Lane line detection method and device for non-standard road scene and electronic equipment

The roadside sensors obtain information on driving vehicles and screen and line up the vehicles, detecting lane lines on non-standard roads, solving the problem of difficulty in lane lines detection in non-standard road scenarios by traditional methods and improving the adaptability of the autonomous driving system.

CN118968435APending Publication Date: 2024-11-15ZHIDAO NETWORK TECH (BEIJING) CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411108846.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-13
Publication Date
2024-11-15

AI Technical Summary

Technical Problem

In non-standard road scenarios, traditional lane line detection methods cannot effectively identify lane lines, resulting in difficulty in driving and making decisions in complex road environments.

Method used

The roadside sensors obtain detection information of the driving vehicles, filter out side-by-side vehicles, and use side-by-side vehicle constraints to detect lane lines, so as to realize lane lines detection of the travelable road without relying on lane line markings.

Benefits of technology

Accurate lane line detection results are provided in non-standard road scenarios, providing data support for path planning and decision-making of autonomous driving systems, and improving adaptability to complex road environments.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118968435A_ABST
    Figure CN118968435A_ABST
Patent Text Reader

Abstract

The invention discloses a road detection method and device for a non-standard road scene and electronic equipment, the method is executed by roadside equipment, and the method comprises the following steps: obtaining vehicle detection information of a plurality of driving vehicles corresponding to each drivable road in a monitoring area through a roadside sensor; acquiring side-by-side driving vehicles corresponding to each drivable road according to the vehicle detection information of the plurality of driving vehicles; and lane line detection is carried out according to the side-by-side driving vehicles corresponding to each drivable road, and a lane line detection result corresponding to each drivable road is obtained. According to the technical scheme, the lane line detection result meeting the actual driving condition can be constructed without depending on traditional lane line marks, and data support is provided for track planning and decision making of an automatic driving system.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of intelligent transportation technology, and in particular to a lane line detection method, device and electronic equipment for non-standard road scenes. Background Art

[0002] Autonomous driving technology is a technology that uses computers to achieve assisted driving or unmanned driving. It relies on sensor systems such as visible light cameras, millimeter wave radars, lidars, inertial navigation systems, global positioning systems, etc., so that computers can partially or completely replace human drivers to automatically and safely operate vehicles.

[0003] At present, autonomous driving technology is mainly used in standard road scenarios. Standard roads are roads with specific markings, signs and other information as stipulated by road traffic laws and regulations. In standard road scenarios, autonomous driving technology can identify standardized road markings, signs and other environmental perception information through the above-mentioned sensor system. However, in non-standard road scenarios such as roads with unclear lane lines or roads without lane lines, sensors such as visible light cameras, millimeter wave radars and lidars cannot collect lane line markings. Traditional lane line detection solutions based on lane line markings cannot obtain reliable and accurate detection results, which may pose challenges to the driving and decision-making of the autonomous driving system. Summary of the invention

[0004] The embodiments of the present application provide a lane line detection method, device and electronic device for non-standard road scenarios to accurately identify the lane lines of each drivable road in non-standard road scenarios, thereby providing support for the path planning of an autonomous driving system.

[0005] The present application embodiment adopts the following technical solutions:

[0006] In a first aspect, an embodiment of the present application provides a lane line detection method for a non-standard road scene, the lane line detection method for a non-standard road scene being performed by a roadside device, the method comprising:

[0007] Acquire vehicle detection information of multiple vehicles corresponding to each drivable road in the monitoring area through roadside sensors;

[0008] Acquire the parallel-traveling vehicles corresponding to each drivable road according to the vehicle detection information of the plurality of traveling vehicles;

[0009] Lane line detection is performed according to the vehicles traveling side by side corresponding to each drivable road, and the lane line detection result corresponding to each drivable road is obtained.

[0010] In a second aspect, an embodiment of the present application further provides a lane line detection device for a non-standard road scene, wherein the lane line detection device for a non-standard road scene is applied to a roadside device, and the device comprises:

[0011] A data acquisition unit, used to acquire vehicle detection information of a plurality of traveling vehicles corresponding to each drivable road in a monitoring area through a roadside sensor;

[0012] A parallel detection unit, used for acquiring the parallel vehicles corresponding to each drivable road according to the vehicle detection information of the plurality of traveling vehicles;

[0013] The lane line detection unit is used to perform lane line detection according to the parallel vehicles corresponding to each drivable road, and obtain the lane line detection result corresponding to each drivable road.

[0014] In a third aspect, an embodiment of the present application further provides an electronic device, including:

[0015] A processor; and a memory arranged to store computer executable instructions, which, when executed, cause the processor to perform a lane line detection method for a non-standard road scene.

[0016] In a fourth aspect, an embodiment of the present application also provides a computer-readable storage medium, which stores one or more programs, and when the one or more programs are executed by an electronic device including multiple application programs, the electronic device performs a lane line detection method for non-standard road scenes.

[0017] At least one of the above technical solutions adopted in the embodiments of the present application can achieve the following beneficial effects:

[0018] The roadside equipment of the embodiment of the present application obtains vehicle detection information of multiple driving vehicles corresponding to each drivable road through roadside sensors, and selects vehicles driving side by side from multiple driving vehicles based on the vehicle detection information. Since vehicles driving side by side can constrain the range of each lane of each drivable road, the lane lines of each drivable road can be detected using vehicles driving side by side. The embodiment of the present application can construct lane line detection results that conform to actual driving conditions for drivable roads in non-standard road scenarios without relying on lane line markings, provide data support for trajectory planning and decision-making of the autonomous driving system, and improve the adaptability of the autonomous driving system to complex road environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0020] Figure 1 A schematic diagram of a process flow of a lane line detection method for a non-standard road scene in an embodiment of the present application;

[0021] Figure 2 This is a schematic diagram of the principle of vehicle target detection based on the CenterPoint algorithm in an embodiment of the present application;

[0022] Figure 3 A road schematic diagram of an intersection area in an embodiment of the present application;

[0023] Figure 4 This is a schematic diagram of a point cloud detection result in an embodiment of the present application;

[0024] Figure 5 This is a schematic diagram of lane line division in an embodiment of the present application;

[0025] Figure 6 This is a schematic diagram of the structure of a lane line detection device for a non-standard road scene in an embodiment of the present application;

[0026] Figure 7 This is a schematic diagram of the structure of an electronic device in an embodiment of the present application. DETAILED DESCRIPTION

[0027] In order to make the purpose, technical solution and advantages of the present application clearer, the technical solution of the present application will be clearly and completely described below in combination with the specific embodiments of the present application and the corresponding drawings. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present application.

[0028] The technical solutions provided by various embodiments of the present application are described in detail below in conjunction with the accompanying drawings.

[0029] The embodiment of the present application provides a lane line detection method for a non-standard road scene. The following embodiment of the present application uses the roadside device as the execution subject of the lane line detection method for a non-standard road scene for example. It can be understood that the execution subject of the embodiment of the present application can also be other devices or systems, such as a roadside unit, a roadside edge computing system or a chip system, and the chip system includes at least one processor, and a communication interface or a memory, wherein the memory is used to store programs or instructions, and the communication interface is used to provide the processor with information or data input and output functions. The execution subject of the embodiment of the present application can also be that the roadside sends the acquired information to the cloud and executes it by the cloud. Alternatively, the execution subject of the lane line detection method for a non-standard road scene in the embodiment of the present application can also be software or hardware, and different execution steps can be centralized or distributed.

[0030] To facilitate understanding of the following embodiments of the present application, the technical terms involved in the embodiments of the present application are first explained.

[0031] A standard road scene refers to a standardized road that complies with relevant laws and regulations on road traffic and has specific markings, signs and other information. In a standard road scene, the autonomous driving system can obtain accurate road information through standardized road markings, signs and other information.

[0032] Non-standard road scenes are relative to standard road scenes. In non-standard road scenes, the lane lines on the road surface are unclear or not set, and the lane line data on the road surface cannot be collected by roadside sensors such as visible light cameras, millimeter wave radars and lidars.

[0033] The intersection area, also known as the crossroad area, includes, for example, a crossroad area, a T-junction area, etc. Each drivable road in the intersection area usually has a vehicle stop line. Figure 3 Lane markings will be laid in the road area within the vehicle stop line, while lane markings will generally not be laid in the intersection area outside the vehicle stop line.

[0034] Vehicles traveling in the same row refer to vehicles that are roughly in the same row in the transverse direction of the road. The lateral spacing between adjacent vehicles in the same row may not be uniform; Figure 4 As shown, there are three rows of vehicles on the oncoming road at the upper left corner of the figure, the first row of vehicles includes 5 vehicles (CarA1, CarA2, CarA3, CarA4 and CarA5), the second row of vehicles includes 3 vehicles (CarA1, CarB1, CarB2), and the third row of vehicles includes 2 vehicles (CarC1, CarC2), wherein the lateral spacing distribution of adjacent vehicles in the first row of vehicles is approximately equally spaced, while the adjacent vehicles in the second and third rows of vehicles present unequal spacing distribution.

[0035] It should be noted that when a certain moving vehicle is located in two rows at the same time in the longitudinal direction, for example Figure 4 The vehicle CarA1 shown can be divided into different ranks as shown in this embodiment. In other embodiments, the vehicle can also be divided into one rank and cannot be divided into other ranks. For example, when CarA1 is divided into the first row, CarA1 should no longer be included in the second row.

[0036] Parallel vehicles are a special arrangement of vehicles in the same row. The lateral spacing between adjacent vehicles in each group of parallel vehicles is approximately equal and should not be too large. The vehicles at both ends of the parallel vehicles are also located at the edge of the drivable road where they are located. Figure 4 As shown, the first row of vehicles on the incoming road (CarA1, CarA2, CarA3, CarA4 and CarA5) are side-by-side vehicles, and the first row of vehicles on the outgoing road (CarD1, CarD2, CarD3, CarD4 and CarD5) are also side-by-side vehicles. Although the second row of vehicles on the outgoing road (CarE1, CarE2 and CarE3) conform to the equally spaced distribution, the right-end vehicle CarE3 is not located at the right road edge of the outgoing road, and therefore does not belong to the side-by-side vehicles specified in the embodiment of the present application.

[0037] Edge-driving vehicles refer to vehicles that travel on the edge of a drivable road, including edge-driving vehicles on both sides of the road. For ease of description, this embodiment is abbreviated as left-side edge-driving vehicles and right-side edge-driving vehicles.

[0038] In order to enable the autonomous driving system to better adapt to non-standard road scenarios, the embodiments of the present application are intended to identify the lane lines of each drivable road without relying on traditional lane line markings, thereby providing support for the adaptability of the autonomous driving system in non-standard road scenarios.

[0039] like Figure 1 As shown, a flow chart of a lane line detection method for a non-standard road scene in an embodiment of the present application is provided, and the method at least includes the following steps S110 to S130:

[0040] Step S110: Acquire vehicle detection information of multiple driving vehicles corresponding to each drivable road in the monitoring area through roadside sensors.

[0041] When lane line detection is required for non-standard road scenes, the roadside equipment of the embodiment of the present application first collects roadside environmental data through roadside sensors, such as collecting point clouds of the monitoring area through laser radar, and processes the point clouds according to algorithms such as semantic segmentation and neural networks, so as to obtain vehicle detection information of each drivable road in the monitoring area and multiple vehicles on each drivable road. The vehicle detection information includes vehicle position, heading angle, vehicle width, detection frame information (detection frame information includes the center point of the detection frame and the length, width and height of the detection frame), etc., wherein the vehicle position includes the lateral position and the longitudinal position.

[0042] In actual application scenarios, roadside equipment can determine whether it is necessary to perform lane line detection on the monitoring area of ​​non-standard road scenes based on set strategies. Set strategies include, for example, periodic detection, specific event-triggered detection, etc. Periodic detection can be understood as performing lane line detection on the monitoring area according to a set detection cycle. Specific events include, for example, detection instructions received by the roadside equipment, and events such as the roadside equipment judging that the current lane line accuracy of the monitoring area is insufficient.

[0043] Step S120: acquiring the parallel-traveling vehicles corresponding to each drivable road according to the vehicle detection information of the plurality of traveling vehicles.

[0044] In non-standard road scenarios, the lane lines on the drivable road are blurred and cannot be sensed by the sensor. However, in actual applications, the lane lines on the road surface may be sensed by the eyes of the driver of the vehicle, so the vehicle on the drivable road will still drive on the road according to the constraints of the lane lines; or even if the lane lines of the current section cannot be sensed by the eyes of the driver of the vehicle, the driver will refer to the lane lines of the adjacent sections that have been traveled or the driving mode of the preceding vehicle based on the vehicle driving experience to drive on the current road.

[0045] It can be seen that the distribution of actual vehicles on the drivable road is related to the lanes of the drivable road to a certain extent. When the drivable road is a straight road or a road approximately straight with a large curvature radius, the side-by-side vehicles on the drivable road identify the range of each lane of the drivable road to a certain extent. Based on this, this embodiment obtains the side-by-side vehicles corresponding to each drivable road, and then detects the lane line of each drivable road based on the side-by-side vehicles.

[0046] Step S130, performing lane line detection according to the parallel vehicles corresponding to each drivable road, and obtaining the lane line detection result corresponding to each drivable road.

[0047] In non-standard road scenarios, after the roadside equipment obtains the lane line detection results, it can also broadcast the lane line detection results so that nearby vehicles can plan their paths based on the lane line detection results, which helps to improve the driving performance of the autonomous driving system.

[0048] like Figure 1It can be seen from the lane line detection method for non-standard road scenes that the roadside equipment of the embodiment of the present application obtains the vehicle detection information of multiple driving vehicles corresponding to each drivable road through the roadside sensor, and selects the side-by-side driving vehicles from the multiple driving vehicles according to the vehicle detection information. Since the side-by-side driving vehicles can constrain the range of each lane of each drivable road, the lane line of each drivable road can be detected by using the side-by-side driving vehicles. Without relying on lane line markings, the embodiment of the present application constructs lane line detection results that conform to actual driving conditions for drivable roads in non-standard road scenes, provides data support for trajectory planning and decision-making of the autonomous driving system, and improves the adaptability of the autonomous driving system to complex road environments.

[0049] Since laser radar has the advantages of high resolution, strong anti-active interference ability, and good low-altitude detection performance, in some embodiments of the present application, multi-frame point clouds of the monitoring area are obtained by laser radar, and vehicle detection information of multiple moving vehicles corresponding to each drivable road is obtained through the multi-frame point clouds, so that the multiple moving vehicles corresponding to each drivable road include at least one row of vehicles traveling side by side.

[0050] Accordingly, in the above step S110, the vehicle detection information of multiple driving vehicles corresponding to each drivable road in the monitoring area is obtained through the roadside sensor, specifically including:

[0051] The target detection is performed on multiple frames of point cloud respectively through a preset target detection model to obtain vehicle detection information of the moving vehicle corresponding to each frame of point cloud;

[0052] The moving vehicles corresponding to the multiple frame point clouds are synthesized into the same frame point cloud according to the vehicle detection information, and the vehicle detection information of the multiple moving vehicles corresponding to each drivable road is obtained through the synthesized point cloud.

[0053] The preset target detection model of this embodiment is, for example, a vehicle target detection model built based on the deep learning algorithm CenterPoint. The CenterPoint algorithm is a framework of a 3D target detection and tracking algorithm for laser point clouds. The algorithm does not use bounding boxes but key points to represent, detect and track 3D targets.

[0054] like Figure 2 As shown, in the detection process, in the first stage, a key point detector is used to detect the center point of the target, and then the center point features are used to regress the target's 3D size, direction, speed and other attributes. In the second stage, these attributes are optimized based on the additional point features of the target, and finally a more accurate detection frame is obtained, which includes the center point (x, y, z) and the length, width and height (L, W, H) of the vehicle. Since the lane line is on the ground plane, z can be set to 0, that is, the virtual lane line information of the embodiment of the present application represents two-dimensional plane information.

[0055] It should be understood that the embodiments of the present application may also perform vehicle detection through other target detection algorithms, and the embodiments of the present application do not specifically limit this. In addition, the embodiments of the present application illustrate an implementation scheme for obtaining vehicle detection information through a laser radar; in other embodiments of the present application, vehicle detection information of multiple traveling vehicles may also be obtained through a camera; or vehicle detection information of multiple traveling vehicles may also be obtained based on fusion information of a camera and a laser radar.

[0056] In this embodiment, when synthesizing point clouds, any frame point cloud in multiple frame point clouds can be used as a synthesized point cloud, and the driving vehicles in other frame point clouds can be added to the synthesized point cloud. During the synthesis process, if the driving vehicles in other frame point clouds overlap with the driving vehicles in the synthesized point cloud, the driving vehicles in the other frame point clouds will not be added to the synthesized point cloud; if the driving vehicles in other frame point clouds do not overlap with any driving vehicles in the synthesized point cloud, the driving vehicles in the other frame point clouds will be added to the synthesized point cloud. There is no overlap between multiple driving vehicles in the synthesized point cloud.

[0057] When obtaining vehicle detection information of a plurality of driving vehicles corresponding to each drivable road through the synthetic point cloud, this embodiment further includes:

[0058] Acquiring information about drivable roads within the monitoring area, wherein the drivable road information includes the range of each drivable road;

[0059] According to the vehicle detection information of each traveling vehicle in the synthetic point cloud and the range of each drivable road, the traveling vehicles belonging to each drivable road in the synthetic point cloud are determined.

[0060] by Figure 4 Taking the scenario shown in the figure as an example, the solution of this embodiment can determine that the vehicles corresponding to the incoming road include CarA1~CarA5, CarB1, Car B2, CarC1 and Car C2, and the vehicles corresponding to the outgoing road include CarD1~CarD5 and CarE1~CarE3, and Figure 4 CarX1, CarX2 and CarX3 in the figure are neither incoming roads nor outgoing roads.

[0061] In this way, all the moving vehicles in the synthetic point cloud can be grouped according to the drivable roads to obtain multiple moving vehicles belonging to each drivable road, which can avoid using the moving vehicles in the intersection area to fit the lane lines and improve the lane line detection accuracy.

[0062] It should be noted that the present embodiment shows an implementation scheme for obtaining vehicle detection information of multiple driving vehicles corresponding to each drivable road based on multi-frame point clouds. In other embodiments, vehicle detection information of multiple driving vehicles corresponding to each drivable road can also be obtained based on a single-frame point cloud.

[0063] In some embodiments of the present application, the step S120 above acquires the parallel-traveling vehicles corresponding to each drivable road according to the vehicle detection information of the plurality of traveling vehicles, specifically including:

[0064] Performing a ranking detection on the plurality of traveling vehicles according to the vehicle positions to obtain a plurality of rows of traveling vehicles;

[0065] According to the vehicle positions, vehicles traveling in parallel that meet a vehicle parallel distribution condition are obtained from among the multiple rows of vehicles traveling in parallel.

[0066] When obtaining multiple rows of moving vehicles, this embodiment also filters the moving vehicles according to the number of vehicles included in each row of moving vehicles. If one or several rows of moving vehicles include only one vehicle, such moving vehicles are deleted; if all rows of moving vehicles include only one vehicle, it means that the current drivable road includes one lane. At this time, the lane line shape point pairs on both sides of the moving vehicle can be obtained according to the vehicle detection information, and the lane line fitting is performed according to the lane line shape point pairs.

[0067] In some possible implementation schemes of this embodiment, performing a ranking detection on the multiple moving vehicles according to the vehicle positions to obtain multiple rows of moving vehicles specifically includes:

[0068] Acquire a ranking detection primitive, wherein the ranking detection primitive includes a ranking detection window or a ranking detection line, wherein the ranking detection window has a height greater than a drivable road width and a width greater than a drivable vehicle width, and the ranking detection line is parallel to a width direction of the drivable road;

[0069] Acquire, according to the vehicle position, a positional relationship between the plurality of traveling vehicles and the ranking detection primitive when the ranking detection primitive is at each sliding position;

[0070] A plurality of rows of traveling vehicles are obtained according to the positional relationship.

[0071] This implementation scheme can construct a ranking detection window or a ranking detection line according to each drivable road information. For example, the width and height information of the ranking detection window can be designed according to the drivable road width and the width of the traveling vehicle, so that the height of the ranking detection window is greater than the drivable road width, and the width of the ranking detection window is greater than the traveling vehicle width. The traveling vehicle width here can be the maximum vehicle width among multiple traveling vehicles, that is, the width of the ranking detection window is greater than the maximum vehicle width. The ranking detection line is designed according to the width direction of the drivable road, so that the ranking detection line is parallel to the width direction of the drivable road. In this way, the ranking detection of multiple traveling vehicles can be completed by sliding the ranking detection window or ranking detection line.

[0072] When sliding the ranking detection primitive, the sliding position of the ranking detection primitive should be determined. In this embodiment, the sliding position of the ranking detection primitive can be determined according to the longitudinal positions of multiple traveling vehicles. For example, multiple traveling vehicles are arranged in a longitudinal queue along the road direction of a drivable road, and the head position of the queue is used as the starting position of the ranking detection primitive. The second sliding position is determined according to the sorting result of the longitudinal queue, and the sliding positions of the ranking detection primitives are determined accordingly.

[0073] For example, Figure 4 The longitudinal position corresponding to the first row of oncoming roads is used as the starting position of the ranking detection primitive, the longitudinal position corresponding to the second row of future roads is used as the second sliding position of the ranking detection primitive, and the longitudinal position corresponding to the third row of future roads is used as the third sliding position of the ranking detection primitive.

[0074] In the process of sliding the above-mentioned row detection primitive, when a row of vehicles traveling side by side is detected, the detection can be stopped. Of course, the detection can also be stopped after all vehicles traveling side by side are detected.

[0075] In this embodiment, the positional relationship between the plurality of vehicles and the ranking detection primitives includes the vehicles intersecting the ranking detection line or the vehicles within the ranking detection window. Figure 4 Taking the scenario shown as an example, when the ranking detection line slides to the first row position, the vehicles CarA1~CarA5 on the oncoming road intersect with the ranking detection line at the same time; when the ranking detection window slides to the first row position, the vehicles CarA1~CarA5 on the oncoming road are also located in the ranking detection window at the same time. Correspondingly, when the ranking detection line slides to the second row position, the vehicles CarA1, CarB1 and CarB2 on the oncoming road intersect with the ranking detection line at the same time; when the ranking detection window slides to the second row position, the vehicles CarA1, CarB1 and CarB2 on the oncoming road are also located in the ranking detection window at the same time.

[0076] It should be noted that the intersection of the moving vehicle and the ranking detection line in this embodiment can be understood as the existence of an intersection between the detection frame of the moving vehicle and the ranking detection line, and the absence of an intersection between the detection frame and the ranking detection line, but the closest longitudinal distance between the detection frame and the ranking detection line is within the threshold range (for example, the closest longitudinal distance is less than half the length of the side of the detection frame).

[0077] For example, when the ranking detection line is at Figure 4 When the first row of the destination road is selected, the ranking detection line may not have an intersection with CarD2, but the longitudinal distance of CarD2 relative to the ranking detection line is less than half of the side length of the CarD2 detection frame. At this time, CarD2 can be classified into the first row.

[0078] In this embodiment, the moving vehicle being located within the ranking detection window can be understood as the moving vehicle being completely located within the ranking detection window and the moving vehicle being partially located within the ranking detection window. When the moving vehicle is completely outside the ranking detection window, the moving vehicle does not belong to the current ranking.

[0079] Thus, this embodiment obtains multiple rows of traveling vehicles according to the position relationship, specifically including:

[0080] The running vehicles that intersect the ranking detection line at the same time and / or the running vehicles that are simultaneously within the ranking detection window are determined as the running vehicles in the same row.

[0081] In actual applications, the ranking detection line and the ranking detection window can be combined to perform ranking detection. For example, when the ranking detection line and the ranking detection window slide to the same position, if a moving vehicle that intersects the ranking detection line at the same time is also in the ranking detection window at the same time, it is determined that the moving vehicle is a vehicle in the same row.

[0082] Of course, those skilled in the art may also use the ranking detection line or the ranking detection window alone to perform ranking detection, and the embodiments of the present application do not impose specific limitations on this.

[0083] In some other possible implementation schemes of this embodiment, the vehicle side-by-side distribution conditions include lateral spacing distribution constraints and edge distribution constraints. The lateral spacing distribution constraints are used to determine whether adjacent vehicles in the same row are distributed at equal intervals, and the edge distribution constraints are used to determine whether the vehicles at both ends of the same row are edge vehicles.

[0084] Specifically, the lateral interval distribution constraint includes a difference threshold and an interval threshold, and the edge distribution constraint includes an edge interval threshold; wherein the difference threshold indicates the maximum lateral interval difference corresponding to the equal interval distribution, and the interval threshold indicates the maximum lateral interval distance corresponding to the equal interval distribution; the edge interval threshold indicates the maximum lateral distance between the edge-driving vehicle and the edge of the drivable road, and these three thresholds can be designed based on experience.

[0085] It should be noted that when the vehicles traveling side by side include more than three vehicles, the difference threshold is used to determine whether the vehicles traveling side by side meet the lateral spacing distribution constraint; and when the vehicles traveling side by side include two vehicles, the spacing threshold is used to determine whether the vehicles traveling side by side meet the lateral spacing distribution constraint.

[0086] Accordingly, determining whether the vehicles traveling in the same row meet the vehicle side-by-side distribution condition according to the vehicle positions specifically includes:

[0087] Determining whether the vehicles at both ends of each row of vehicles meet the edge distribution constraint according to the lateral position, and determining whether the lateral intervals between adjacent vehicles in each row of vehicles meet the lateral interval distribution constraint;

[0088] If both the lateral spacing distribution constraint and the edge distribution constraint are satisfied, the row of vehicles is determined to be parallel-running vehicles that satisfy the vehicle parallel distribution condition;

[0089] Otherwise, it is determined that the row of traveling vehicles does not meet the vehicle side-by-side distribution condition.

[0090] In this embodiment, the lateral distances between the two end vehicles in each row of vehicles and the edge of the drivable road are calculated. When the left lateral distance between the left end vehicle and the left edge of the drivable road is less than the edge interval threshold, and the right lateral distance between the right end vehicle and the right edge of the drivable road is less than the edge interval threshold, it is determined that the row of vehicles meets the edge distribution constraint. And according to the number of vehicles in each row of vehicles, when the number of vehicles is more than three, the lateral interval difference between adjacent vehicles in the same row is calculated. When the lateral interval difference is less than the difference threshold, it is determined that the vehicles in the same row meet the lateral interval distribution constraint; when the number of vehicles is two, the lateral interval between adjacent vehicles in the same row is calculated. When the lateral interval is less than the interval threshold, it is determined that the vehicles in the same row meet the lateral interval distribution constraint.

[0091] Continue with Figure 4 Taking the scenario shown as an example, the first row of vehicles on the incoming road and the outgoing road meet both the lateral spacing distribution constraint and the edge distribution constraint. Therefore, the first row of vehicles on the incoming road and the outgoing road are side-by-side vehicles. The second row of vehicles on the incoming road does not meet either the lateral spacing distribution constraint or the edge distribution constraint, so it does not belong to side-by-side vehicles; the third row of vehicles on the incoming road does not meet the edge distribution constraint, so it does not belong to side-by-side vehicles; the second row of vehicles on the outgoing road meets the lateral spacing distribution constraint, but does not meet the edge distribution constraint, so it does not belong to side-by-side vehicles.

[0092] In some embodiments of the present application, the above step S130 performs lane line detection according to the parallel vehicles corresponding to each drivable road to obtain the lane line detection result corresponding to each drivable road, specifically including:

[0093] Acquire lane line shape point pairs on both sides of each vehicle traveling side by side according to the vehicle detection information;

[0094] Lane line fitting is performed according to the lane line shape point pairs, and a lane line detection result corresponding to each drivable road is obtained according to the lane line fitting result.

[0095] In some possible implementations of this embodiment, obtaining lane line shape point pairs on both sides of each vehicle in the parallel driving vehicles according to the vehicle detection information specifically includes:

[0096] Acquire a lateral distance between the detection frames of each vehicle traveling side by side and an adjacent vehicle according to the detection frame information;

[0097] According to the lateral spacing of the detection frames, a pair of lane line shape points on each side is obtained on the extension line of the side width of the detection frame of each moving vehicle.

[0098] like Figure 5 As shown in the figure, assuming that the center point of the detection frame of the moving vehicle car1 is (x1, y1, z1), the length, width and height of the detection frame are (L1, W1, H1), and the center point of the detection frame of the moving vehicle car2 is (x2, y2, z2), the length, width and height of the detection frame are (L2, W2, H2), the horizontal distance d1 between the detection frames of the moving vehicle car1 and the moving vehicle car2 is calculated according to the center point of the detection frame and the length, width and height of the detection frame. The coordinates are (2*x1+2*x2-W1+W2). On the extended lines l1 and l2 of the detection frame width of the moving vehicle car1, the points P1 and P2 that are d1 / 2 away from the left side of the detection frame are calculated as the left lane line shape point pair. The two-dimensional coordinates of P1 are ((2*x1+2*x2-W1+W2) / 4, y1-L1 / 2), and the two-dimensional coordinates of P2 are ((2*x1+2*x2-W1+W2) / 4, y1+L1 / 2). Similarly, on the extended lines l1 and l2 of the detection frame width of the moving vehicle car1, the points that are d1 / 2 away from the right side of the detection frame are calculated as the right lane line shape point pair P3 and P4. At this point, the mathematical expression of the lane line Line1 can be constructed through the left lane line shape point pair P1 and P2. Similarly, the mathematical expression of the right lane line (also known as the right road boundary line Road1) can also be calculated.

[0099] In some other possible implementations of this embodiment, when performing lane line fitting according to the lane line shape point pairs, the method further includes:

[0100] Determining whether each drivable road has a vehicle stop line;

[0101] When there is a vehicle stop line, the vehicle stop line is used as the lane line endpoint, and lane line fitting is performed toward the inside of the vehicle stop line according to the lane line shape point.

[0102] This embodiment can identify the vehicle stop line based on the roadside environment data collected by the roadside sensor, for example, by using semantic segmentation, neural network and other algorithms to identify the vehicle stop line on the point cloud, and determine whether each drivable road has a vehicle stop line based on the identification result. It can also be determined whether each drivable road has a vehicle stop line based on historical vehicle data, and this embodiment does not limit the detection method of the vehicle stop line.

[0103] When it is determined that a drivable road has a vehicle stop line, lane line fitting should be performed on the inside of the vehicle stop line to avoid constructing lane lines in the intersection area and improve the accuracy of lane line fitting.

[0104] In addition, after obtaining the lane line detection result, the embodiment of the present application can also correct the lane line detection result according to the real-time driving trajectory of the vehicle to make the corrected lane line detection result more accurate.

[0105] The embodiment of the present application also provides a lane line detection device 600 for a non-standard road scene, such as Figure 5 As shown, a schematic diagram of the structure of a lane line detection device for a non-standard road scene in an embodiment of the present application is provided, wherein the device 600 includes: a data acquisition unit 610, a side-by-side detection unit 620 and a lane line detection unit 630, wherein:

[0106] The data acquisition unit 610 is used to acquire vehicle detection information of multiple driving vehicles corresponding to each drivable road in the monitoring area through a roadside sensor;

[0107] A parallel detection unit 620, configured to obtain the parallel vehicles corresponding to each drivable road according to the vehicle detection information of the plurality of traveling vehicles;

[0108] The lane line detection unit 630 is used to perform lane line detection according to the parallel vehicles corresponding to each drivable road, and obtain the lane line detection result corresponding to each drivable road.

[0109] In some embodiments of the present application, the vehicle detection information includes the vehicle position, and the side-by-side detection unit 620 includes a position detection module and a side-by-side distribution detection module;

[0110] A ranking detection module, used for performing ranking detection on the plurality of moving vehicles according to the vehicle positions to obtain a plurality of rows of moving vehicles;

[0111] The parallel distribution detection module is used to obtain the parallel running vehicles that meet the vehicle parallel distribution conditions in the multiple rows of running vehicles according to the vehicle positions.

[0112] In some embodiments of the present application, the ranking detection module is specifically used to obtain a ranking detection primitive, the ranking detection primitive including a ranking detection window or a ranking detection line, the ranking detection window having a height greater than a drivable road width and a width greater than a traveling vehicle width, and the ranking detection line being parallel to the width direction of the drivable road; obtaining, according to the vehicle position, a positional relationship between the multiple traveling vehicles and the ranking detection primitive when the ranking detection primitive is at each sliding position; and obtaining the multiple rows of traveling vehicles according to the positional relationship.

[0113] In some embodiments of the present application, the positional relationship includes moving vehicles that intersect the ranking detection line or moving vehicles that are within the ranking detection window, and the ranking detection module is specifically used to determine moving vehicles that simultaneously intersect the ranking detection line and / or moving vehicles that are simultaneously within the ranking detection window as vehicles in the same row.

[0114] In some embodiments of the present application, the vehicle position includes a lateral position and a longitudinal position, the vehicle side-by-side distribution condition includes a lateral spacing distribution constraint and an edge distribution constraint, and the side-by-side distribution detection module is specifically used to determine whether the vehicles at both ends of each row of vehicles satisfy the edge distribution constraint based on the lateral position, and to determine whether the lateral spacing between adjacent vehicles in each row of vehicles satisfies the lateral spacing distribution constraint; if both the lateral spacing distribution constraint and the edge distribution constraint are satisfied at the same time, the row of vehicles is determined to be side-by-side vehicles that satisfy the vehicle side-by-side distribution condition.

[0115] In some embodiments of the present application, the lane line detection unit 630 is specifically used to obtain lane line shape point pairs on both sides of each vehicle traveling side by side according to the vehicle detection information; perform lane line fitting according to the lane line shape point pairs, and obtain lane line detection results corresponding to each drivable road according to the lane line fitting results.

[0116] In some embodiments of the present application, the vehicle detection information includes vehicle position and detection frame information, and the lane line detection unit 630 is specifically used to obtain the lateral spacing of the detection frames between each vehicle traveling side by side and the adjacent vehicle according to the detection frame information; and obtain the lane line shape point pairs on each side of the extension line of the detection frame side width of each vehicle according to the lateral spacing of the detection frames.

[0117] In some embodiments of the present application, the lane line detection unit 630 is also used to determine whether each drivable road has a vehicle stop line; when there is a vehicle stop line, the vehicle stop line is used as the lane line endpoint, and the lane line is fitted toward the inside of the vehicle stop line according to the lane line shape point.

[0118] It can be understood that the above-mentioned lane line detection device for non-standard road scenes can implement the various steps of the lane line detection method for non-standard road scenes provided in the aforementioned embodiments, and the relevant explanations about the lane line detection method for non-standard road scenes are applicable to the lane line detection device for non-standard road scenes, which will not be repeated here.

[0119] Figure 7 This is a schematic diagram of the structure of an electronic device according to an embodiment of the present application. Figure 7 At the hardware level, the electronic device includes a processor, and optionally also includes an internal bus, a network interface, and a memory. The memory may include a memory, such as a high-speed random access memory (RAM), and may also include a non-volatile memory (non-volatile memory), such as at least one disk storage. Of course, the electronic device may also include hardware required for other services.

[0120] The processor, network interface and memory can be interconnected through an internal bus, which can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 7 Only one bidirectional arrow is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0121] The memory is used to store the program. Specifically, the program may include a program code, and the program code includes a computer operation instruction. The memory may include a memory and a non-volatile memory, and provides instructions and data to the processor.

[0122] The processor reads the corresponding computer program from the non-volatile memory into the memory and then runs it, forming a lane line detection device for non-standard road scenes at the logical level. The processor executes the program stored in the memory and is specifically used to perform the following operations:

[0123] Acquire vehicle detection information of multiple vehicles corresponding to each drivable road in the monitoring area through roadside sensors;

[0124] Acquire the parallel-traveling vehicles corresponding to each drivable road according to the vehicle detection information of the plurality of traveling vehicles;

[0125] Lane line detection is performed according to the vehicles traveling side by side corresponding to each drivable road, and the lane line detection result corresponding to each drivable road is obtained.

[0126] The above application Figure 1 The method performed by the lane line detection device for non-standard road scenes disclosed in the illustrated embodiment can be applied to a processor or implemented by a processor. The processor may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by an integrated logic circuit of hardware in the processor or an instruction in the form of software. The above processor may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it may also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The methods, steps and logic block diagrams disclosed in the embodiments of the present application can be implemented or executed. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The steps of the method disclosed in conjunction with the embodiments of the present application can be directly embodied as being executed by a hardware decoding processor, or executed by a combination of hardware and software modules in a decoding processor. The software module can be located in a random access memory, flash memory, read-only memory, programmable read-only memory or electrically erasable programmable memory, register or other mature storage media in the art. The storage medium is located in the memory, and the processor reads the information in the memory and completes the steps of the lane line detection method for non-standard road scenes in combination with its hardware.

[0127] The electronic device may also perform Figure 1 A method for executing a lane line detection device for a non-standard road scene in a vehicle lane detection system Figure 1 The functions of the illustrated embodiment will not be described in detail in the embodiments of the present application.

[0128] The present application also provides a computer-readable storage medium, which stores one or more programs, wherein the one or more programs include instructions, which, when executed by an electronic device including multiple application programs, enable the electronic device to execute Figure 1 The method performed by the lane line detection device for non-standard road scenes in the illustrated embodiment is specifically used to perform:

[0129] Acquire vehicle detection information of multiple vehicles corresponding to each drivable road in the monitoring area through roadside sensors;

[0130] Acquire the parallel-traveling vehicles corresponding to each drivable road according to the vehicle detection information of the plurality of traveling vehicles;

[0131] Lane line detection is performed according to the vehicles traveling side by side corresponding to each drivable road, and the lane line detection result corresponding to each drivable road is obtained.

[0132] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0133] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0134] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0135] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0136] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0137] The memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0138] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0139] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0140] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0141] The above is only an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.

Claims

1. A lane detection method for non-standard road scenes, It is characterized in that The lane line detection method for the non-standard road scene is performed by a roadside device, and the method includes: Acquire vehicle detection information of multiple vehicles corresponding to each drivable road in the monitoring area through roadside sensors; Acquire the parallel-traveling vehicles corresponding to each drivable road according to the vehicle detection information of the plurality of traveling vehicles; Lane line detection is performed according to the vehicles traveling side by side corresponding to each drivable road, and the lane line detection result corresponding to each drivable road is obtained.

2. The lane line detection method for non-standard road scenes according to claim 1, It is characterized in that The vehicle detection information includes vehicle positions, and obtaining the parallel-traveling vehicles corresponding to each drivable road according to the vehicle detection information of the plurality of traveling vehicles includes: Performing a ranking detection on the plurality of traveling vehicles according to the vehicle positions to obtain a plurality of rows of traveling vehicles; According to the vehicle positions, vehicles traveling in parallel that meet a vehicle parallel distribution condition are obtained from among the multiple rows of vehicles traveling in parallel.

3. The lane line detection method for non-standard road scenes according to claim 2, It is characterized in that The step of performing ranking detection on the plurality of traveling vehicles according to the vehicle positions to obtain a plurality of rows of traveling vehicles comprises: Acquire a ranking detection primitive, wherein the ranking detection primitive includes a ranking detection window or a ranking detection line, wherein the ranking detection window has a height greater than a drivable road width and a width greater than a drivable vehicle width, and the ranking detection line is parallel to a width direction of the drivable road; Acquire, according to the vehicle position, a positional relationship between the plurality of traveling vehicles and the ranking detection primitive when the ranking detection primitive is at each sliding position; The multiple rows of traveling vehicles are acquired according to the positional relationship.

4. The lane line detection method for non-standard road scenes according to claim 3, It is characterized in that The position relationship includes the running vehicles intersecting the ranking detection line or the running vehicles in the ranking detection window, and the acquiring the multiple rows of running vehicles according to the position relationship includes: The running vehicles that intersect the ranking detection line at the same time and / or the running vehicles that are simultaneously within the ranking detection window are determined as the running vehicles in the same row.

5. The lane line detection method for non-standard road scenes according to claim 2, It is characterized in that The vehicle position includes a lateral position and a longitudinal position, the vehicle side-by-side distribution condition includes a lateral spacing distribution constraint and an edge distribution constraint, and the step of obtaining the side-by-side running vehicles that meet the vehicle side-by-side distribution condition in the multiple rows of running vehicles according to the vehicle position includes: Determining whether the vehicles at both ends of each row of vehicles meet the edge distribution constraint according to the lateral position, and determining whether the lateral intervals between adjacent vehicles in each row of vehicles meet the lateral interval distribution constraint; If both the lateral spacing distribution constraint and the edge distribution constraint are satisfied at the same time, the row of traveling vehicles is determined to be parallel traveling vehicles that satisfy the vehicle parallel distribution condition.

6. The lane line detection method for non-standard road scenes according to claim 1, It is characterized in that The performing lane line detection according to the parallel vehicles corresponding to each drivable road to obtain the lane line detection result corresponding to each drivable road includes: Acquire lane line shape point pairs on both sides of each vehicle traveling side by side according to the vehicle detection information; Lane line fitting is performed according to the lane line shape point pairs, and a lane line detection result corresponding to each drivable road is obtained according to the lane line fitting result.

7. The lane line detection method for non-standard road scenes according to claim 6, It is characterized in that The vehicle detection information includes vehicle position and detection frame information, and the step of obtaining lane line shape point pairs on both sides of each vehicle in the parallel driving vehicles according to the vehicle detection information includes: Acquire a lateral distance between the detection frames of each vehicle traveling side by side and an adjacent vehicle according to the detection frame information; According to the lateral spacing of the detection frames, a pair of lane line shape points on each side is obtained on the extension line of the side width of the detection frame of each moving vehicle.

8. The lane line detection method for non-standard road scenes according to claim 6, It is characterized in that When performing lane line fitting according to the lane line shape point pairs, the method further includes: Determining whether each drivable road has a vehicle stop line; When there is a vehicle stop line, the vehicle stop line is used as the lane line endpoint, and lane line fitting is performed toward the inside of the vehicle stop line according to the lane line shape point.

9. A lane line detection device for non-standard road scenes, It is characterized in that The lane line detection device for non-standard road scenes is applied to roadside equipment, and the device includes: A data acquisition unit, used to acquire vehicle detection information of a plurality of traveling vehicles corresponding to each drivable road in a monitoring area through a roadside sensor; A parallel detection unit, used for acquiring the parallel running vehicles corresponding to each drivable road according to the vehicle detection information of the plurality of running vehicles; The lane line detection unit is used to perform lane line detection according to the parallel vehicles corresponding to each drivable road, and obtain the lane line detection result corresponding to each drivable road.

10. An electronic device, include: processor; as well as A memory arranged to store computer executable instructions, wherein when the executable instructions are executed, the processor executes the lane line detection method for a non-standard road scene according to any one of claims 1 to 8.