location information includes lateral location information. Based on the location information of multiple sampling points on the traffic corridor line, the lateral information of the traffic corridor line is determined, including: the sum of the lateral location information of multiple sampling points on the traffic corridor line is used to determine the lateral information of the traffic corridor line.

[0015] According to another aspect of the embodiments of this application, a vehicle is also provided, including: a memory storing an executable program; and a processor for running the program, wherein the program executes the methods in various embodiments of this application when it runs.

[0016] According to another aspect of the embodiments of this application, an electronic device is also provided, including: a memory storing an executable program; and a processor for running the program, wherein the program runs the methods of various embodiments of this application.

[0017] According to another aspect of the embodiments of this application, a computer-readable storage medium is also provided, the computer-readable storage medium including a stored executable program, wherein, when the executable program is running, it controls the device where the computer-readable storage medium is located to perform the methods of various embodiments of this application.

[0018] According to another aspect of the embodiments of this application, a computer program product is also provided, including a computer program that, when executed by a processor, implements the methods of various embodiments of this application.

[0019] According to another aspect of the embodiments of this application, a computer program product is also provided, including a non-volatile computer-readable storage medium storing a computer program, which, when executed by a processor, implements the methods in various embodiments of this application.

[0020] According to another aspect of the embodiments of this application, a computer program is also provided, which, when executed by a processor, implements the methods of the various embodiments of this application.

[0021] According to another aspect of the embodiments of this application, a device for determining the drivable area of ​​a vehicle is also provided. The device may include: an acquisition module for acquiring a set of traffic lane lines in a traffic corridor where the vehicle is traveling; a determination module for determining a first subset of traffic lane lines based on attribute information of the traffic lane lines in the set, wherein the attribute information represents the state of the traffic lane lines in a time dimension and / or spatial dimension, and the first subset of traffic lane lines satisfies the time dimension and / or spatial dimension requirements for determining the drivable area; a second determination module for determining a second subset of traffic lane lines based on lateral information of the traffic lane lines in the first subset of traffic lane lines, wherein the lateral information represents the positional relationship between multiple traffic lane lines in the lateral direction of the traffic corridor, and the second subset of traffic lane lines satisfies the positional relationship requirements for determining the drivable area; a third determination module for determining the drivable boundary of the vehicle and the two endpoints of the traffic lane lines in the second subset of traffic lane lines based on the second subset of traffic lane lines; and a fourth determination module for determining the area enclosed by the drivable boundary and the two endpoints along lateral straight lines as the drivable area.

[0022] In this embodiment, a set of traffic lane lines in the traffic corridor where the vehicle is traveling is obtained; based on the attribute information of the traffic lane lines in the set, a first subset of traffic lane lines is determined, wherein the attribute information is used to represent the state of the traffic lane lines in the time dimension and / or spatial dimension, and the first subset of traffic lane lines satisfies the time dimension and / or spatial dimension requirements for determining the drivable area; based on the lateral information of the traffic lane lines in the first subset of traffic lane lines, a second subset of traffic lane lines is determined, wherein the lateral information is used to represent the positional relationship between multiple traffic lane lines in the lateral direction of the traffic corridor, and the second subset of traffic lane lines satisfies the positional relationship requirements for determining the drivable area; based on the second subset of traffic lane lines, the drivable boundary of the vehicle and the two endpoints of the traffic lane lines in the second subset of traffic lane lines are determined; the area enclosed by the drivable boundary and the two endpoints along the lateral straight lines is determined as the drivable area. In other words, in this embodiment, effective traffic lane lines in the time and / or spatial dimensions are filtered based on attribute information, and the drivable boundaries are further filtered based on the lateral information of the aforementioned traffic lane lines. The left and right boundaries of the traffic lane where the vehicle is located are dynamically and adaptively reconstructed, thereby generating a drivable area without relying on lidar or high computing power. This method overcomes the obstacles in related technologies where determining the drivable area of ​​a vehicle is easily affected by lighting, shadows, and rain / snow, and is prone to failure on open roads without physical curbs. It avoids the low accuracy of determining the drivable area of ​​a vehicle due to distortion or incompleteness, thus solving the technical problem of low accuracy in determining the drivable area of ​​a vehicle and achieving the technical effect of improving the accuracy of determining the drivable area of ​​a vehicle. Attached Figure Description

[0023] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0024] Figure 1 This is a schematic diagram of a scenario for determining the drivable area of ​​a vehicle according to an embodiment of this application;

[0025] Figure 2 This is a flowchart of a method for determining the drivable area of ​​a vehicle according to an embodiment of this application;

[0026] Figure 3 This is a schematic diagram of a generation architecture for real-time reconstruction of drivable road surface areas based on single-frame lane line equations, according to an embodiment of this application.

[0027] Figure 4 This is a schematic diagram of a coordinate system according to an embodiment of this application;

[0028] Figure 5 This is a flowchart of a method for generating a drivable road surface area in real time based on a single-frame lane line equation, according to an embodiment of this application.

[0029] Figure 6 This is a schematic diagram of a vehicle drivable area determination device according to an embodiment of this application. Detailed Implementation

[0030] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present application.

[0031] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0032] According to an embodiment of this application, a method for determining the drivable area of ​​a vehicle is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0033] Figure 1 This is a schematic diagram illustrating a scenario for determining the drivable area of ​​a vehicle according to an embodiment of this application, such as... Figure 1 As shown, the scenario described above may include terminal device 10, network 20, server 30, and vehicle 40. Terminal device 10 can be used to obtain a vehicle's drivable area determination instruction, which can be sent to vehicle 40 via network 20. After receiving the drivable area determination instruction, vehicle 40 can send a single-frame image request to server 30. Upon receiving the single-frame image request from vehicle 40, server 30 can identify the traffic lane where the vehicle is located in real time, obtaining single-frame image information. This single-frame image information can then be sent to vehicle 40. At this point, vehicle 40 needs to identify the traffic lane on the road it is currently on from the single-frame image information. After obtaining the traffic lane, vehicle 40 can execute steps S102 to S110 to determine the vehicle's drivable area.

[0034] The following steps can be performed by vehicle 40: Step S102, obtain the traffic lane line set of the traffic lane in which the vehicle is traveling; Step S104, determine a first traffic lane line subset based on the attribute information of the traffic lane lines in the traffic lane line set; Step S106, determine a second traffic lane line subset based on the lateral information of the traffic lane lines in the first traffic lane line subset; Step S108, determine the drivable boundary of the vehicle and the two endpoints of the traffic lane lines in the second traffic lane line subset based on the second traffic lane line subset; Step S110, determine the drivable area as the area enclosed by the drivable boundary and the two endpoints along the lateral straight lines respectively.

[0035] In this embodiment, effective traffic lane lines in the time and / or spatial dimensions are filtered based on attribute information. These traffic lane lines are further filtered based on their lateral information to determine drivable boundaries. The left and right boundaries of the traffic lane where the vehicle is located are dynamically and adaptively reconstructed, thereby generating a drivable area without relying on lidar or high computing power. This method overcomes the limitations of related technologies, such as susceptibility to lighting, shadows, rain, and snow, and the tendency to fail on open roads without physical curbs. It thus solves the technical problem of low accuracy in determining the drivable area of ​​a vehicle, achieving the technical effect of improving the accuracy of drivable area determination.

[0036] This embodiment provides a method for determining the drivable area of ​​a vehicle. Figure 2 This is a flowchart of a method for determining the drivable area of ​​a vehicle according to an embodiment of this application, such as... Figure 2 As shown, the method may include the following steps.

[0037] Step S202: Obtain the traffic channel line set of the traffic channel in which the vehicle is traveling.

[0038] In the technical solution provided by step S202 of this application embodiment, the vehicle can be a vehicle equipped with various cameras or sensors. The traffic channel can be used to represent a continuous spatial area defined by road markings where vehicles can legally travel, and can include straight roads, curves, ramps, construction zones, or open road sections without physical guardrails. The traffic channel line set can be used to represent a dataset of traffic channel lines within a single frame of an image acquired by the various cameras or sensors equipped on the vehicle. For example, the traffic channel can include, but is not limited to, lanes; correspondingly, the traffic channel lines can be lane lines. The traffic channel line set can be a single-frame lane line dataset.

[0039] Optionally, if the vehicle is traveling in a traffic lane, the traffic lane line set of the traffic lane in which the vehicle is traveling can be obtained.

[0040] Optionally, images of traffic lanes are captured by the vehicle's camera, and these images are input into a pre-trained deep learning model. The deep learning model can then identify at least one traffic lane line in the images, and the identified at least one traffic lane line can be combined into a traffic lane line set.

[0041] It should be noted that the above method of identifying traffic lane lines in images using deep learning models is only an example and does not impose specific limitations on the method of obtaining traffic lane line sets.

[0042] In this embodiment of the application, the traffic lane set of the traffic lane in which the vehicle is traveling can be obtained through the above step S202, which provides an efficient and practical technical input for subsequent steps.

[0043] Step S204: Based on the attribute information of traffic corridor lines in the traffic corridor line set, determine the first traffic corridor line subset.

[0044] In the technical solution provided by step S204 of this application embodiment, the aforementioned attribute information can be used to represent the state of traffic lane lines in the time and / or spatial dimensions. The attribute information may include the cubic parametric equation of the traffic lane line within a single frame image, the longitudinal coordinates of the starting point and the ending point, the semantic type of the traffic lane line, and the unique identifier (ID) of the traffic lane line. The aforementioned first subset of traffic lane lines can be used to represent the set of traffic lane lines that, after time filtering (e.g., temporal filtering of the unique ID of each traffic lane line) and / or spatial filtering, meet the time and / or spatial requirements of the current driving scenario. The aforementioned time dimension requirement can be used to represent the temporal continuity that traffic lane lines need to possess. The aforementioned spatial dimension requirement can be used to represent that traffic lane lines need to cover the area of ​​core driver focus, avoiding interference from invalid traffic lane lines at a distance or behind.

[0045] Optionally, after obtaining the set of traffic lane lines in the traffic lanes where the vehicle is traveling, a first subset of traffic lane lines can be determined based on the attribute information of the traffic lane lines in the set.

[0046] Optionally, time-domain filtering can be performed on the unique ID of each traffic corridor line to retain traffic corridor lines that meet the time dimension requirements (e.g., the duration of continuous occurrence is not less than a preset threshold); spatial filtering can be performed on traffic corridor lines in the first subset of traffic corridor lines to retain traffic corridor lines that meet the spatial dimension requirements (e.g., the longitudinal coordinates of the starting point and the longitudinal coordinates of the ending point meet preset conditions); traffic corridor lines that meet the above time dimension requirements and / or spatial dimension requirements can be combined into the first subset of traffic corridor lines.

[0047] For example, the least squares method can be used to generate the cubic parametric equations corresponding to the traffic lane lines; the semantic classification network can be combined to determine the semantic type of the traffic lane lines (e.g., ordinary lane lines or roadside), and a unique ID can be assigned to the traffic lane lines based on a multi-frame tracking algorithm.

[0048] It should be noted that the above-mentioned methods of generating cubic parametric equations corresponding to traffic corridor lines using the least squares method, determining the semantic type of traffic corridor lines using a semantic classification network, and assigning unique IDs to traffic corridor lines using a multi-frame tracking algorithm are only illustrative examples and do not impose specific restrictions on the methods for obtaining attribute information of traffic corridor line sets.

[0049] In this embodiment of the application, a stable and effective traffic corridor line screening mechanism can be constructed through the above step S204, which retains only traffic corridor lines that meet the requirements of the time dimension and / or the spatial dimension, thereby suppressing perceived noise interference.

[0050] Step S206: Based on the lateral information of the traffic corridor lines in the first traffic corridor line subset, determine the second traffic corridor line subset.

[0051] In the technical solution provided by step S206 of the embodiments of this application, the aforementioned lateral information can be used to represent the lateral positional relationship between multiple traffic channel lines within a traffic channel. The aforementioned second subset of traffic channel lines can be used to represent a set of traffic channel lines that can characterize the left and right boundaries of a traffic channel, further filtered based on the lateral positional relationship from the first subset of traffic channel lines.

[0052] Optionally, after determining the first subset of traffic corridors based on the attribute information of the traffic corridors in the traffic corridor set, the second subset of traffic corridors can be determined based on the lateral information of the traffic corridors in the first subset of traffic corridors.

[0053] Optionally, determining the second subset of traffic lane lines based on the lateral information of the traffic lane lines in the first subset may include obtaining lateral information, i.e., the lateral offset of the traffic lane lines (also known as the lateral position score), based on the cubic parametric equations corresponding to the traffic lane lines in the first subset. The lateral offsets can be sorted from largest to smallest, and the traffic lane line corresponding to the largest lateral offset can be selected as the leftmost traffic lane line (also known as the leftmost boundary, denoted by L1), and the traffic lane line corresponding to the smallest lateral offset can be selected as the rightmost traffic lane line (also known as the rightmost boundary, denoted by R1). The set formed by the leftmost and rightmost traffic lane lines can be considered the second subset of traffic lane lines. For example, the leftmost traffic lane line can be the leftmost lane line, and the rightmost traffic lane line can be the rightmost lane line.

[0054] In this embodiment of the application, the traffic lane line corresponding to the maximum lateral offset in the current road environment can be identified and extracted as the leftmost traffic lane line through the above step S206, and the traffic lane line corresponding to the minimum lateral offset can be extracted as the rightmost traffic lane line. The leftmost traffic lane line and the rightmost traffic lane line constitute the second traffic lane line subset, thereby determining the left and right boundary range of the traffic lane where the vehicle is currently located, and providing an effective benchmark reference for subsequent steps.

[0055] Step S208: Based on the second traffic channel line subset, determine the drivable boundary of the vehicle and the two endpoints of the traffic channel lines in the second traffic channel line subset.

[0056] In the technical solution provided by step S208 in the embodiments of this application, the drivable boundary can be used to represent the left boundary curve (which can be represented by L_boundary) generated based on the leftmost boundary of the second traffic channel line subset, and the right boundary curve (which can be represented by R_boundary) generated based on the rightmost boundary of the second traffic channel line subset; the endpoint can be used to represent the start and end points of the leftmost and rightmost boundaries of the second traffic channel line subset in the longitudinal direction (x-axis direction).

[0057] Optionally, after determining the second traffic channel subset based on the lateral information of the traffic channel lines in the first traffic channel subset, the drivable boundary of the vehicle and the two endpoints of the traffic channel lines in the second traffic channel subset can be determined based on the second traffic channel subset.

[0058] Optionally, based on the cubic parametric equations of the leftmost and rightmost boundaries in the above-mentioned second traffic channel line subset, the intercepts corresponding to the cubic parametric equations of the leftmost and rightmost boundaries are averaged to generate the left boundary curve and the right boundary curve; the starting point and ending point of the leftmost and rightmost boundaries in the longitudinal direction can be used as the two endpoints of the traffic channel lines in the second traffic channel line subset.

[0059] In this embodiment of the application, the above step S208 can reconstruct a drivable boundary without jitter and with high consistency based on the leftmost and rightmost boundaries of the second traffic channel line subset, thereby achieving a balance between morphological fidelity and stability of the drivable boundary.

[0060] Step S210: The area enclosed by the drivable boundary and the two endpoints along the horizontal straight lines is defined as the drivable area.

[0061] In the technical solution provided by step S210 in the embodiments of this application, the aforementioned drivable area can be used to represent the road surface range in front of the vehicle that is safe to pass through, and can also be referred to as the drivable road surface area.

[0062] Optionally, after determining the drivable boundary of the vehicle and the two endpoints of the traffic channel lines in the second traffic channel line subset based on the drivable boundary, the area enclosed by the drivable boundary and the two endpoints along the lateral straight lines can be defined as the drivable area.

[0063] Optionally, the generated left and right boundary curves are used as the two sides of the drivable area. Perpendicular lines are drawn from the two endpoints along the horizontal direction (y-axis) to form a front (starting point) perpendicular line (which can be represented by x=xs) and a rear (ending point) perpendicular line (which can be represented by x=xe). The left and right boundary curves are then connected. The closed polygonal area enclosed by the left and right boundary curves, the front and rear perpendicular lines is defined as the drivable area.

[0064] In steps S202 to S210 of this embodiment, a set of traffic lane lines in the traffic lanes in which the vehicle travels is obtained; based on the attribute information of the traffic lane lines in the set, a first subset of traffic lane lines is determined, wherein the attribute information is used to represent the state of the traffic lane lines in the time dimension and / or spatial dimension, and the first subset of traffic lane lines meets the time dimension and / or spatial dimension requirements for determining the drivable area; based on the lateral information of the traffic lane lines in the first subset of traffic lane lines, a second subset of traffic lane lines is determined, wherein the lateral information is used to represent the positional relationship between multiple traffic lane lines in the lateral direction of the traffic lane, and the second subset of traffic lane lines meets the positional relationship requirements for determining the drivable area; based on the second subset of traffic lane lines, the drivable boundary of the vehicle and the two endpoints of the traffic lane lines in the second subset of traffic lane lines are determined; the area enclosed by the drivable boundary and the two endpoints along the lateral straight lines is determined as the drivable area. In other words, in this embodiment, effective traffic lane lines in the time and / or spatial dimensions are filtered based on attribute information, and the drivable boundaries are further filtered based on the lateral information of the aforementioned traffic lane lines. The left and right boundaries of the traffic lane where the vehicle is located are dynamically and adaptively reconstructed, thereby generating a drivable area without relying on lidar or high computing power. This method overcomes the obstacles of related technologies, such as susceptibility to the effects of lighting, shadows, rain, and snow, and the tendency to fail on open roads without physical curbs. Therefore, it solves the technical problem of low accuracy in determining the drivable area of ​​a vehicle, achieving the technical effect of improving the accuracy of determining the drivable area of ​​a vehicle.

[0065] The embodiments of this application will be described in detail below with reference to the steps described above.

[0066] As an optional implementation, step S204, based on the attribute information of traffic corridors in the traffic corridor set, determines a first subset of traffic corridors, including: determining traffic corridors in the traffic corridor set whose attribute information satisfies time-domain conditions as elements of the first subset of traffic corridors, wherein the time-domain conditions are used to represent the accuracy requirements that the traffic corridors need to meet in the time dimension; and / or, determining traffic corridors in the first subset of traffic corridors whose attribute information satisfies spatial conditions as elements of the first subset of traffic corridors, wherein the spatial conditions are used to represent the accuracy requirements that the traffic corridors need to meet in the spatial dimension.

[0067] In the embodiments of this application, the aforementioned time-domain conditions can be used to represent the accuracy requirements that a traffic corridor concentration needs to meet in the time dimension. The aforementioned spatial conditions can be used to represent the accuracy requirements that a traffic corridor concentration needs to meet in the spatial dimension.

[0068] Optionally, for traffic corridors in the traffic corridor set, based on the unique ID corresponding to the traffic corridor, if the number of frames in which the traffic corridor consistently appears in consecutive frames is not less than a preset threshold, then the traffic corridor can be determined to meet the temporal condition, and the traffic corridor that consistently appears in consecutive frames for a number of frames not less than the preset threshold can be included in the first subset of traffic corridors. If the traffic corridor does not meet the temporal condition, that is, if the number of frames in which the traffic corridor consistently appears in consecutive frames is less than the preset threshold, then it can be excluded from the first subset of the first traffic corridors.

[0069] Optionally, if a traffic lane line appears for the first time, the historical cache of the aforementioned traffic lane line can be initialized and counting can begin; if a unique ID is lost for more than a preset threshold frame consecutively, the traffic lane line corresponding to the aforementioned unique ID can be deleted from the cache, considered as an invalid detection, and not included in the first subset of the first traffic lane lines.

[0070] Optionally, for each traffic corridor in the set of traffic corridor lines, it is determined whether the longitudinal starting point and longitudinal ending point of the traffic corridor line meet the spatial conditions. If the longitudinal starting point and longitudinal ending point of the traffic corridor line meet the preset conditions, that is, if the traffic corridor line meets the spatial conditions, then the traffic corridor line whose longitudinal starting point and longitudinal ending point meet the preset conditions can be included in the first subset of the first traffic corridor lines. If the spatial conditions are not met, then it can be excluded from the first subset of the first traffic corridor lines.

[0071] In this embodiment of the application, the above method can actively filter traffic channels that meet the temporal and / or spatial conditions based on the attribute information of the traffic channels, and exclude traffic channels that do not meet the temporal and / or spatial conditions, i.e., noise interference, thereby obtaining a stable and reliable first subset of the first traffic channels.

[0072] As an optional implementation, the attribute information includes the identification information of traffic corridor lines. Traffic corridor lines whose attribute information satisfies a time-domain condition are grouped together and determined as a first subset of traffic corridor lines, including at least one of the following: In response to traffic corridor lines with the same identification information appearing consecutively for a duration greater than or equal to a duration threshold, the identification information is determined to satisfy the time-domain condition, and the traffic corridor lines with the identification information are determined as elements of the first subset of traffic corridor lines; In response to the appearance of a traffic corridor line with target identification information, the identification information is determined to satisfy the time-domain condition, and the traffic corridor line with the target identification information is determined as an element of the first subset of traffic corridor lines, wherein the target identification information is identification information that has not appeared in historical moments before the current moment; In response to the identification information being in a non-lost state, the identification information is determined to satisfy the time-domain condition, and the traffic corridor lines with the identification information are determined as elements of the first subset of traffic corridor lines.

[0073] In this embodiment, the aforementioned identification information can be used to represent the unique ID of each traffic corridor line. The aforementioned duration threshold can be used to represent the shortest duration of continuous observation in the time dimension required to determine that a traffic corridor line is stable and valid. The aforementioned target identification information can be identification information that has not appeared in historical moments before the current moment, used to identify newly emerging traffic corridor lines. The aforementioned non-loss status can be used to indicate that a certain traffic corridor line has been successfully detected in the most recent consecutive frames without interruption.

[0074] Optionally, for each traffic corridor line with the same unique ID, the number of frames that appear stably in consecutive frames is accumulated; when the accumulated number of frames reaches or exceeds a preset threshold (which can be used...), the threshold is determined. When the above unique ID is used, the traffic channel line corresponding to it will be included in the first traffic channel line subset; if the above cumulative frame count does not meet the standard, it will still be kept in the cache for continuous monitoring until the above traffic channel line's cumulative frame count reaches or exceeds the preset threshold or is cleared by timeout.

[0075] Optionally, the detected unique IDs are traversed; if a unique ID has never appeared within a preset time window (e.g., 5 to 10 seconds), the unique ID can be marked as target identification information; the historical cache of the target identification information can be initialized, the time of the first frame appearance of the target identification information can be recorded, and frame counting can be started. If the number of frames in which the target identification information appears stably in consecutive frames reaches or exceeds the preset threshold, the target identification information can be determined as an element in the first traffic channel line subset.

[0076] Optionally, for a series of unique IDs already existing in the cache, the detection continuity of the aforementioned unique IDs in the most recent few frames (e.g., 3 frames) can be continuously monitored; if the aforementioned unique IDs are continuously detected, it can be determined that the aforementioned unique IDs are not lost and can still be retained in the first traffic channel line subset to continue participating in subsequent processing; otherwise, the aforementioned unique IDs will be removed from the first traffic channel line subset.

[0077] For example, if the current unique ID is appearing for the first time, the unique ID can be written to the circular cache History[id] to record the time of its first appearance; if the unique ID appears consecutively for more than 10 frames .... ( =10, corresponding to 1000 milliseconds), then the above unique ID is allowed to participate in subsequent filtering; if the above unique ID appears consecutively for a number of frames < If the vehicle's position is constantly changing, the previous traffic lanes are no longer available.

[0078] In this embodiment of the application, an adaptive traffic corridor line time-domain state management mechanism can be constructed by the above method. By combining the duration threshold to continuously monitor the unique ID to achieve dynamic filtering, the scene adaptability of the first traffic corridor line subset is improved.

[0079] As an optional implementation, the attribute information includes first endpoint information and second endpoint information of the traffic channel line. The first endpoint information indicates the location of the first endpoint of the traffic channel line, and the second endpoint information indicates the location of the second endpoint of the traffic channel line. The distance between the second endpoint and the vehicle is greater than the distance between the first endpoint and the vehicle. The traffic channel lines whose attribute information satisfies the spatial conditions are determined as elements of the first traffic channel line subset, including: in response to the first endpoint information of the traffic channel line being less than or equal to a first attention distance and the second endpoint information being greater than or equal to a second attention distance, determining that the first endpoint information and the second endpoint information satisfy the spatial conditions, wherein the first attention distance is less than the second attention distance; the traffic channel lines whose first endpoint information and the second endpoint information satisfy the spatial conditions are determined as elements of the first traffic channel line subset.

[0080] In this embodiment, the first endpoint information can be used to indicate the location of the first endpoint of the traffic corridor line. The second endpoint information is used to indicate the location of the second endpoint of the traffic corridor line. The distance between the second endpoint and the vehicle is greater than the distance between the first endpoint and the vehicle. The first attention distance can be used to indicate the minimum effective observation distance threshold (which can be represented by S) that needs to be ensured to cover the area in front of the vehicle. The second attention distance can be used to indicate the minimum effective extension distance threshold (which can be represented by E) that needs to be maintained behind the vehicle.

[0081] Optionally, determine the longitudinal starting point of each traffic corridor line obtained above (which can be used as a reference). (representation), longitudinal endpoint (can be represented by) (Indicates) whether the first focus distance (S) and the second focus distance (E) obtained from a preset lookup table based on the vehicle's current speed satisfy the following conditions: Less than or equal to -S and If the condition is greater than or equal to E, it can be determined that the above traffic corridor line meets the spatial conditions.

[0082] Optionally, for each of the current traffic corridor lines, determine the corresponding traffic corridor lines one by one. and Does it meet the requirements? Less than or equal to -S and If the condition is greater than or equal to E, the unique ID, cubic parametric equation, semantic type, and other attribute information of the traffic corridor line can be recorded in the first traffic corridor line subset only if both conditions are met; if either of the above conditions is not met, the traffic corridor line can be discarded and will not participate in subsequent processing.

[0083] For example, to ensure that the generated drivable area always covers the driver's most important front and rear visible areas, and to avoid long lane lines outside these areas that would prevent the subsequently drawn road area from covering the vehicle's location, the traffic lane line is discarded if its longitudinal starting point is less than -S or its longitudinal ending point is greater than E.

[0084] In this embodiment, the above method can constrain the input for generating the drivable area boundary. It compares the first and second points of interest (S) with the longitudinal start and end points of the traffic corridor lines by obtaining the results from a preset lookup table based on the vehicle's current speed. Traffic corridor lines that do not meet the above conditions, i.e., false positives and invalid traffic corridor lines, are eliminated to ensure that the selected traffic corridor lines have complete spatial extensibility.

[0085] As an optional implementation, the second traffic corridor line subset includes the first traffic corridor line and the second traffic corridor line. Step S206, based on the lateral information of the traffic corridor lines in the first traffic corridor line subset, determines the second traffic corridor line subset, including: determining the traffic corridor line with the largest lateral information as the first traffic corridor line, and determining the traffic corridor line with the smallest lateral information as the second traffic corridor line.

[0086] In this embodiment of the application, the first traffic lane line can be used to represent the traffic lane line located on the far left of the vehicle in a valid traffic lane line. The second traffic lane line can be used to represent the traffic lane line located on the far right of the vehicle in a valid traffic lane line.

[0087] Optionally, for each traffic lane in the first traffic lane subset, N points (e.g., N=10) are sampled at equal intervals along the longitudinal direction of the traffic lane within a common effective observation interval. The lateral coordinates of each sampled point are calculated, and the arithmetic sum of the lateral coordinates of each sampled point is calculated as the lateral position score of that lane. The lane with the largest score can be identified as the first traffic lane; the lane with the smallest score can be identified as the second traffic lane.

[0088] In this embodiment, the above method is used to determine the boundaries of the drivable areas on the left and right sides of the traffic lanes based on the cubic parametric equation of the traffic lane lines. By sampling multiple traffic lane lines at equal intervals and calculating the lateral position score, the lateral position of the traffic lane lines is objectively quantified, avoiding the influence of relying on single-point intercepts or endpoints, thereby solving problems such as misjudgment of left and right boundaries caused by local noise and curve distortion.

[0089] As an optional implementation, determining the drivable boundary of a vehicle based on a subset of second traffic lane lines includes: determining the target longitudinal length of the traffic lane lines in the subset of second traffic lane lines; determining a reference line set corresponding to the subset of second traffic lane lines based on the target longitudinal length; and determining the drivable boundary based on the reference line set.

[0090] In this embodiment, the aforementioned target longitudinal length can be used to represent the actual effective extension distance of each traffic lane line within the effective observation range in front of the vehicle, and the weighted longitudinal representation value (which can be represented by Range′). The aforementioned reference line set can be used to represent the longest longitudinally selected traffic lane lines from the second traffic lane line subset after sorting them according to their target longitudinal length, including the left longest reference line L_max, i.e., the most representative left traffic lane line, and the right longest reference line R_max, i.e., the most representative right traffic lane line, forming the candidate boundary set.

[0091] Optionally, the traffic lane lines on the left and right sides, namely the first traffic lane line and the second traffic lane line, are sorted in descending order using the target longitudinal length of the first traffic lane line and the second traffic lane line as the sorting key; among them, the traffic lane line with the largest target longitudinal length on the left side can be recorded as the longest left reference line L_max; the traffic lane line with the largest target longitudinal length on the right side can be used as the longest right traffic lane line, which can be recorded as the longest right reference line R_max.

[0092] Optionally, the intercept of the cubic parametric equation of the longest left reference line is replaced with the left soft roadway reference offset after moving average processing to generate the equation of the corresponding drivable boundary on the left; the intercept of the cubic parametric equation of the longest right reference line is replaced with the right soft roadway reference offset after moving average processing to generate the equation of the corresponding drivable boundary on the right.

[0093] In the embodiments of this application, the above method can achieve the generation of highly continuous and semantically consistent boundaries of drivable areas with relatively low computing power.

[0094] As an optional implementation, the attribute information includes first endpoint information, second endpoint information, and type information of the traffic channel line. The first endpoint information indicates the location of the first endpoint of the traffic channel line, the second endpoint information indicates the location of the second endpoint of the traffic channel line, and the distance between the second endpoint and the vehicle is greater than the distance between the first endpoint and the vehicle. The type information indicates whether there are physical obstacles in the traffic channel line. Determining the target longitudinal length of the traffic channel lines in the second traffic channel line subset includes: determining the initial longitudinal length of the traffic channel lines in the second traffic channel line subset based on the difference between the first endpoint information and the second endpoint information; and weighting the initial longitudinal length based on the type information to obtain the target longitudinal length.

[0095] In this embodiment of the application, the aforementioned type information can be used to indicate whether there are physical obstacles on the traffic lane. The aforementioned first endpoint information can be used to indicate the location of the first endpoint of the traffic lane. The aforementioned second endpoint information can be used to indicate the location of the second endpoint of the traffic lane, wherein the distance between the second endpoint and the vehicle is greater than the distance between the first endpoint and the vehicle.

[0096] The aforementioned initial longitudinal length can be used to represent the original geometric extension distance of the traffic corridor along the longitudinal direction.

[0097] Optionally, for the traffic corridor lines in the second traffic corridor line subset, the initial longitudinal length of the corresponding traffic corridor line is obtained by calculating the difference between the longitudinal coordinates of the first endpoint information and the second endpoint information of the traffic corridor line, which can be represented by Range; the initial longitudinal length of the corresponding traffic corridor line can be calculated by the following formula.

[0098]

[0099] Optionally, the initial longitudinal length is weighted according to the type information of the traffic corridor (e.g., ordinary, curb, etc.) to generate the target longitudinal length. The target longitudinal length can be calculated using the following formula:

[0100]

[0101] in, This can be used to represent the coefficients used for weighting. Where the type information is "roadside," then... Greater than 1, otherwise The value is 1, and the target longitudinal length is equal to the initial longitudinal length.

[0102] As an optional implementation, a reference line set corresponding to the second traffic corridor line subset is determined based on the target longitudinal length, including: determining the traffic corridor line with the largest target longitudinal length in the second traffic corridor line subset as an element in the reference line set.

[0103] In this embodiment of the application, the aforementioned reference line set can be used to represent a set of boundary candidates composed of the target with the largest longitudinal length among the traffic channel lines on the left and right sides.

[0104] Optionally, a max-heap data structure is established for the traffic lane lines in the second traffic lane line subset, namely the first traffic lane line and the second traffic lane line, and sorted in descending order using the target longitudinal length of each traffic lane line as the sorting key; the sorting can be from largest to smallest. The left max-heap, i.e., the max-heap data structure established for the first traffic lane line, can have its top element taken as the left lane line with the largest target longitudinal length, which can be an element from the reference line set; the right max-heap, i.e., the max-heap data structure established for the second traffic lane line, can have its top element taken as the right lane line with the largest target longitudinal length, which can also be an element from the reference line set.

[0105] In this embodiment of the application, the above method can be used to construct a reference line set generation mechanism that is independent on the left and right sides and prioritizes the maximum longitudinal length of the target based on the identified traffic channel lines. This ensures that the most representative reference lines in the left and right lateral directions can still be selected in complex scenarios such as curves and ramps, thereby improving the robustness of the generation of the drivable area boundary.

[0106] As an optional implementation, determining the drivable boundary based on a set of reference lines includes: obtaining the equation intercepts corresponding to traffic lane lines in the second traffic lane line subset; smoothing the equation intercepts to obtain smoothed equation intercepts, and determining the smoothed equation intercepts as the reference offset of the second traffic lane line subset; replacing the equation intercepts of the reference line equations of the reference lines in the reference line subset with the reference offsets, and determining the replaced reference line equations as the boundary equations of the drivable boundary.

[0107] In this embodiment, the intercept of the above equation can be used to represent the lateral offset of the cubic parametric equation corresponding to the traffic channel line at the longitudinal coordinate. The above smoothing process can be used to represent the moving average operation of the intercept of the equation corresponding to the traffic channel line in the second traffic channel line subset. The above reference offset can be used to represent the statistical characteristics of the lateral position of the reference lines on both sides after temporal moving average filtering, which are stably converged in historical multiple frames. The above boundary equation can be used to represent the drivable area boundary curve finally used for rendering, which is composed of the original curvature structure of the reference line and the statistical smoothing offset. The boundary equation of the drivable boundary can include the left drivable boundary equation of the traffic channel line (which can be represented by L_boundary) and the right drivable boundary equation of the traffic channel line (which can be represented by R_boundary).

[0108] Optionally, for each reference line in the reference line set, the cubic parametric equation of the corresponding traffic corridor line can be extracted, which can be expressed by the following formula:

[0109]

[0110] in, It can represent the intercept of a cubic parametric equation. It can represent the initial slope of a cubic parametric equation. It can represent the curvature coefficients of a cubic parametric equation. It can represent the coefficient of the rate of change of curvature in a cubic parametric equation. It can represent the coordinate values ​​along the vehicle's forward direction (longitudinal direction). It can represent the coordinate value in the direction perpendicular to the vehicle's forward movement (lateral direction).

[0111] Optionally, the intercept of the equation corresponding to the left traffic corridor line in the second traffic corridor line subset is... (L1) is the right-hand intercept of the equation corresponding to the right-hand traffic corridor line in the second traffic corridor line subset. (R1) The sliding average value within the current valid frame is used as the reference offset, which can be expressed by the following formula:

[0112]

[0113] in, It can be used to represent offset values. It can be used to represent the intercept of the cubic parametric equation corresponding to the reference line. K can be used to represent the preset maximum capacity of the moving average buffer queue. It can be used to represent a buffer queue. n can be used to represent the actual number of valid frames in the buffer queue.

[0114] Optionally, the intercept term in the cubic parametric equation of the longest reference line L_max on the left is replaced with the left reference offset (which can be represented by osl) to obtain the equation of the drivable boundary on the left (which can be represented by L_boundary); the intercept term in the cubic parametric equation of the longest reference line R_max on the right is replaced with the right reference offset (which can be represented by osr) to obtain the equation of the drivable boundary on the right (which can be represented by R_boundary). In this case, the a0 of L_max and R_max is directly replaced with osl and osr, respectively, while the other coefficients remain unchanged to obtain the curve equation of the left line L_boundary, which can be expressed by the following formula.

[0115]

[0116] The equation of the right-hand boundary curve R_boundary can be expressed by the following formula.

[0117]

[0118] In this embodiment of the application, the above method generates drivable boundaries on both sides of the vehicle by identifying traffic lane lines, effectively preserving the true curvature of the road and improving the continuity and stability of the drivable area display.

[0119] As an optional implementation, the drivable boundary includes a first drivable boundary and a second drivable boundary, and the endpoints include a first endpoint and a second endpoint. The distance between the second endpoint and the vehicle is greater than the distance between the first endpoint and the vehicle. The area enclosed by the drivable boundary and the two endpoints along transverse straight lines is defined as the drivable area, including:

[0120] The area enclosed by the boundary equations of the first drivable boundary, the second drivable boundary, the straight line equation along the lateral direction of the first endpoint, and the straight line equation along the lateral direction of the second endpoint is defined as the drivable area.

[0121] In this embodiment, the equation of the line at the first endpoint can be used to represent a perpendicular line to the longitudinal axis that passes through the left reference line L_boundary at x=xs. The equation of the line at the second endpoint can be used to represent a perpendicular line to the longitudinal axis that passes through the right reference line R_boundary at x=xe.

[0122] Optionally, L_boundary, R_boundary, perpendicular lines x=xs and x=xe can be connected end to end in sequence to form a closed quadrilateral region, which will serve as the boundary of the drivable road surface in the final output.

[0123] In the embodiments of this application, the above method can be used to construct a drivable area boundary model with complete geometric closed loop and in line with driving cognition under a lightweight architecture that relies only on a single frame of traffic channel line equations.

[0124] As an optional implementation, the method further includes: sampling traffic corridor lines in the first traffic corridor line subset along the longitudinal direction of the traffic corridor to obtain multiple sampling points on the traffic corridor lines; and determining the lateral information of the traffic corridor lines based on the position information of the multiple sampling points on the traffic corridor lines.

[0125] In this embodiment, the aforementioned sampling points can be used to represent discrete spatial location points evenly distributed along the three-dimensional trajectory of the traffic lane line in the vehicle's forward direction (longitudinal direction). The aforementioned location information can be used to represent the two-dimensional spatial coordinates of each sampling point in the vehicle's coordinate system. The aforementioned lateral information represents the statistical characteristics of the overall lateral offset of the traffic lane line within the longitudinal interval of interest.

[0126] Optionally, based on the effective longitudinal range of the traffic corridor line retained in the current frame, N sampling points are uniformly generated within the interval at a fixed step size, where N is the preset number of sampling points; the longitudinal coordinates of each sampling point can be substituted into the cubic parametric equation of the corresponding traffic corridor line to calculate the corresponding transverse coordinates, thereby obtaining the set of N sampling points for the traffic corridor line.

[0127] Optionally, the lateral coordinates of the sampling points can be summed algebraically to obtain the lateral information of the corresponding traffic corridor.

[0128] For example, for each retained traffic corridor, a common range of horizontal axis values ​​is determined ( , Within the aforementioned range, N sampling points (N can be 10) are sampled at equal intervals along the horizontal axis. The above intervals are calculated... It can be expressed by the following formula.

[0129]

[0130] in, It can be used to represent the interval between sampling points. The vertical coordinates of the above sampling points can be calculated using the following formula.

[0131]

[0132] in, It can be used to represent the ordinate of a sampling point. It can be used to represent the longitudinal starting point of a traffic corridor. The lateral coordinates of the above sampling points can be calculated using the following formula.

[0133]

[0134] in, It can be used to represent the x-coordinate of a sampling point. The lateral information of the corresponding traffic corridor line can be calculated using the following formula.

[0135]

[0136] The Score can be used to represent the lateral information of the corresponding traffic lane. On the same road, the more left a traffic lane is, the higher its lateral information. The larger the value, the larger the score, and thus the leftmost; conversely, the smaller the value, the rightmost. The above algorithm has a complexity of O(N·M), where N is the number of points selected and M is the number of traffic corridor lines selected.

[0137] In this embodiment of the application, the leftmost and rightmost boundaries of the preserved traffic lane lines can be identified by the above method, thereby generating a drivable area boundary that conforms to the real road.

[0138] As an optional implementation, the location information includes lateral location information. Based on the location information of multiple sampling points on the traffic corridor line, the lateral information of the traffic corridor line is determined, including: the sum of the lateral location information of multiple sampling points on the traffic corridor line is used to determine the lateral information of the traffic corridor line.

[0139] In this embodiment of the application, the aforementioned lateral position information can be used to represent the actual offset distance of each sampling point relative to the origin of the coordinate system in the lateral direction.

[0140] Optionally, for the traffic corridor line, within its effective longitudinal interval, N sampling points are selected at equal intervals along the x-axis, and the lateral coordinates of each sampling point are obtained. and will obtain Add to, with The sum is used as lateral information for the lane lines.

[0141] In this embodiment, the above method can be used to obtain a set of traffic lane lines in the traffic corridor where the vehicle is traveling; based on the attribute information of the traffic lane lines in the set, a first subset of traffic lane lines is determined, wherein the attribute information is used to represent the state of the traffic lane lines in the time dimension and / or spatial dimension, and the first subset of traffic lane lines satisfies the time dimension and / or spatial dimension requirements for determining the drivable area; based on the lateral information of the traffic lane lines in the first subset of traffic lane lines, a second subset of traffic lane lines is determined, wherein the lateral information is used to represent the positional relationship between multiple traffic lane lines in the lateral direction of the traffic corridor, and the second subset of traffic lane lines satisfies the positional relationship requirements for determining the drivable area; based on the second subset of traffic lane lines, the drivable boundary of the vehicle and the two endpoints of the traffic lane lines in the second subset of traffic lane lines are determined; the area enclosed by the drivable boundary and the two endpoints along the lateral perpendicular lines is determined as the drivable area. In other words, in this embodiment, effective traffic lane lines in the time and / or spatial dimensions are filtered based on attribute information, and the drivable boundaries are further filtered based on the lateral information of the aforementioned traffic lane lines. The left and right boundaries of the traffic lane where the vehicle is located are dynamically and adaptively reconstructed, thereby generating a drivable area without relying on lidar or high computing power. This method overcomes the obstacles of related technologies, such as susceptibility to the effects of lighting, shadows, rain, and snow, and the tendency to fail on open roads without physical curbs. Therefore, it solves the technical problem of low accuracy in determining the drivable area of ​​a vehicle, achieving the technical effect of improving the accuracy of determining the drivable area of ​​a vehicle.

[0142] The technical solutions of the embodiments of this application will be illustrated below with reference to preferred embodiments.

[0143] Currently, existing autonomous driving perception systems typically employ two methods to obtain drivable areas: visual semantic segmentation and LiDAR point cloud clustering. Visual semantic segmentation performs pixel-level classification of images to obtain a drivable area mask. This method is sensitive to lighting, shadows, and rain / snow, and requires a large amount of labeled data to train deep networks, demanding high hardware computing power. LiDAR point cloud clustering projects 3D point clouds onto a bird's-eye view and then performs clustering to extract curb or obstacle boundaries. This method requires high-beam LiDAR, is costly, and is prone to failure on open roads without physical curbs (e.g., construction zones, rural roads). Furthermore, some mass-produced L2+ (Level 2+ Autonomous Driving System) autonomous driving vehicles attempt to directly widen and generate virtual road surfaces using lane line equations output by the perception module.

[0144] In view of the defects in the related art, such as jitter in perception results, fixed boundaries, and failure in scenarios without physical curbs, the present application proposes a method for dynamically generating soft curbs based on historical statistics of lane line equations, which can achieve the following without increasing the cost of additional sensors: smooth boundaries without jitter; the widening distance adapts to the curvature; and it can be compatible with open roads without physical curbs.

[0145] In an embodiment of the present application, a method for generating the boundary of a road surface area is provided, which runs on an in-vehicle computing unit and includes: obtaining a single-frame lane line data set. The above data set at least includes the cubic parameter equation of the traffic channel line (for example, ), the starting longitudinal coordinate xs and the ending longitudinal coordinate xe, the semantic type T of the traffic channel line ∈ {ordinary, curb}, and the unique ID.

[0146] Optionally, perform time-domain filtering on the above data set, including: if the duration of continuous appearance of the same unique ID < t1, then discard it; if a new unique ID appears for the first time, it can be written into the historical cache; if a certain unique ID is lost, delete the corresponding historical record.

[0147] Optionally, only retain the lane lines that satisfy xs ≤ –S and xe ≥ E, where S and E are obtained by dynamically looking up a table according to the current speed of the vehicle itself to ensure the rendering effect before and after the vehicle itself.

[0148] Optionally, for the retained lane lines, sample N points at equal intervals longitudinally and calculate the abscissa of each point for the sum as the lateral position score Score of the lane line.

[0149] Optionally, mark the one with the largest Score as the leftmost boundary L1, and the one with the smallest Score as the rightmost boundary R1, and record its equation intercept (L1), (R1).

[0150] Optionally, perform a moving average on (L1), (R1) of the nearest K frames respectively to obtain the smoothed left and right soft curb reference offsets osl and osr.

[0151] Optionally, calculate the longitudinal length Range = xe – xs of each lane line; if the semantic type is "curb", then Range′ = Range · β, β > 1; otherwise Range′ = Range.

[0152] Optionally, sort the left and right sides respectively according to Range′, and take the ones with the largest Range′ on the left and right sides as the left longest reference line L_max and the right longest reference line R_max respectively.

[0153] Optionally, the intercept of the equation for L_max can be replaced with osl, and the intercept of the equation for R_max can be replaced with osr to generate new left boundary equation L_boundary and right boundary equation R_boundary.

[0154] Optionally, L_boundary, R_boundary, = , = The closed area enclosed by the four curves is the drivable road surface area in this frame, which is output to the human-machine interface (HMI) for display.

[0155] Optionally, compared with related technologies, this application has at least the advantages of eliminating the need for LiDAR, reusing existing Advanced Driver Assistance Systems Cameras (ADAS cameras) or domain controllers to achieve zero additional hardware costs, and suppressing single-frame noise and reducing data jitter through time and space filtering.

[0156] Optionally, when identifying curbs, curb data can be used as much as possible to make the driving area closer to the real road surface. However, when the curb recognition effect is poor and the curb is very short, it can be replaced by the traffic lane line with better recognition effect in the middle. The widening distance is adaptively generated by statistical value osl or osr, and when there is a jump in the traffic lane line of perception recognition, it can be smoother and the display effect is better.

[0157] Optionally, to make the purpose, technical solution and advantages of this application clearer, the parameters, formulas and lookup table relationships in the embodiments of this application are all real vehicle values ​​and have repeatability; if alternative devices or platforms are used, they can be scaled up proportionally and still fall within the protection scope of this application.

[0158] Optionally, a single-frame lane line dataset can be obtained.

[0159] Optionally, to suppress false detections in a single frame, this application embodiment employs a "three-stage" finite state machine, including: when id first appears, writing it into the circular buffer History[id], and recording the time of the first frame. ; The number of consecutive frames is ≥ ( =10 (corresponding to 1000 milliseconds) is allowed to participate in subsequent spatial filtering; if a frame is lost, it is directly deleted from the History because the position of the vehicle is constantly changing and the previous traffic lane is no longer available.

[0160] Optionally, to ensure that the rendering area always covers the driver's most visible front and rear sections, and to avoid long lane lines outside the area of ​​interest that would prevent the subsequently drawn road area from covering the vehicle's area, if... <–S or If >E, then discard the aforementioned traffic corridor lines.

[0161] Optionally, to balance smoothness and response speed, this embodiment employs a circular queue and an arithmetic mean. The queue a0list has a length K = 10 (corresponding to 1000 milliseconds, maintaining consistency with the lane line time-domain filtering value). When a new frame arrives, the head pointer is incremented by 1, overwriting the oldest data. The offset value can be represented by the following formula.

[0162]

[0163] Here, offset can be used to represent the offset value. This can be used to represent the intercept of the cubic parametric equation corresponding to the reference line. K can be used to represent the preset maximum capacity of the moving average buffer queue. n can be used to represent the actual number of valid frames in the current buffer queue. The above calculation of the average value needs to take into account the actual number. When the queue is not full at the beginning, it needs to be calculated according to the actual length. The left and right offset values ​​osl and osr are calculated according to the formula. It can be used to represent a buffer queue. n can be used to represent the actual number of valid frames in the buffer queue.

[0164] Optionally, open roads often have no guardrails or road markings. In such cases, the perception module can still identify asphalt and grass edges as curb types. To encourage the algorithm to prioritize the aforementioned asphalt and grass edges, this embodiment multiplies the longitudinal length of the curb type by β=1.2, and the ordinary lane line by β=1.0; using the formula Range′=(x_end–x_start). β is the weighted length of all lane lines.

[0165] Optionally, a max-heap can be built on the left and right sides respectively, with the top of the heap being the one with the largest Range′. The top of the left heap is denoted as... The top of the pile on the right is If one side of the pile is empty, then L1 or R1 is reused; if both sides are empty, meaning there are no traffic lane lines, then subsequent operations are prevented. The reason for handling the left and right sides separately is because the road surface widens before the on-ramps and off-ramps, and the trajectories of the left and right lane lines differ significantly, requiring separate handling.

[0166] Optionally, generating new boundary equations can replace , The equation intercept is obtained by keeping the other coefficients unchanged, resulting in the equations for the left and right curves.

[0167] Optionally, the obtained two lane line equations and the corresponding range of x values ​​are […]. , The data is sent to the downstream module. The display module can draw a closed region based on the range of the two equations, and then fade the edges to achieve a visually pleasing road surface effect.

[0168] Optionally, the output of this application is only used for HMI display, which is an enhancement of the display effect under the condition of limited hardware capabilities, and does not directly control the vehicle; this application strictly relies on the lane line data of the perception module. When the recognition accuracy of perception is poor, the generated road surface area will also have a large error. However, under normal circumstances, it has a certain filtering capability for perception jitter.

[0169] The methods of the embodiments of this application will be further illustrated below.

[0170] Figure 3 This is a schematic diagram of a generation architecture for real-time reconstruction of drivable road surface areas based on single-frame lane line equations, according to an embodiment of this application. Figure 3 As shown, the architecture defining the drivable area of ​​the vehicle includes: a camera, an intelligent driving domain (including camera services, perception modules, human-machine interaction services, and Ethernet communication services), and a cockpit domain, including the QNX software system (QNX Software Systems, hereinafter referred to as QNX) server process and the human-machine interaction service. Specifically, the camera sends image data to the intelligent driving domain controller via a private CAN bus at a frequency of 10 Hz; the camera service is responsible for receiving data from the camera and publishing messages subscribed to by downstream modules based on the software architecture's message subscription and distribution mechanism; the perception module is responsible for sensor data reception, time synchronization, coordinate transformation, image recognition, and fusion; the human-machine interaction service is responsible for receiving all data required for display and performing post-processing; the Ethernet communication service is responsible for cross-domain network communication, and in this application, it is the module used for data communication with the cockpit domain; the QNX server process is mainly used for communication with the intelligent driving domain and then forwards the data to other modules.

[0171] Figure 4 This is a schematic diagram of a coordinate system according to an embodiment of this application, such as... Figure 4 As shown, the origin of this coordinate system is at the center of the rear axle, x is forward, y is to the left, and z is upward; the lane line equations are unified as cubic parametric curves. The "vertical" refers to the x-direction, and the "horizontal" refers to the y-direction; the unit of length is meter; the unit of time is second; and the unit of angle is radian.

[0172] Figure 5 This is a flowchart illustrating a method for generating a drivable road surface region in real time based on a single-frame lane line equation, according to an embodiment of this application. Figure 5As shown, the method for determining the drivable area of ​​the vehicle may include the following steps.

[0173] Step S502: Obtain a single-frame lane line dataset.

[0174] Optionally, multiple lane line data output from the current frame are received from the perception module. Each lane line contains its cubic polynomial equation, longitudinal start point, end point, semantic type (ordinary lane line or curb), and unique ID.

[0175] Step S504: Perform temporal filtering on the single-frame lane line dataset.

[0176] Optionally, if a lane line ID appears for less than 10 consecutive frames (i.e., less than 1 second), it is discarded; if it is a new ID, it is written to the history cache; if an ID is lost for more than 1 consecutive frame, it is deleted from the history cache.

[0177] Step S506, spatial filtering.

[0178] Optionally, the effective rendering range [-S, E] before and after the current vehicle speed is obtained by looking up the table, and only lane lines that meet the preset conditions are retained to ensure that the generated drivable area covers the driver's main gaze area.

[0179] Step S508: Vertical sampling to calculate the horizontal position.

[0180] Optionally, for each retained lane line, 10 points are sampled at equal intervals within the interval, the lateral coordinates of each sampled point are calculated, and all y_i are summed to obtain the lateral position score of the lane line.

[0181] Step S510: Sort horizontally to obtain the leftmost and rightmost boundaries and obtain their intercepts.

[0182] Optionally, the score values ​​of all lane lines are sorted, the one with the largest score is taken as the leftmost boundary L_1, and the one with the smallest score is taken as the rightmost boundary R_1, and the equation intercept is recorded. If only one lane line is detected, the output of this frame is skipped.

[0183] Step S512: Obtain the final left and right intercepts using a moving average.

[0184] Optionally, the a_0(L_1) and a_0(R_1) of the most recent 10 frames are stored in a circular queue of length 10, and the average value of the valid values ​​in the current queue is calculated to obtain the smoothed left intercept offset and right intercept offset.

[0185] Step S514: Calculate the weighted vertical length.

[0186] Optionally, for each lane line, calculate the longitudinal length; if the semantic type is "roadside", multiply the length by the weighting coefficient, otherwise leave it unchanged to obtain the weighted length.

[0187] Step S516: Select the longest reference line.

[0188] Optionally, the lane lines on the left and right sides are sorted from largest to smallest, and the longest one is taken as the left reference line and the right reference line. If there is no valid lane line on a certain side, L_1 or R_1 selected in step S510 is reused.

[0189] Step S518: Generate new boundary equations.

[0190] Optionally, the intercepts of the left and right reference lines are replaced with osl and osr, respectively, while the other coefficients remain unchanged, to obtain the left boundary equation.

[0191] Step S520: Output to the downstream display module.

[0192] Optionally, the generated left and right boundary equations and their corresponding longitudinal ranges are sent to the display module; a closed region is drawn and its edges are faded, and it is displayed on the vehicle screen as a drivable region.

[0193] In this embodiment, a set of traffic lane lines in the traffic corridor where the vehicle is traveling is obtained; based on the attribute information of the traffic lane lines in the set, a first subset of traffic lane lines is determined, wherein the attribute information is used to represent the state of the traffic lane lines in the time dimension and / or spatial dimension, and the first subset of traffic lane lines satisfies the time dimension and / or spatial dimension requirements for determining the drivable area; based on the lateral information of the traffic lane lines in the first subset of traffic lane lines, a second subset of traffic lane lines is determined, wherein the lateral information is used to represent the positional relationship between multiple traffic lane lines in the lateral direction of the traffic corridor, and the second subset of traffic lane lines satisfies the positional relationship requirements for determining the drivable area; based on the second subset of traffic lane lines, the drivable boundary of the vehicle and the two endpoints of the traffic lane lines in the second subset of traffic lane lines are determined; the area enclosed by the drivable boundary and the two endpoints along the lateral perpendicular lines is determined as the drivable area. In other words, in this embodiment, effective traffic lane lines in the time and / or spatial dimensions are filtered based on attribute information, and the drivable boundaries are further filtered based on the lateral information of the aforementioned traffic lane lines. The left and right boundaries of the traffic lane where the vehicle is located are dynamically and adaptively reconstructed, thereby generating a drivable area without relying on lidar or high computing power. This method overcomes the obstacles of related technologies, such as susceptibility to the effects of lighting, shadows, rain, and snow, and the tendency to fail on open roads without physical curbs. Therefore, it solves the technical problem of low accuracy in determining the drivable area of ​​a vehicle, achieving the technical effect of improving the accuracy of determining the drivable area of ​​a vehicle.

[0194] Figure 6 This is a schematic diagram of a vehicle drivable area determination device according to an embodiment of this application. Figure 6 As shown, the vehicle's drivable area determination device 60 includes: an acquisition module 602, used to acquire the traffic lane line set of the traffic lane in which the vehicle is traveling.

[0195] The first determining module 604 is used to determine a first subset of traffic corridors based on the attribute information of traffic corridors in the traffic corridor set. The attribute information is used to represent the state of the traffic corridors in the time dimension and / or spatial dimension. The first subset of traffic corridors satisfies the time dimension requirement and / or spatial dimension requirement for determining the drivable area.

[0196] The second determining module 606 is used to determine a second subset of traffic channels based on the lateral information of the traffic channels in the first subset of traffic channels. The lateral information is used to represent the positional relationship between multiple traffic channels in the lateral direction of the traffic channel. The second subset of traffic channels satisfies the positional relationship requirement for determining the drivable area.

[0197] The third determining module 608 is used to determine the drivable boundary of the vehicle and the two endpoints of the traffic channel lines in the second traffic channel line subset based on the second traffic channel line subset.

[0198] The fourth determining module 610 is used to determine the drivable area as the area enclosed by the drivable boundary and the two endpoints along the horizontal straight lines.

[0199] In this embodiment, the acquisition module 602 acquires the set of traffic lane lines in the traffic lane where the vehicle is traveling. The first determination module 604 determines a first subset of traffic lane lines based on the attribute information of the traffic lane lines in the set, wherein the attribute information represents the state of the traffic lane lines in the time and / or spatial dimensions, and the first subset of traffic lane lines satisfies the time and / or spatial dimension requirements for determining the drivable area. The second determination module 606 determines a second subset of traffic lane lines based on the lateral information of the traffic lane lines in the first subset of traffic lane lines, wherein the lateral information represents the positional relationship between multiple traffic lane lines in the lateral direction of the traffic lane, and the second subset of traffic lane lines satisfies the positional relationship requirements for determining the drivable area. The third determination module 608 determines the drivable boundary of the vehicle and the two endpoints of the traffic lane lines in the second subset of traffic lane lines based on the second subset of traffic lane lines. The fourth determination module 610 determines the area enclosed by the drivable boundary and the two endpoints along lateral straight lines as the drivable area.

[0200] According to another aspect of the embodiments of this application, a vehicle is also provided, including: a memory storing an executable program; and a processor for running the program, wherein the program executes the methods in various embodiments of this application when it runs.

[0201] According to another aspect of the embodiments of this application, an electronic device is also provided, including: a memory storing an executable program; and a processor for running the program, wherein the program runs the methods of various embodiments of this application.

[0202] According to another aspect of the embodiments of this application, a computer-readable storage medium is also provided, the computer-readable storage medium including a stored executable program, wherein, when the executable program is running, it controls the device where the computer-readable storage medium is located to perform the methods of various embodiments of this application.

[0203] Embodiments of this application also provide a computer program product, including a computer program that, when executed by a processor, implements the methods of various embodiments of this application.

[0204] Embodiments of this application also provide a computer program product, including a non-volatile computer-readable storage medium for storing a computer program that, when executed by a processor, implements the methods in various embodiments of this application.

[0205] Embodiments of this application also provide a computer program that, when executed by a processor, implements the methods described in the various embodiments of this application.

[0206] In the above embodiments of this application, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0207] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units can be a logical functional division, and in actual implementation, there may be other division methods. For instance, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units or modules may be electrical or other forms.

[0208] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0209] Furthermore, the functional units in the various embodiments of this application 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.

[0210] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0211] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for determining the drivable area of ​​a vehicle, characterized in that, include: Obtain the traffic lane set of the traffic lanes in which the vehicle is traveling; Based on the attribute information of the traffic corridors in the traffic corridor set, a first subset of traffic corridors is determined, wherein the attribute information is used to represent the state of the traffic corridors in the time dimension and / or spatial dimension, and the first subset of traffic corridors satisfies the time dimension requirement and / or spatial dimension requirement for determining the drivable area. Based on the lateral information of the traffic corridor lines in the first traffic corridor line subset, a second traffic corridor line subset is determined, wherein the lateral information is used to represent the positional relationship between multiple traffic corridor lines in the lateral direction of the traffic corridor, and the second traffic corridor line subset satisfies the positional relationship requirement for determining the drivable area; Based on the second subset of traffic lane lines, the drivable boundary of the vehicle and the two endpoints of the traffic lane lines in the second subset of traffic lane lines are determined. The drivable area is defined as the area enclosed by the drivable boundary and the two endpoints along the lateral straight lines.

2. The method according to claim 1, characterized in that, Based on the attribute information of the traffic corridors in the traffic corridor set, a first subset of traffic corridors is determined, including: The traffic corridor lines are grouped together, and those whose attribute information satisfies the time-domain conditions are determined as elements of the first subset of traffic corridor lines, wherein the time-domain conditions represent the accuracy requirements that the traffic corridor lines need to meet in the time dimension; and / or, The traffic corridors that satisfy the spatial conditions in the first traffic corridor set are identified as elements in the first traffic corridor line subset, wherein the spatial conditions represent the accuracy requirements that the traffic corridor lines need to meet in the spatial dimension.

3. The method according to claim 2, characterized in that, The attribute information includes the identification information of the traffic corridor lines. Traffic corridor lines whose attribute information satisfies the time-domain condition are grouped together and determined as the first subset of traffic corridor lines, including at least one of the following: In response to the traffic corridor lines of the same identification information, if the duration of consecutive occurrences is greater than or equal to the duration threshold, it is determined that the identification information satisfies the time domain condition, and the traffic corridor lines of the identification information are determined as elements in the first traffic corridor line subset; In response to the traffic corridor line where target identification information appears, it is determined that the identification information satisfies the time domain condition, and the traffic corridor line with the target identification information is determined as an element in the first traffic corridor line subset, wherein the target identification information is identification information that has not appeared in historical times before the current time; In response to the fact that the identification information is not lost, it is determined that the identification information satisfies the time domain condition, and the traffic corridor line of the identification information is determined as an element in the first traffic corridor line subset; And / or, The attribute information includes first endpoint information and second endpoint information of the traffic corridor line. The first endpoint information indicates the location of the first endpoint of the traffic corridor line, and the second endpoint information indicates the location of the second endpoint of the traffic corridor line. The distance between the second endpoint and the vehicle is greater than the distance between the first endpoint and the vehicle. The traffic corridor lines whose attribute information satisfies the spatial conditions are determined as elements in the first traffic corridor line subset, including: In response to the fact that the first endpoint information of the traffic corridor is less than or equal to the first attention distance and the second endpoint information is greater than or equal to the second attention distance, it is determined that the first endpoint information and the second endpoint information satisfy the spatial condition, wherein the first attention distance is less than the second attention distance; The traffic corridors that satisfy the spatial conditions in terms of their first endpoint information and second endpoint information are grouped together and determined as elements in the first traffic corridor line subset.

4. The method according to claim 1, characterized in that, The second subset of traffic corridor lines includes the first traffic corridor line and the second traffic corridor line. Based on the lateral information of the traffic corridor lines in the first subset of traffic corridor lines, the second subset of traffic corridor lines is determined, including: The traffic corridor lines are grouped together, and the traffic corridor line with the largest lateral information is determined as the first traffic corridor line; and the traffic corridor lines are grouped together, and the traffic corridor line with the smallest lateral information is determined as the second traffic corridor line. And / or, Based on the second subset of traffic corridor lines, the drivable boundaries of the vehicle are determined, including: Determine the target longitudinal length of the traffic corridor line in the second traffic corridor sub-group; Based on the target longitudinal length, determine the reference line set corresponding to the second traffic corridor line subset; The drivable boundary is determined based on the reference line set.

5. The method according to claim 4, characterized in that, The attribute information includes first endpoint information, second endpoint information, and type information of the traffic corridor line. The first endpoint information indicates the location of the first endpoint of the traffic corridor line, the second endpoint information indicates the location of the second endpoint of the traffic corridor line, and the distance between the second endpoint and the vehicle is greater than the distance between the first endpoint and the vehicle. The type information indicates whether there are physical obstacles in the traffic corridor line. Determining the target longitudinal length of the traffic corridor lines in the second traffic corridor line subset includes: Based on the difference between the first endpoint information and the second endpoint information, the initial longitudinal length of the traffic corridor line in the second traffic corridor line subset is determined; Based on the type information, the initial longitudinal length is weighted to obtain the target longitudinal length.

6. The method according to claim 5, characterized in that, Based on the target longitudinal length, determine the reference line set corresponding to the second traffic corridor line subset, including: The traffic corridor line with the largest target longitudinal length is selected as an element in the reference line set from the second traffic corridor line subset. And / or, Determining the drivable boundary based on the reference line set includes: Obtain the equation intercepts corresponding to the traffic corridor lines in the second traffic corridor line subset; The intercept of the equation is smoothed to obtain the smoothed intercept of the equation, and the smoothed intercept of the equation is determined as the reference offset of the second traffic corridor line subset. The intercept of the reference line equation of the reference line in the reference line set is replaced with the reference offset, and the replaced reference line equation is determined as the boundary equation of the drivable boundary.

7. The method according to claim 6, characterized in that, The drivable boundary includes a first drivable boundary and a second drivable boundary. The endpoints include a first endpoint and a second endpoint. The distance between the second endpoint and the vehicle is greater than the distance between the first endpoint and the vehicle. The drivable area is defined as the region enclosed by the drivable boundary and the two endpoints along the lateral straight lines, including: The area enclosed by the boundary equation of the first drivable boundary, the boundary equation of the second drivable boundary, the straight line equation of the first endpoint along the lateral direction, and the straight line equation of the second endpoint along the lateral direction is defined as the drivable area.

8. The method according to any one of claims 1 to 7, characterized in that, The method further includes: Along the longitudinal direction of the traffic corridor, the traffic corridor lines in the first traffic corridor line subset are sampled to obtain multiple sampling points on the traffic corridor lines; Based on the location information of multiple sampling points along the traffic corridor, the lateral information of the traffic corridor is determined.

9. The method according to claim 8, characterized in that, The location information includes lateral location information. Based on the location information of multiple sampling points along the traffic corridor, the lateral information of the traffic corridor is determined, including: The sum of the lateral position information of multiple sampling points on the traffic corridor line is determined as the lateral information of the traffic corridor line.

10. A vehicle, characterized in that, include: Memory, which stores executable programs; A processor for running the program, wherein the program, when running, performs the method according to any one of claims 1 to 9.

Images