Method and device for searching driving lane of unmanned vehicle, storage medium and equipment

Abstract

The application discloses a driving lane searching method and device of an unmanned vehicle, a storage medium and equipment, and belongs to the technical field of automatic driving. The method comprises the following steps: acquiring lane line information of a plurality of lanes from a high-definition map; generating an influence domain of each lane according to the lane line information of each lane, wherein the influence domain corresponds to a coordinate range of two lane lines of the lane; searching for a relevant lane intersecting with the unmanned vehicle according to the influence domain; and searching for a current driving lane of the unmanned vehicle from the relevant lane by using a ray method. According to the application, the relevant influence domain in which the unmanned vehicle is located is searched from each influence domain, and the relevant lane corresponding to the relevant influence domain is determined, so that the thousands of lanes can be coarsely screened, and the relevant relevant lane can be quickly obtained. Then, the driving lane of the unmanned vehicle is searched from the relevant lane by using the ray method, the relevant relevant lane can be finely screened, and the searching speed and searching precision of the lane are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of automatic driving, in particular to a driving lane searching method and device for an unmanned vehicle, a storage medium and equipment. BACKGROUND

[0002] In the driving process of an unmanned vehicle, the driving lane of the unmanned vehicle needs to be searched first, and then the path of the unmanned vehicle is planned based on the driving lane. However, there are thousands of lanes in a high-definition map, which greatly affects the searching efficiency of the driving lane of the unmanned vehicle. SUMMARY

[0003] The present application provides a driving lane searching method and device for an unmanned vehicle, a storage medium and equipment, which are used to solve the problem that there are thousands of lanes in a high-definition map, which greatly affects the searching efficiency of the driving lane of the unmanned vehicle. The technical solution is as follows:

[0004] In one aspect, a driving lane searching method for an unmanned vehicle is provided, and the method comprises the following steps:

[0005] Obtaining lane line information of a plurality of lanes from a high-definition map;

[0006] Generating an influence domain of each lane according to the lane line information of each lane, wherein the influence domain corresponds to a coordinate range of two lane lines of the lane;

[0007] Searching for a relevant influence domain in which the unmanned vehicle currently locates from the influence domains, and determining a relevant lane corresponding to the relevant influence domain;

[0008] Searching for a driving lane of the unmanned vehicle from the relevant lane by using a ray method.

[0009] In one possible implementation, the searching for the driving lane of the unmanned vehicle from the relevant lane by using the ray method comprises the following steps:

[0010] Obtaining coordinate information of the unmanned vehicle;

[0011] For each relevant lane, connecting adjacent reference points in the relevant lane, connecting each two reference points to form a reference edge, so that each lane line has a plurality of reference edges, and two lane lines are closed to finally form a closed polygon having a plurality of reference edges; screening effective reference edges from the plurality of reference edges according to the coordinate information, wherein the effective reference edges are within a range through which a ray with the unmanned vehicle as a starting point can pass; determining a relative position relationship between the unmanned vehicle and the effective reference edges by using the ray method, and determining the driving lane of the unmanned vehicle according to the relative position relationship.

[0012] In a possible implementation, the filtering of the effective reference edge from the plurality of reference edges according to the coordinate information comprises the following steps.

[0013] Obtaining an x-axis coordinate value and a y-axis coordinate value in the coordinate information;

[0014] Obtaining an x-axis minimum coordinate value, an x-axis maximum coordinate value, a y-axis minimum coordinate value and a y-axis maximum coordinate value of each reference edge;

[0015] If the direction of the ray is horizontal right, filtering, from the plurality of reference edges, a reference edge with the x-axis maximum coordinate value greater than or equal to the x-axis coordinate value, the y-axis maximum coordinate value greater than or equal to the y-axis coordinate value and the y-axis minimum coordinate value less than or equal to the y-axis coordinate value as the effective reference edge;

[0016] If the direction of the ray is horizontal left, filtering, from the plurality of reference edges, a reference edge with the x-axis minimum coordinate value less than or equal to the x-axis coordinate value, the y-axis maximum coordinate value greater than or equal to the y-axis coordinate value and the y-axis minimum coordinate value less than or equal to the y-axis coordinate value as the effective reference edge;

[0017] If the direction of the ray is vertical up, filtering, from the plurality of reference edges, a reference edge with the y-axis maximum coordinate value greater than or equal to the y-axis coordinate value, the x-axis maximum coordinate value greater than or equal to the x-axis coordinate value and the x-axis minimum coordinate value less than or equal to the x-axis coordinate value as the effective reference edge;

[0018] If the direction of the ray is vertical down, filtering, from the plurality of reference edges, a reference edge with the y-axis minimum coordinate value less than or equal to the y-axis coordinate value, the x-axis maximum coordinate value greater than or equal to the x-axis coordinate value and the x-axis minimum coordinate value less than or equal to the x-axis coordinate value as the effective reference edge.

[0019] In a possible implementation, before the step of searching for the driving lane of the unmanned vehicle from the relevant lanes by using the ray method, the method further comprises the following steps.

[0020] Obtaining a driving direction of the unmanned vehicle;

[0021] Searching for a relevant lane with the lane direction consistent with the driving direction from the relevant lanes.

[0022] In a possible implementation, the step of searching for a relevant lane with the lane direction consistent with the driving direction from the relevant lanes comprises the following steps.

[0023] For each relevant lane, determining a tangent direction of a lane center line point closest to the unmanned vehicle as the lane direction of the relevant lane;

[0024] Calculating an included angle between the driving direction and the lane direction.

[0025] If the absolute value of the included angle is less than or equal to 90 degrees, it is determined that the lane direction is consistent with the driving direction, and the relevant lane is retained.

[0026] If the absolute value of the included angle is greater than 90 degrees, it is determined that the lane direction is inconsistent with the driving direction, and the relevant lane is excluded.

[0027] In a possible implementation, when the lane line information includes coordinate information of reference points in lane lines, the generating, by the processor, of an influence domain for each lane according to lane line information of the lane includes:

[0028] For each lane, selecting, from the coordinate information of the reference points of the lane, an x-axis minimum coordinate value, an x-axis maximum coordinate value, a y-axis minimum coordinate value, and a y-axis maximum coordinate value;

[0029] Generating an influence domain of a predetermined shape, wherein an x-axis coordinate value of any point in the influence domain is greater than or equal to the x-axis minimum coordinate value and less than or equal to the x-axis maximum coordinate value, and a y-axis coordinate value of any point in the influence domain is greater than or equal to the y-axis minimum coordinate value and less than or equal to the y-axis maximum coordinate value.

[0030] In a possible implementation, the finding, by the processor, of a relevant influence domain in which the unmanned vehicle is currently located from the influence domains includes:

[0031] Obtaining an x-axis coordinate value and a y-axis coordinate value of the unmanned vehicle;

[0032] Obtaining an x-axis minimum coordinate value, an x-axis maximum coordinate value, a y-axis minimum coordinate value, and a y-axis maximum coordinate value of each influence domain;

[0033] From the influence domains, screening an influence domain in which an x-axis maximum coordinate value is greater than or equal to the x-axis coordinate value, an x-axis minimum coordinate value is less than or equal to the x-axis coordinate value, a y-axis maximum coordinate value is greater than or equal to the y-axis coordinate value, and a y-axis minimum coordinate value is less than or equal to the y-axis coordinate value as a relevant influence domain.

[0034] In an aspect, a driving lane finding device for an unmanned vehicle is provided, and the device includes:

[0035] An obtaining module, configured to obtain lane line information of a plurality of lanes from a high-definition map;

[0036] A generating module, configured to generate an influence domain for each lane according to lane line information of the lane, the influence domain corresponding to a coordinate range corresponding to two lane lines of the lane;

[0037] The searching module is configured to search, from each influence domain, a relevant influence domain in which the unmanned vehicle currently locates, and determine a relevant lane corresponding to the relevant influence domain.

[0038] The searching module is further configured to search, from the relevant lane, a driving lane of the unmanned vehicle by using a ray method.

[0039] In one aspect, a computer readable storage medium is provided, and the storage medium stores at least one instruction, which is loaded and executed by a processor to implement the method for searching the driving lane of the unmanned vehicle.

[0040] In one aspect, a perception device is provided, and the perception device comprises a processor and a memory, and the memory stores at least one instruction, which is loaded and executed by the processor to implement the method for searching the driving lane of the unmanned vehicle.

[0041] The technical scheme provided by the present application has at least the following beneficial effects:

[0042] First, the influence domain of each lane in the high-definition map is generated; then, a relevant influence domain in which the unmanned vehicle currently locates is searched from each influence domain, and a relevant lane corresponding to the relevant influence domain is determined, which can perform coarse-grained screening on thousands of lanes to quickly obtain the relevant relevant lane; finally, the driving lane of the unmanned vehicle is searched from the relevant lane by using the ray method, which can perform fine-grained screening on the relevant lane, thereby improving the searching speed and accuracy of the lane.

[0043] When the driving lane of the unmanned vehicle is searched from the relevant lane by using the ray method, the adjacent reference points in the relevant lane are connected first, and the connecting line between each two reference points forms a reference edge, so that each lane line has multiple reference edges, and two lane lines are closed to finally form a closed polygon with multiple reference edges; then, the effective reference edges are screened from the multiple reference edges according to the coordinate information of the unmanned vehicle, which can quickly screen the effective reference edges located in the range that can be penetrated by the ray from the multiple reference edges, thereby reducing the reference edges that need to be screened by using the ray method and greatly reducing the calculation amount.

[0044] Before the driving lane of the unmanned vehicle is searched from the relevant lane by using the ray method, the relevant lane with the lane direction consistent with the driving direction of the unmanned vehicle can be searched from each relevant lane, which can eliminate the relevant lane with the lane direction opposite to the driving direction, thereby reducing the relevant lane that needs to be screened by using the ray method and further reducing the calculation amount. BRIEF DESCRIPTION OF DRAWINGS

[0045] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the embodiments description will be briefly introduced as follows. Obviously, the drawings in the following description only constitute some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained from these drawings without creative labor.

[0046] Figure 1 is a method flowchart of a driving lane searching method of an unmanned vehicle provided by an embodiment of the present application;

[0047] Figure 2 is a method flowchart of a driving lane searching method of an unmanned vehicle provided by another embodiment of the present application;

[0048] Figure 3 is a schematic diagram of an influence domain provided by another embodiment of the present application;

[0049] Figure 4 is a structural schematic diagram of a driving direction and a lane direction provided by another embodiment of the present application;

[0050] Figure 5 is a schematic diagram of a ray method provided by another embodiment of the present application;

[0051] Figure 6 is an effect diagram of a driving lane searching method provided by another embodiment of the present application;

[0052] Figure 7 is a structural block diagram of a driving lane searching device of an unmanned vehicle provided by still another embodiment of the present application. DETAILED DESCRIPTION

[0053] In order to make the purpose, technical solutions and advantages of the embodiments of the present application more clear, the embodiments of the present application will be further described in detail below with reference to the drawings.

[0054] Please refer to Figure 1 which shows a method flowchart of a driving lane searching method of an unmanned vehicle provided by an embodiment of the present application. The driving lane searching method of the unmanned vehicle can be applied to a perception device of the unmanned vehicle. The driving lane searching method of the unmanned vehicle can include:

[0055] In step 101, lane line information of a plurality of lanes is acquired from a high-definition map.

[0056] There are thousands of lanes drawn in the high-definition map, and each lane is composed of two lane lines.

[0057] The lane line information can include coordinate information of a plurality of reference points collected in the lane line, each lane has left and right lane lines, and each lane line is composed of a group of lines of road points (x, y), wherein the reference points are sparse at straight sections and dense at curved sections.

[0058] In step 102, an influence domain of each lane is generated according to the lane line information of each lane, and the influence domain corresponds to a coordinate range corresponding to the two lane lines of the lane.

[0059] The perception device can determine the lane according to the lane line information and then generate the influence domain containing the lane. The influence domain can be of any shape.

[0060] In step 103, a relevant influence domain in which the unmanned vehicle is currently located is found from the influence domains, and a relevant lane corresponding to the relevant influence domain is determined.

[0061] The perception device can obtain the position of the unmanned vehicle by using the sensor in the unmanned vehicle, and then find the influence domain in which the position is located. If the finding result is that the unmanned vehicle is not in any influence domain, it is determined that the unmanned vehicle is not in the high-definition map, and the search fails. If the finding result is that the unmanned vehicle is in one influence domain, the lane corresponding to the influence domain is determined as the driving lane of the unmanned vehicle, and the search succeeds. If the finding result is that the unmanned vehicle is in at least two influence domains, the at least two influence domains are determined as relevant influence domains, and the lane corresponding to each relevant influence domain is determined as a relevant lane, and step 104 is performed to find the driving lane of the unmanned vehicle.

[0062] In step 104, the driving lane of the unmanned vehicle is found from the relevant lanes by using a ray method.

[0063] The ray method refers to using a horizontal ray or a vertical ray to determine whether a point is in a polygon region.

[0064] In this embodiment, the perception device can generate a polygon region for each relevant lane based on the reference points, and then determine whether the unmanned vehicle is in the polygon region by using the ray method. If the determination result is that the unmanned vehicle is not in any polygon region, it is determined that the unmanned vehicle is not in the high-definition map, and the search fails. If the determination result is that the unmanned vehicle is in one polygon region, the relevant lane corresponding to the polygon region is determined as the driving lane of the unmanned vehicle, and the search succeeds. If the determination result is that the unmanned vehicle is in at least two polygon regions, the lane center lines corresponding to the at least two polygon regions are calculated, the lane center line closest to the unmanned vehicle is found, and the relevant lane corresponding to the lane center line is determined as the driving lane of the unmanned vehicle.

[0065] In summary, the driving lane searching method of the unmanned vehicle provided by the embodiment of the application first generates the influence domain of each lane in the high-definition map; then searches for the relevant influence domain where the unmanned vehicle is currently located from each influence domain, and determines the relevant lane corresponding to the relevant influence domain, which can perform coarse-grained screening on thousands of lanes to quickly obtain the relevant lane; finally, the ray method is used to search for the driving lane of the unmanned vehicle from the relevant lane, which can perform fine-grained screening on the relevant lane, thereby improving the searching speed and accuracy of the lane.

[0066] Reference is made to Figure 2 which shows the method flowchart of the driving lane searching method of the unmanned vehicle provided by an embodiment of the application. The driving lane searching method of the unmanned vehicle can be applied to the perception device of the unmanned vehicle. The driving lane searching method of the unmanned vehicle can include:

[0067] Step 201: obtaining lane line information of a plurality of lanes from a high-definition map.

[0068] There are thousands of lanes in the high-definition map, and each lane is composed of two lane lines.

[0069] The lane line information can include coordinate information of a plurality of reference points collected in the lane line. Each lane has left and right lane lines, and each lane line is composed of a group of connected road points (x, y). The reference points are relatively sparse at straight lanes, and the reference points are relatively dense at curved lanes.

[0070] Step 202: generating an influence domain of each lane according to the lane line information of each lane. The influence domain corresponds to the coordinate range of the two lane lines of the lane.

[0071] The perception device can determine the lane according to the lane line information, and then generate the influence domain containing the lane. The influence domain can be of any shape.

[0072] Specifically, for each lane, the minimum x-axis coordinate value, the maximum x-axis coordinate value, the minimum y-axis coordinate value, and the maximum y-axis coordinate value are selected from the coordinate information of each reference point of the lane; an influence domain of a predetermined shape is generated, wherein the x-axis coordinate value of any point in the influence domain is greater than or equal to the minimum x-axis coordinate value and less than or equal to the maximum x-axis coordinate value, and the y-axis coordinate value of any point in the influence domain is greater than or equal to the minimum y-axis coordinate value and less than or equal to the maximum y-axis coordinate value. The predetermined shape can be a rectangle, a square, a circle, a polygon, etc., which is not limited in the embodiment.

[0073] Figure 3Taking a rectangle as an example, a schematic diagram of the influence domain is shown in FIG. 1, wherein x_min represents the minimum coordinate value of the x-axis, y_min represents the minimum coordinate value of the y-axis, x_max represents the maximum coordinate value of the x-axis, and y_max represents the maximum coordinate value of the y-axis. Any point (x, y) in the influence domain satisfies x_min≤x≤x_max and y_min≤y≤y_max.

[0074] In step 203, the relevant influence domain in which the ego vehicle currently locates is searched from the influence domains, and a relevant lane corresponding to the relevant influence domain is determined.

[0075] Specifically, searching the relevant influence domain in which the ego vehicle currently locates from the influence domains can include the following sub-steps:

[0076] (1) Obtain the x-axis coordinate value and the y-axis coordinate value of the ego vehicle.

[0077] The perception device can obtain the x-axis coordinate value and the y-axis coordinate value of the ego vehicle by using the sensors in the ego vehicle.

[0078] (2) Obtain the minimum x-axis coordinate value, the maximum x-axis coordinate value, the minimum y-axis coordinate value, and the maximum y-axis coordinate value of each influence domain.

[0079] For example, an influence domain is represented by x_min, x_max, y_min, and y_max shown in FIG. 1. Figure 3

[0080] (3) From the influence domains, select the influence domain in which the maximum x-axis coordinate value is greater than or equal to the x-axis coordinate value, the minimum x-axis coordinate value is less than or equal to the x-axis coordinate value, the maximum y-axis coordinate value is greater than or equal to the y-axis coordinate value, and the minimum y-axis coordinate value is less than or equal to the y-axis coordinate value as the relevant influence domain.

[0081] Suppose the coordinate value of the ego vehicle is (x_ego, y_ego). When x_min≤x_ego≤x_max and y_min≤y_ego≤y_max, it is determined that the ego vehicle is in the influence domain. When x_ego

[0082] If the search result is that the ego vehicle is not in any influence domain, it is determined that the ego vehicle is not in the high-definition map, and the search fails. If the search result is that the ego vehicle is in an influence domain, the lane corresponding to the influence domain is determined as the driving lane of the ego vehicle, and the search succeeds. If the search result is that the ego vehicle is in at least two influence domains, the at least two influence domains are determined as the relevant influence domains, and the lane corresponding to each relevant influence domain is determined as a relevant lane. Step 204 is performed to search for the driving lane of the ego vehicle. ​

[0083] Step 204, obtaining the driving direction of the unmanned vehicle, and searching for the relevant lane whose lane direction is consistent with the driving direction from each relevant lane.

[0084] The perception device can obtain the driving direction of the unmanned vehicle by using the sensors in the unmanned vehicle. Then, for each relevant lane, the perception device determines the tangent direction of the lane center line point closest to the unmanned vehicle as the lane direction of the relevant lane; calculates the included angle between the driving direction and the lane direction; if the absolute value of the included angle is less than or equal to 90 degrees, it is determined that the lane direction is consistent with the driving direction, and the relevant lane is retained; if the absolute value of the included angle is greater than 90 degrees, it is determined that the lane direction is inconsistent with the driving direction, and the relevant lane is excluded.

[0085] Figure 4 A schematic diagram of the driving direction and the lane direction is shown, in which the solid arrow represents the driving direction of the unmanned vehicle, and the dashed arrow represents the lane direction.

[0086] Those skilled in the art should understand that step 204 is an optional step, and the perception device can perform step 204 to screen the relevant lanes whose lane directions are consistent with the driving direction, so as to reduce the number of relevant lanes searched by the ray method and improve the searching efficiency.

[0087] Step 205, obtaining the coordinate information of the unmanned vehicle.

[0088] The perception device can obtain the coordinate information of the unmanned vehicle by using the sensors in the unmanned vehicle.

[0089] Step 206, for each relevant lane, connecting the adjacent reference points in the relevant lane, and forming a reference edge by connecting each two reference points, so that each lane line has multiple reference edges, and two lane lines are closed to finally form a closed polygon with multiple reference edges; screening effective reference edges from the multiple reference edges according to the coordinate information, the effective reference edges being within the range that the ray can pass through from the unmanned vehicle as the starting point; determining the relative position relationship between the unmanned vehicle and the effective reference edge by using the ray method, and determining the current driving lane of the unmanned vehicle according to the relative position relationship.

[0090] The ray method refers to using horizontal rays or vertical rays to judge whether a point is within a polygon region.

[0091] The traditional ray method needs to determine the relative position relationship between the unmanned vehicle and each reference edge in a relevant lane, and if there are many reference edges, the ray method needs to consume a lot of computing resources to search for the driving lane of the unmanned vehicle. In this embodiment, the perception device can first screen the effective reference edges within the range that the ray can pass through from the multiple reference edges according to the coordinate information of the unmanned vehicle, and then determine the relative position relationship between the unmanned vehicle and the effective reference edge by using the ray method, so as to reduce the computing resources consumed.

[0092] Figure 5 A schematic diagram of the ray method is shown, Figure 5 The ray in is a horizontal right ray, and the dashed line segment represents the valid reference edge, and the solid line segment represents the reference edge to be removed.

[0093] Specifically, according to the coordinate information, the valid reference edge is selected from the plurality of reference edges, which can include the following sub-steps:

[0094] (1) Obtain the x-axis coordinate value and the y-axis coordinate value in the coordinate information.

[0095] The sensing device can obtain the x-axis coordinate value and the y-axis coordinate value of the unmanned vehicle by using the sensor in the unmanned vehicle.

[0096] (2) Obtain the x-axis minimum coordinate value, the x-axis maximum coordinate value, the y-axis minimum coordinate value and the y-axis maximum coordinate value of each reference edge.

[0097] The two endpoints of each reference edge are two reference points, and the sensing device can obtain the x-axis coordinate value and the y-axis coordinate value of the two reference points to obtain the x-axis minimum coordinate value, the x-axis maximum coordinate value, the y-axis minimum coordinate value and the y-axis maximum coordinate value of each reference edge.

[0098] The ray can be horizontal, including horizontal right direction and horizontal left direction, or the ray can be vertical, including vertical upward direction and vertical downward direction. The selection method of the valid reference edge according to the direction of the ray is described below.

[0099] (3) If the direction of the ray is horizontal right, select the reference edge with the x-axis maximum coordinate value greater than or equal to the x-axis coordinate value, the y-axis maximum coordinate value greater than or equal to the y-axis coordinate value, and the y-axis minimum coordinate value less than or equal to the y-axis coordinate value from the plurality of reference edges as the valid reference edge.

[0100] That is, the sensing device selects the valid reference edge located on the right side of the unmanned vehicle and intersecting with the ray.

[0101] For example, Figure 5 Assuming that the coordinate value of the unmanned vehicle is (x_ego, y_ego), and the coordinate value of the two endpoints of any reference edge is x_min, x_max, y_min and y_max, the selected valid reference edge needs to satisfy x_max≥x_ego and y_min≤y_ego≤y_max.

[0102] (4) If the direction of the ray is horizontal left, select the reference edge with the x-axis minimum coordinate value less than or equal to the x-axis coordinate value, the y-axis maximum coordinate value greater than or equal to the y-axis coordinate value, and the y-axis minimum coordinate value less than or equal to the y-axis coordinate value from the plurality of reference edges as the valid reference edge.

[0103] That is, the sensing device screens the effective reference edge which is located on the left side of the ego vehicle and will intersect with the ray.

[0104] (5) If the direction of the ray is vertically upward, the reference edge with the y-axis maximum coordinate value greater than or equal to the y-axis coordinate value, the x-axis maximum coordinate value greater than or equal to the x-axis coordinate value, and the x-axis minimum coordinate value less than or equal to the x-axis coordinate value is screened from the plurality of reference edges as the effective reference edge.

[0105] That is, the sensing device screens the effective reference edge which is located above the ego vehicle and will intersect with the ray.

[0106] (6) If the direction of the ray is vertically downward, the reference edge with the y-axis minimum coordinate value less than or equal to the y-axis coordinate value, the x-axis maximum coordinate value greater than or equal to the x-axis coordinate value, and the x-axis minimum coordinate value less than or equal to the x-axis coordinate value is screened from the plurality of reference edges as the effective reference edge.

[0107] That is, the sensing device screens the effective reference edge which is located below the ego vehicle and will intersect with the ray.

[0108] If the sensing device does not screen the effective reference edge, it is determined that the ego vehicle is not in the high-definition map, and the search fails; if the sensing device screens the effective reference edge, the ray in the horizontal direction to the right shown in formula (1) is taken as an example. Figure 5 The x is calculated according to formula (2), wherein y_ego represents the y-axis coordinate value of the ego vehicle, and x1, x2, y1 and y2 respectively represent the coordinate values of the two end points of an effective reference edge.

[0109] If x=x_ego, it means that the ego vehicle is located on the lane line of a related lane, and the screening ends; if x>x_ego, it means that the effective reference edge is located on the right side of the ego vehicle, the number of intersection points between the ego vehicle and the effective reference edge is counted, and if the number of intersection points is odd, it means that the ego vehicle is located in the related lane, otherwise it means that the ego vehicle is not in the related lane.

[0110] If it is determined that the ego vehicle is in at least two related lanes, the lane center lines of the at least two related lanes are calculated, the lane center line closest to the ego vehicle is found, and the related lane corresponding to the lane center line is determined as the driving lane of the ego vehicle.

[0111] Figure 6 An effect diagram of a driving lane searching method is shown, after the driving lane of the ego vehicle is determined, the lane center line can be queried for automatic driving tracking.

[0112] ​In summary, the driving lane searching method of the unmanned vehicle provided by the embodiments of the present application first generates the influence domain of each lane in the high-definition map; then searches for the relevant influence domain in which the unmanned vehicle is currently located from the influence domains, and determines the relevant lane corresponding to the relevant influence domain, which can perform coarse-grained screening on the thousands of lanes to quickly obtain the relevant relevant lane; finally, the ray method is used to search for the driving lane of the unmanned vehicle from the relevant lane, which can perform fine-grained screening on the relevant lane, thereby improving the searching speed and accuracy of the lane.

[0113] When searching for the driving lane of the unmanned vehicle from the relevant lane by using the ray method, first, the adjacent reference points in the relevant lane are connected, and the connecting line of each two reference points forms a reference edge, so that each lane line has multiple reference edges, and two lane lines are closed to finally form a closed polygon with multiple reference edges; then, the effective reference edge is screened from the multiple reference edges according to the coordinate information of the unmanned vehicle, which can quickly screen the effective reference edge located in the range that the ray can pass through from the multiple reference edges, thereby reducing the reference edges that need to be screened by using the ray method and greatly reducing the calculation amount.

[0114] Before searching for the driving lane of the unmanned vehicle from the relevant lane by using the ray method, the relevant lane with the lane direction consistent with the driving direction of the unmanned vehicle can be searched from the relevant lanes, which can eliminate the relevant lane with the lane direction opposite to the driving direction, thereby reducing the relevant lanes that need to be screened by using the ray method and further reducing the calculation amount.

[0115] Please refer to Figure 7 which shows the structure block diagram of the driving lane searching device of the unmanned vehicle provided by an embodiment of the present application. The driving lane searching device of the unmanned vehicle can be applied to the perception device of the unmanned vehicle. The driving lane searching device of the unmanned vehicle can include:

[0116] The acquisition module 710 is configured to acquire the lane line information of the multiple lanes from the high-definition map.

[0117] The generation module 720 is configured to generate the influence domain of each lane according to the lane line information of each lane, and the influence domain corresponds to the coordinate range of the two lane lines of the lane.

[0118] The searching module 730 is configured to search for the relevant influence domain in which the unmanned vehicle is currently located from the influence domains, and determine the relevant lane corresponding to the relevant influence domain.

[0119] The searching module 730 is further configured to search for the driving lane of the unmanned vehicle from the relevant lane by using the ray method.

[0120] In an optional embodiment, the searching module 730 is further configured to:

[0121] obtain coordinate information of the unmanned vehicle;

[0122] For each relevant lane, adjacent reference points in the relevant lane are connected, and each two reference points are connected to form a reference edge, so that each lane line has a plurality of reference edges, two lane lines are closed, and finally a closed polygon having a plurality of reference edges is formed; the effective reference edge is screened from the plurality of reference edges according to the coordinate information, and the effective reference edge is in a range that can be passed through by a ray with the unmanned vehicle as a starting point; the relative position relationship between the unmanned vehicle and the effective reference edge is determined by using the ray method, and the current driving lane of the unmanned vehicle is determined according to the relative position relationship.

[0123] In an optional embodiment, the searching module 730 is further configured to:

[0124] obtain an x-axis coordinate value and a y-axis coordinate value in the coordinate information;

[0125] obtain an x-axis minimum coordinate value, an x-axis maximum coordinate value, a y-axis minimum coordinate value, and a y-axis maximum coordinate value of each reference edge;

[0126] if the direction of the ray is horizontal to the right, the reference edge with the x-axis maximum coordinate value greater than or equal to the x-axis coordinate value, the y-axis maximum coordinate value greater than or equal to the y-axis coordinate value, and the y-axis minimum coordinate value less than or equal to the y-axis coordinate value is screened from the plurality of reference edges as the effective reference edge;

[0127] if the direction of the ray is horizontal to the left, the reference edge with the x-axis minimum coordinate value less than or equal to the x-axis coordinate value, the y-axis maximum coordinate value greater than or equal to the y-axis coordinate value, and the y-axis minimum coordinate value less than or equal to the y-axis coordinate value is screened from the plurality of reference edges as the effective reference edge;

[0128] if the direction of the ray is vertical upward, the reference edge with the y-axis maximum coordinate value greater than or equal to the y-axis coordinate value, the x-axis maximum coordinate value greater than or equal to the x-axis coordinate value, and the x-axis minimum coordinate value less than or equal to the x-axis coordinate value is screened from the plurality of reference edges as the effective reference edge;

[0129] if the direction of the ray is vertical downward, the reference edge with the y-axis minimum coordinate value less than or equal to the y-axis coordinate value, the x-axis maximum coordinate value greater than or equal to the x-axis coordinate value, and the x-axis minimum coordinate value less than or equal to the x-axis coordinate value is screened from the plurality of reference edges as the effective reference edge.

[0130] In an optional embodiment, the searching module 730 is further configured to:

[0131] before searching for the driving lane of the unmanned vehicle from the relevant lanes by using the ray method, obtaining a driving direction of the unmanned vehicle;

[0132] searching for a relevant lane with a lane direction consistent with the driving direction from the relevant lanes.

[0133] In an optional embodiment, the searching module 730 is further configured to:

[0134] For each relevant lane, a tangent direction of a lane center line point closest to the unmanned vehicle is determined as a lane direction of the relevant lane;

[0135] An included angle between the driving direction and the lane direction is calculated;

[0136] If an absolute value of the included angle is less than or equal to 90 degrees, it is determined that the lane direction is consistent with the driving direction, and the relevant lane is retained;

[0137] If the absolute value of the included angle is greater than 90 degrees, it is determined that the lane direction is inconsistent with the driving direction, and the relevant lane is excluded.

[0138] In an optional embodiment, when the lane line information comprises coordinate information of reference points in the lane line, the generating module 720 is further configured to:

[0139] For each lane, an x-axis minimum coordinate value, an x-axis maximum coordinate value, a y-axis minimum coordinate value and a y-axis maximum coordinate value are selected from the coordinate information of the reference points of the lane;

[0140] An influence domain of a predetermined shape is generated, wherein an x-axis coordinate value of any point in the influence domain is greater than or equal to the x-axis minimum coordinate value and less than or equal to the x-axis maximum coordinate value, and a y-axis coordinate value of any point in the influence domain is greater than or equal to the y-axis minimum coordinate value and less than or equal to the y-axis maximum coordinate value.

[0141] In an optional embodiment, the searching module 730 is further configured to:

[0142] An x-axis coordinate value and a y-axis coordinate value of the unmanned vehicle are obtained;

[0143] An x-axis minimum coordinate value, an x-axis maximum coordinate value, a y-axis minimum coordinate value and a y-axis maximum coordinate value of each influence domain are obtained;

[0144] From the influence domains, an influence domain in which the x-axis maximum coordinate value is greater than or equal to the x-axis coordinate value, the x-axis minimum coordinate value is less than or equal to the x-axis coordinate value, the y-axis maximum coordinate value is greater than or equal to the y-axis coordinate value, and the y-axis minimum coordinate value is less than or equal to the y-axis coordinate value is selected as a relevant influence domain.

[0145] To sum up, the driving lane searching device of the unmanned vehicle provided by the embodiments of the present application firstly generates the influence domain of each lane in the high-definition map; then searches for the relevant influence domain where the unmanned vehicle is currently located from the influence domains, and determines the relevant lane corresponding to the relevant influence domain, which can perform coarse-grained screening on thousands of lanes to quickly obtain the relevant lane; finally, the ray method is used to search for the driving lane of the unmanned vehicle from the relevant lane, which can perform fine-grained screening on the relevant lane, thereby improving the searching speed and accuracy of the lane.

[0146] When searching for the driving lane of the unmanned vehicle from the relevant lane by using the ray method, the adjacent reference points in the relevant lane are connected first, and the connecting line of each two reference points forms a reference edge, so that each lane line has multiple reference edges, and two lane lines are closed to finally form a closed polygon with multiple reference edges; then, the effective reference edge is screened from the multiple reference edges according to the coordinate information of the unmanned vehicle, which can quickly screen the effective reference edge located in the range that can be penetrated by the ray from the multiple reference edges, thereby reducing the reference edges that need to be screened by using the ray method and greatly reducing the calculation amount.

[0147] Before searching for the driving lane of the unmanned vehicle from the relevant lane by using the ray method, the relevant lane with the same lane direction as the driving direction of the unmanned vehicle can be searched from the relevant lanes, which can eliminate the relevant lane with the opposite lane direction to the driving direction, thereby reducing the relevant lane that needs to be screened by using the ray method and further reducing the calculation amount.

[0148] An embodiment of the present application provides a computer readable storage medium, the storage medium stores at least one instruction, the at least one instruction is loaded and executed by a processor to implement the driving lane searching method of the unmanned vehicle as described above.

[0149] An embodiment of the present application provides a perception device, the perception device includes a processor and a memory, the memory stores at least one instruction, the instruction is loaded and executed by the processor to implement the driving lane searching method of the unmanned vehicle as described above.

[0150] It should be noted that: the driving lane searching device of the unmanned vehicle provided by the above embodiments is only used as an example to illustrate the division of the above functional modules during the driving lane searching of the unmanned vehicle, and in actual application, the above functions can be completed by different functional modules according to needs, that is, the internal structure of the driving lane searching device of the unmanned vehicle is divided into different functional modules to complete all or part of the functions described above. In addition, the driving lane searching device of the unmanned vehicle and the driving lane searching method embodiment provided by the above embodiments belong to the same concept, and the specific implementation process is described in the method embodiment, which will not be repeated here.

[0151] Those skilled in the art can understand that all or part of the steps of the above-mentioned embodiments can be completed by hardware, or can be instructed by a program to complete the related hardware, and the program can be stored in a computer readable storage medium. The storage medium mentioned above can be a read-only memory, a magnetic disk or an optical disk, etc.

[0152] The above description is not intended to limit the embodiments of the present application, and any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the embodiments of the present application shall be included in the protection scope of the embodiments of the present application.

Claims

1. A method for finding a lane for an unmanned vehicle, characterized in that: The method comprises: Obtain lane line information for multiple lanes from high-precision maps; Generate an influence domain for each lane based on lane line information of each lane, where the influence domain corresponds to a coordinate range corresponding to two lane lines of the lane; Find the relevant influence domain where the unmanned vehicle is currently located from each influence domain, and determine the relevant lane corresponding to the relevant influence domain; Using a ray method to find the driving lane of the unmanned vehicle from the relevant lanes; The method of using the ray method to find the current driving lane of the unmanned vehicle from the relevant lanes includes: obtaining the coordinate information of the unmanned vehicle; for each relevant lane, connecting adjacent reference points in the relevant lane, and forming a reference edge with the line connecting every two reference points, so that each lane line has multiple reference edges, closing the two lane lines, and finally forming a closed polygon with multiple reference edges; screening valid reference edges from the multiple reference edges according to the coordinate information, and the valid reference edges are within the range that can be passed by the ray starting from the unmanned vehicle; using the ray method to determine the relative position relationship between the unmanned vehicle and the valid reference edge, and determining the current driving lane of the unmanned vehicle based on the relative position relationship.

2. The lane finding method for an unmanned vehicle according to claim 1, characterized in that: The filtering of a valid reference edge from the plurality of reference edges according to the coordinate information includes: Obtaining the x-axis coordinate value and the y-axis coordinate value in the coordinate information; Get the minimum x-axis coordinate value, maximum x-axis coordinate value, minimum y-axis coordinate value, and maximum y-axis coordinate value of each reference edge; If the direction of the ray is horizontal to the right, a reference edge whose maximum x-axis coordinate value is greater than or equal to the x-axis coordinate value, whose maximum y-axis coordinate value is greater than or equal to the y-axis coordinate value, and whose minimum y-axis coordinate value is less than or equal to the y-axis coordinate value is selected from the multiple reference edges as a valid reference edge; If the direction of the ray is horizontal to the left, a reference edge whose minimum x-axis coordinate value is less than or equal to the x-axis coordinate value, whose maximum y-axis coordinate value is greater than or equal to the y-axis coordinate value, and whose minimum y-axis coordinate value is less than or equal to the y-axis coordinate value is selected from the multiple reference edges as a valid reference edge; If the direction of the ray is vertically upward, a reference edge whose maximum y-axis coordinate value is greater than or equal to the y-axis coordinate value, whose maximum x-axis coordinate value is greater than or equal to the x-axis coordinate value, and whose minimum x-axis coordinate value is less than or equal to the x-axis coordinate value is selected from the multiple reference edges as a valid reference edge; If the direction of the ray is vertically downward, a reference edge whose minimum y-axis coordinate value is less than or equal to the y-axis coordinate value, whose maximum x-axis coordinate value is greater than or equal to the x-axis coordinate value, and whose minimum x-axis coordinate value is less than or equal to the x-axis coordinate value is selected from the multiple reference edges as a valid reference edge.

3. The method for finding a lane for an unmanned vehicle according to claim 1, wherein: Before searching the driving lane of the unmanned vehicle from the relevant lanes using the ray method, the method further includes: Obtaining the driving direction of the unmanned vehicle; A relevant lane having a lane direction consistent with the driving direction is searched from among the relevant lanes.

4. The lane finding method for an unmanned vehicle according to claim 3, characterized in that: The step of searching for a lane having a direction consistent with the driving direction from among the relevant lanes includes: For each relevant lane, determining the tangent direction of the lane centerline point closest to the unmanned vehicle as the lane direction of the relevant lane; Calculating the angle between the driving direction and the lane direction; If the absolute value of the included angle is less than or equal to 90 degrees, it is determined that the lane direction is consistent with the driving direction, and the relevant lane is retained; If the absolute value of the angle is greater than 90 degrees, it is determined that the lane direction is inconsistent with the driving direction, and the relevant lane is removed.

5. The method for finding a lane for an unmanned vehicle according to any one of claims 1 to 4, characterized in that: When the lane line information includes coordinate information of a reference point in the lane line, generating the influence domain of each lane according to the lane line information of each lane includes: For each lane, selecting the minimum x-axis coordinate value, the maximum x-axis coordinate value, the minimum y-axis coordinate value, and the maximum y-axis coordinate value from the coordinate information of each reference point of the lane; Generate an influence domain of a predetermined shape, wherein the x-axis coordinate value of any point in the influence domain is greater than or equal to the x-axis minimum coordinate value and less than or equal to the x-axis maximum coordinate value, and the y-axis coordinate value of any point in the influence domain is greater than or equal to the y-axis minimum coordinate value and less than or equal to the y-axis maximum coordinate value.

6. The lane finding method for an unmanned vehicle according to claim 5, characterized in that: The process of searching for the relevant influence domain where the unmanned vehicle is currently located from each influence domain includes: Obtaining the x-axis coordinate value and the y-axis coordinate value of the unmanned vehicle; Obtain the minimum x-axis coordinate value, maximum x-axis coordinate value, minimum y-axis coordinate value, and maximum y-axis coordinate value of each influence domain; From each influence domain, filter out the influence domain whose maximum x-axis coordinate value is greater than or equal to the x-axis coordinate value, whose minimum x-axis coordinate value is less than or equal to the x-axis coordinate value, whose maximum y-axis coordinate value is greater than or equal to the y-axis coordinate value, and whose minimum y-axis coordinate value is less than or equal to the y-axis coordinate value as the relevant influence domain.

7. A lane finding device for an unmanned vehicle, characterized in that: The device comprises: The acquisition module is used to obtain lane line information of multiple lanes from the high-precision map; A generation module, configured to generate an influence domain for each lane based on lane line information of each lane, wherein the influence domain corresponds to a coordinate range corresponding to two lane lines of the lane; A search module is used to search the relevant influence domain where the unmanned vehicle is currently located from each influence domain, and determine the relevant lane corresponding to the relevant influence domain; The search module is further configured to use a ray method to search for the driving lane of the unmanned vehicle from the relevant lanes; The search module is also used to: obtain the coordinate information of the unmanned vehicle; for each relevant lane, connect the adjacent reference points in the relevant lane, and form a reference edge with the line connecting every two reference points, so that each lane line has multiple reference edges, close the two lane lines, and finally form a closed polygon with multiple reference edges; filter valid reference edges from the multiple reference edges according to the coordinate information, and the valid reference edges are within the range that can be passed by the ray starting from the unmanned vehicle; use the ray method to determine the relative position relationship between the unmanned vehicle and the valid reference edge, and determine the current driving lane of the unmanned vehicle based on the relative position relationship.

8. A computer-readable storage medium, characterized in that The storage medium stores at least one instruction, and the at least one instruction is loaded and executed by the processor to implement the lane finding method for an unmanned vehicle as described in any one of claims 1 to 6.

9. A sensing device, characterized in that: The perception device includes a processor and a memory, wherein the memory stores at least one instruction, and the instruction is loaded and executed by the processor to implement the driving lane search method for an unmanned vehicle as described in any one of claims 1 to 6.

Images