Automatic parking method and apparatus, and vehicle

Abstract

An automatic parking method and apparatus, and a vehicle. The method comprises: acquiring information of obstacles and position information of a target parking space; on the basis of the information of the obstacles and the position information of the target parking space, determining whether the target parking space is a narrow and elongated parking space, wherein the narrow and elongated parking space is a parking space having at least two parking space boundary lines constrained by target obstacles; and if the target parking space is the narrow and elongated parking space, performing automatic parking in a head-in parking mode. In the automatic parking method, when the target parking space is the narrow and elongated parking space, automatic parking can be performed in the head-in parking mode, thereby reducing the length of a parking path and the number of parking operations, reducing the parking error, and improving the accuracy and safety of parking.

Description

Automatic parking method, device and vehicle

[0001] Cross-reference to Related Applications

[0002] This application claims priority to Chinese application No. 202410890460X filed on July 3, 2024, which is hereby incorporated by reference in its entirety for all purposes. TECHNICAL FIELD

[0003] The present application relates to the technical field of vehicles, and more particularly, to an automatic parking method, device and vehicle. BACKGROUND

[0004] In the new wave of reform in the automotive industry, the label of "intelligent" is added and empowered, and the Autonomous Parking Assistance (APA) technology has become a research hotspot. In related ways, the workflow of APA can be divided into three parts: environment perception, decision planning, and path following. In the decision planning stage, path planning can be performed based on the target parking space, the current position of the vehicle, and the rear parking-in mode. However, due to the diversity of parking space deployment in parking lots, the parking environment becomes more complex, such as the existence of some narrow parking spaces and more visual blind areas during parking. It may be difficult to adapt to complex parking environments through related ways for automatic parking, thereby resulting in low accuracy and safety of automatic parking. SUMMARY

[0005] In view of the above problems, the present application provides an automatic parking method, device and vehicle to improve the above problems.

[0006] In a first aspect, the present application provides an automatic parking method, which comprises: obtaining information of an obstacle itself and position information of a target parking space; determining whether the target parking space is a narrow parking space based on the information of the obstacle itself and the position information of the target parking space, the narrow parking space being a parking space in which at least two parking boundary lines are constrained by a target obstacle; and performing automatic parking in a front parking-in mode if the target parking space is a narrow parking space.

[0007] In a second aspect, the present application provides an automatic parking method, which comprises: obtaining an echo time difference corresponding to an ultrasonic wave; obtaining parking space information of the vehicle based on the echo time difference; and performing automatic parking in a front parking-in mode if the parking space information indicates that the available parking space of the vehicle is limited during parking.

[0008] In a third aspect, the present application provides an automatic parking device, comprising: an information acquisition unit configured to acquire information of an obstacle itself and position information of a target parking space; a parking space type acquisition unit configured to determine whether the target parking space is a narrow parking space based on the information of the obstacle itself and the position information of the target parking space, the narrow parking space being a parking space in which at least two parking boundary lines are constrained by a target obstacle; and a parking strategy generation unit configured to perform automatic parking in a head-in manner if the target parking space is a narrow parking space.

[0009] In a fourth aspect, the present application provides an automatic parking device, comprising: an information acquisition unit configured to acquire an echo time difference corresponding to an ultrasonic wave, the echo time difference representing a distance between a vehicle and an obstacle; a parking space information acquisition unit configured to obtain parking space information of the vehicle based on the echo time difference; and a parking strategy generation unit configured to perform automatic parking in a head-in manner if the parking space information indicates that a borrowable parking space of the vehicle is limited during parking.

[0010] In a fifth aspect, the present application provides a vehicle, comprising one or more processors and a memory; one or more programs stored in the memory and configured to be executed by the one or more processors, the one or more programs being configured to perform the method described above.

[0011] In a sixth aspect, the present application provides a computer-readable storage medium, the computer-readable storage medium storing program code, wherein the program code performs the method described above when executed.

[0012] The automatic parking method, device, vehicle and storage medium provided by the present application can determine whether a target parking space is a narrow parking space after acquiring information of an obstacle itself and position information of the target parking space, and perform automatic parking in a head-in manner if the target parking space is a narrow parking space. In this way, the length of a parking path and the number of parking maneuvers can be reduced, and thus the parking error can be reduced, and the accuracy and safety of parking can be improved. BRIEF DESCRIPTION OF DRAWINGS

[0013] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative effort.

[0014] FIG. 1 shows a flow chart of an automatic parking method according to an embodiment of the present application;

[0015] FIG. 2 shows a schematic diagram of a position relationship according to the present application;

[0016] FIG. 3 shows a schematic diagram of a constraint space according to the present application;

[0017] FIG. 4 shows a flow chart of another automatic parking method according to an embodiment of the present application;

[0018] FIG. 5 shows a flow chart of an implementation method of S220 in FIG. 4 according to the present application;

[0019] FIG. 6 shows a flow chart of an implementation method of S224 in FIG. 5 according to the present application;

[0020] FIG. 7 shows a schematic diagram of a reference grid point according to the present application;

[0021] FIG. 8 shows a schematic diagram of a contribution space according to the present application;

[0022] FIG. 9 shows a schematic diagram of a three-dimensional distribution curve according to the present application;

[0023] FIG. 10 shows a schematic diagram of a position relationship between a target obstacle boundary line and a target parking space boundary line according to the present application;

[0024] FIG. 11 shows a flow chart of yet another automatic parking method according to an embodiment of the present application;

[0025] FIG. 12 shows a schematic diagram of a time difference of echoes according to the present application;

[0026] FIG. 13 shows a schematic diagram of a parking mode according to the present application;

[0027] FIG. 14 shows a schematic diagram of another parking mode according to the present application;

[0028] FIG. 15 shows a flow chart of an automatic parking method according to another embodiment of the present application;

[0029] FIG. 16 shows a structural block diagram of an automatic parking device according to an embodiment of the present application;

[0030] FIG. 17 shows a structural block diagram of a vehicle according to the present application. DETAILED DESCRIPTION

[0031] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative effort belong to the scope of protection of the present application.

[0032] In the embodiments of the present application, the inventors provide an automatic parking method, device and vehicle. After obtaining the information of obstacles themselves and the position information of a target parking space, whether the target parking space is a narrow parking space is determined based on the information of obstacles themselves and the position information of the target parking space. If the target parking space is a narrow parking space, automatic parking is performed in a way of parking with a front end of the vehicle. In the above way, whether the target parking space is a narrow parking space can be determined after obtaining the information of obstacles themselves and the position information of the target parking space. When the target parking space is a narrow parking space, automatic parking is performed in a way of parking with a front end of the vehicle. Therefore, the length of a parking path and the number of parking cranks are reduced, and thus the parking error is reduced, and the accuracy and safety of parking are improved.

[0033] Referring to FIG. 1, an automatic parking method provided by an embodiment of the present application includes the following steps.

[0034] S110: Obtain the information of obstacles themselves and the position information of a target parking space.

[0035] The information of obstacles themselves can include the position information of each of a plurality of obstacles in a first coordinate system and the category information of each of the plurality of obstacles. The obstacles can refer to pedestrians, traffic cones, parking space stoppers, load-bearing columns (walls) of an underground garage, other vehicles parked in a parking space, ground locks and the like detected by a visual perception system (such as a camera, a laser radar and the like) of a vehicle within a perception range of the visual perception system. The first coordinate system can refer to a world coordinate system, a vehicle coordinate system or an image coordinate system.

[0036] The target parking space can refer to a parking space selected by a user, that is, an end point of a parking path of the vehicle. The target parking space can be represented by a polygon (such as a rectangle, a parallelogram and the like) composed of a plurality of target parking space boundary lines. The position information of the target parking space can include the position information of the target parking space boundary lines.

[0037] As a way, after an APA function of the vehicle is started, the visual perception system of the vehicle can start to collect the parking space information of all parking spaces in a space where the vehicle is located and the information of obstacles themselves. After the parking space information of all parking spaces and the information of obstacles themselves are collected, the vehicle can display the information in the form of an image on a central control display screen of the vehicle and prompt the user to select a parking space. After receiving the selection information of the user, the parking space selected by the user is taken as a target parking space, and the position information of the target parking space is obtained.

[0038] Optionally, the visual perception system of the vehicle can first collect image information of the space where the vehicle is located, and then obtain parking space information of all parking spaces and information of the obstacle itself through artificial intelligence image processing technology (such as target detection), the parking space information can refer to corner point coordinates of the parking space, so that after the target parking space is determined, the position information of the target parking space can be obtained based on the corner point coordinates of the target parking space.

[0039] S120: Determine whether the target parking space is a narrow parking space based on the information of the obstacle itself and the position information of the target parking space, the narrow parking space being a parking space in which at least two parking boundary lines are constrained by the target obstacle.

[0040] As one way, the boundary information of the target obstacle boundary line can be obtained based on the information of the obstacle itself and the position information, the target obstacle boundary line can be an obstacle boundary line that exists parking space constraint to the target parking space; based on the position information and the boundary information, it is determined whether the target parking space is a narrow parking space.

[0041] The boundary information can include position relationship and position change trend of the target obstacle boundary line and the target parking boundary line. The position relationship can be intersection, parallel, slight separation or absolute separation.

[0042] When the target obstacle boundary line intersects with the target parking boundary line, it can be shown that there is an intersection point between them, and due to the different intersection angles and lengths of the two, the distances between different positions of the boundary lines are also different. Slight separation can refer to that the distance between the target obstacle boundary line and the target parking boundary line is greater than a first preset distance threshold and less than a second preset distance threshold, and absolute separation can refer to that the distance between the target obstacle boundary line and the target parking boundary line is greater than or equal to the second preset distance threshold.

[0043] For example, as shown in FIG. 2, the vehicle body boundary line on the left side of the vehicle intersects with the parking boundary line on the right side of the parking spaces 1-3 at the tail part, is slightly separated at the middle part of the vehicle body, and is absolutely separated at the head part.

[0044] The position change trend can be no intersection, front-end intersection or rear-end intersection, where the front-end intersection can refer to that the intersection point is located in front of the vehicle entry line, and the rear-end intersection can refer to that the intersection point is located in rear of the vehicle entry line, and the vehicle entry line can be understood as a boundary line of the parking space that the vehicle first passes when entering the parking space. For example, referring to FIG. 2, a vehicle can enter the parking space 1-4 from the aisle, and there are two walls that intersect at an acute angle near the parking space 1-4, and the intersection point of the wall 1 and the left boundary line of the parking space 1-4 is in front of the vehicle entry line, and thus the intersection is a front-end intersection. The vehicle can enter the parking space 1-1 from the aisle, and there are two walls that intersect at an obtuse angle near the parking space 1-1, and the intersection point of the wall 3 and the left boundary line of the parking space 1-1 is in rear of the vehicle entry line, and thus the intersection is a rear-end intersection.

[0045] Optionally, the information of the obstacle itself can be smoothed to obtain a reference obstacle boundary line, and the reference obstacle boundary line can be a boundary line formed by obstacles of a specified category; and based on the position information and the reference obstacle boundary line, boundary information of the target obstacle boundary line is obtained.

[0046] Optionally, the specified category can be a wall or the like.

[0047] Optionally, the target parking space area can be obtained based on the position information of the target parking space boundary line; the constraint space area formed by the target obstacle boundary line can be obtained based on the position relationship and the position change trend between the target obstacle boundary line and the target parking space boundary line; and the constraint feature can be obtained based on the target parking space area and the constraint space area; and whether the target parking space is a narrow parking space can be determined based on the constraint feature.

[0048] Optionally, the constraint space can refer to a closed polygon formed by the target obstacle boundary line, as shown in FIG. 3, when there are two or more target obstacle boundary lines, the constraint space can refer to a closed polygon formed by connecting the end points of the target obstacle boundary lines. It should be noted that when there is no target obstacle boundary line or only one target obstacle boundary line, the constraint space cannot be formed.

[0049] In the embodiments of the present application, the constraint feature can be no constraint, one-side constraint, two-side constraint or three-side constraint.

[0050] Optionally, the intersection area can be obtained based on the target parking space area and the constraint space area, and the intersection area can be understood as an area that belongs to both the target parking space and the constraint space; and the constraint feature can be obtained based on the ratio of the intersection area to the constraint space area. For example, in FIG. 3, the intersection area can be the same as the target parking space area.

[0051] Optionally, if the ratio of the intersection area and the constraint space area is greater than the first preset value and less than the second preset value, it can be determined that the constraint feature is a two-side constraint; if the ratio of the intersection area and the constraint space area is greater than or equal to the second preset value and less than the third preset value, it can be determined that the constraint feature is a three-side constraint; if the ratio of the intersection area and the constraint space area is the fourth preset value, it can be determined that the constraint feature is no constraint or a single-side constraint.

[0052] The first preset value, the second preset value, the third preset value and the fourth preset value can be preset values obtained based on multiple tests. For example, the first preset value can be 0, the second preset value can be 0.5, the third preset value can be 1, and the fourth preset value can be ∞. Since there is no target obstacle boundary line or only one target obstacle boundary line, the ratio obtained can be ∞.

[0053] Optionally, if the constraint feature is a two-side constraint or a three-side constraint, it is determined that the parking space type of the target parking space is a narrow and long parking space; if the constraint feature is a single-side constraint or no constraint, it is determined that the parking space type of the target parking space is a non-narrow and long parking space.

[0054] S130: If the target parking space is a narrow and long parking space, automatic parking is performed in a car head parking manner.

[0055] As one way, if the target parking space is a narrow and long parking space, automatic parking can be directly performed in a car head parking manner.

[0056] In the embodiments of the present application, automatic parking is directly performed based on the parking strategy, which can speed up the automatic parking and save the user's time.

[0057] As another way, if the target parking space is a narrow and long parking space, recommendation information can be sent to the user, and the recommendation information can be used to recommend to the user to perform automatic parking in a car head parking manner; if a confirmation operation of the user indicating agreement to the recommendation information is received, automatic parking is performed in a car head parking manner.

[0058] The confirmation operation of the user indicating agreement to the recommendation information can be receiving voice information of the user indicating agreement, or a touch operation of the user indicating agreement to the central control display screen, or recognizing a gesture of the user indicating agreement.

[0059] In the embodiments of the present application, when the parking space type of the target parking space is a narrow and long parking space, recommendation information is sent to the user, and after a confirmation operation of the user indicating agreement to the recommendation information, automatic parking is performed, which can improve the selection right of the user, increase the interactivity of the user and the vehicle, and improve the user experience.

[0060] Optionally, if the parking space type of the target parking space is a non-narrow and long parking space, automatic parking is performed in a car head parking manner.

[0061] The automatic parking method provided in this embodiment determines whether the target parking space is a narrow parking space based on the information of the obstacles and the position information of the target parking space after the information of the obstacles and the position information of the target parking space are acquired, and performs automatic parking in the way of head-in parking if the target parking space is a narrow parking space. In this way, it can be determined whether the target parking space is a narrow parking space after the information of the obstacles and the position information of the target parking space are acquired, and automatic parking is performed in the way of head-in parking when the target parking space is a narrow parking space, thereby reducing the length of the parking path and the number of parking operations, and further reducing the parking error and improving the accuracy and safety of parking.

[0062] Referring to FIG. 4, the automatic parking method provided in this embodiment includes the following steps.

[0063] S210: Acquire the information of the obstacles and the position information of the target parking space.

[0064] S220: Perform smoothing operation on the information of the obstacles to obtain a reference obstacle boundary line, which is a boundary line formed by obstacles of a specified category.

[0065] As a way, the smoothing operation can be performed based on the position information of each of the plurality of obstacles in the first coordinate system to obtain a plurality of obstacle boundary lines, and the reference obstacle boundary line can be obtained based on the category information of each of the plurality of obstacles and the plurality of obstacle boundary lines.

[0066] In the prior art, straight line detection can be performed by Hough Transform, that is, a plurality of obstacle boundary lines are obtained, and the main implementation method of Hough Transform can be as follows.

[0067] 1. Any straight line in a 2D-cartesian coordinate system can be expressed as a straight line expression in Hough space: ρ = cosθ·x + sinθ·y, where x and y can be the position information of the obstacles in the first coordinate system, and ρ and θ can be the single injection of the intercept and slope of the straight line, respectively, θ∈[0,2π];

[0068] 2. Assuming that there are n points in a 2D-cartesian coordinate system, which form a point set N = {(x i ,y i )|i = 1, 2, 3...n}. Bringing all points in the point set into the straight line expression in Hough space will form n trigonometric function expansions with θ as the parameter {(cosθ·x i + sinθ·y i= p) || i = 1, 2, 3...n}, the above three trigonometric function expansions are combined into The form of is taken out respectively, and the function curve of each combined trigonometric function expression is drawn in the 2D-cartesian coordinate system ρ-θ plane with ρ as the dependent variable and θ as the independent variable, to obtain n sinusoidal function curves.

[0069] 3. When the n points of the above assumption are on the same straight line, it means that there is a unique slope and a unique intercept, so the n sinusoidal function curves must intersect at a certain point (θ0, ρ0) on the 2D-cartesian ρ-θ plane.

[0070] 4. After calculating the intersection point (θ0, ρ0), it is substituted into ρ0 = cosθ0·x + sinθ0·y, and the mathematical expression of the straight line in the 2D-cartesian coordinate system can be obtained.

[0071] 5. If there are multiple points in the 2D-cartesian coordinate system that can form multiple straight lines, the greedy strategy is adopted to calculate the intersection points of the sinusoidal function curve cluster on the 2D-cartesian ρ-θ plane according to the above steps 1-3, and the number of sinusoidal function curves passing through each intersection point is counted. The θ and ρ corresponding to the intersection points ranked in the top k are taken out, so as to determine the k straight lines {ρ i = cos i·x + sin i·y | i = 1, 2, 3...k}.

[0072] In the application scenario of the present application, it is not appropriate to directly use the Hough transform to obtain multiple obstacle boundary lines. The reason is that, in ideal conditions (i.e. in a theoretical space), through the Hough transform, a unique straight line can be fitted for each edge of the obstacle; however, in engineering practice, due to factors such as sensor errors, some points may not fall on the straight line, but show a normal distribution near the straight line. Directly applying the theoretical method in engineering practice may have multiple fitting results for the same distribution and the same number of points, that is, sometimes fitting a straight line A and sometimes fitting a straight line B, so that the fitting result is unstable, which further affects the judgment result of the target parking space type. Therefore, in order to improve the accuracy of the judgment, the present application proposes an improved Hough transform, i.e. a Hough transform carrying a locally soft non-maximum suppression (Locally Soft NMS) mechanism, so that in engineering practice, a unique straight line can also be fitted for each edge of the obstacle.

[0073] The steps of obtaining multiple obstacle boundary lines through the improved Hough transform (smoothing operation) proposed in the present application will be introduced below with reference to FIGS. 5 and 6.

[0074] Optionally, as shown in FIG. 5, based on the position information of each of the plurality of obstacles in the first coordinate system, a smoothing operation is performed to obtain a plurality of obstacle boundary lines, including:

[0075] S221: Based on the position information of each of the plurality of obstacles in the first coordinate system, a two-dimensional sinusoidal curve of each of the plurality of obstacles in the second coordinate system is obtained to obtain a plurality of two-dimensional sinusoidal curves.

[0076] The position information of each of the plurality of obstacles in the first coordinate system can be the coordinates of the discrete points corresponding to each of the plurality of obstacles, such as (x, y). The second coordinate system can refer to a 2D-cartersian coordinate system.

[0077] As one way, based on the position information of each of the plurality of obstacles in the first coordinate system, a two-dimensional sinusoidal curve of each of the plurality of obstacles in the second coordinate system can be obtained based on the aforementioned Hough transform method to obtain a plurality of two-dimensional sinusoidal curves, which can form a sinusoidal curve cluster:

[0078] S222: Based on the value range and domain of the plurality of two-dimensional sinusoidal curves, a grid space of the plurality of two-dimensional sinusoidal curves is obtained.

[0079] In the grid space, there can be a plurality of grids with the same shape, and each grid can correspond to four grid points, so that the grid space can include a plurality of grid points. The domain of the plurality of two-dimensional sinusoidal curves can be [0, 2π].

[0080] As one way, by traversing the plurality of two-dimensional sinusoidal curves, the value range of the two-dimensional sinusoidal curve can be obtained, which can be represented as Thus, the ρ-θ plane can be limited to θ∈[0, 2π], and the rectangular region is discretized in the ρ-θ axis with a step size of to obtain a grid space of , wherein represents the upward rounding.

[0081] S223: Discretization processing is performed on the plurality of two-dimensional sinusoidal curves in the grid space to obtain a plurality of feature points.

[0082] As one way, for each two-dimensional sinusoidal curve , starting from θ = 0, increasing θ by a step size of N and calculating ρ successively: Thus, a plurality of feature points (θ k , ρi,k )。

[0083] S224: Obtain the plurality of obstacle boundary lines based on the plurality of feature points.

[0084] As one way, the contribution value of each grid point can be obtained based on the plurality of feature points; the target grid point can be obtained based on the contribution value of each grid point, the target grid point can be the grid point with the top K contribution values, and the contribution value represents the probability of the obstacle being located at the corresponding grid point; and the plurality of obstacle boundary lines can be obtained based on the target grid point.

[0085] Optionally, as shown in FIG. 6, the contribution value of each grid point can be obtained based on the plurality of feature points, including:

[0086] S2241: Obtain the three-dimensional distribution curve of each feature point in the second coordinate system based on the position information of each feature point in the second coordinate system, and the three-dimensional distribution curve represents the energy distribution of the corresponding feature point in the second coordinate system.

[0087] As one way, the three-dimensional distribution curve of each feature point in the second coordinate system can be obtained based on the position information of each feature point in the second coordinate system and a preset formula, the three-dimensional distribution curve can form a peak in the grid space, and the preset formula is:

[0088] wherein (θ k ,ρ i,k ) represents the position information of the feature point, and α, γ, β and γ can be preset constants, and (θ, ρ) can represent any point in the grid space.

[0089] S2242: Obtain the reference grid point corresponding to each feature point, and the reference grid point is the vertex of the grid where the feature point is located.

[0090] As one way, the grid where each feature point is located can be determined based on the position information of each feature point, and then the four vertices of the grid are taken as the reference grid point of the corresponding feature point. For example, as shown in FIG. 7, the reference grid point of the feature point (θ k ,ρ i,k ) is point A, B, C and D.

[0091] S2243: Obtain the contribution degree curve of the reference grid point corresponding to each feature point based on the distance between each feature point and the corresponding reference grid point and the three-dimensional distribution curve corresponding to each feature point, to obtain the contribution degree curve of the reference grid point corresponding to the plurality of feature points, and the contribution degree curve can represent the contribution degree distribution of the reference grid point of the corresponding feature point in the second coordinate system.

[0092] As a manner, the distance between each feature point and the corresponding reference grid point and the three-dimensional distribution curve corresponding to each feature point can be substituted into the following formula to obtain the contribution degree curve of each feature point corresponding to the reference grid point, so that one peak corresponding to the three-dimensional distribution curve can become four peaks distributed on the reference grid point, and the formula is:

[0093] Wherein, d j The distance between the feature point and the corresponding jth reference grid point can be obtained by the coordinates of the reference grid point and the feature point.

[0094] S2244: Obtain the contribution value of each grid point based on the contribution degree curve of the reference grid point corresponding to the plurality of feature points.

[0095] As a manner, the contribution degree curve of the plurality of feature points corresponding to the reference grid point can be as shown in FIG. 8, the values of the contribution degree curve of the plurality of feature points corresponding to the reference grid point on the same reference grid point can be added to obtain the contribution value of the reference grid point, and thus the contribution values of all grid points in the grid space are obtained.

[0096] Optionally, after obtaining the contribution value of each grid point, the entire grid space can form a contribution degree space Ω-ρ-θ, and by performing the following soft-argmax smoothing operation on the contribution degree space multiple times, K target grid points with contribution values greater than a preset contribution value and ranked in the top K positions are screened out, that is, K groups (θ * ,ρ * ). Wherein, K and the preset contribution value can be obtained based on actual experience.

[0097] Exemplarily, the contribution degree space can be as shown in FIG. 8, and after one smoothing operation, a group (θ * ,ρ * ) can be obtained, which (θ * ,ρ * ) can obtain a new three-dimensional distribution curve as shown in FIG. 9 through step S2241, and through the new three-dimensional distribution curve, an obstacle boundary line can be obtained.

[0098] In the embodiments of the present application, by performing smoothing operations in the ρ and θ directions respectively, the influence of all contribution degree curves in the ρ and θ directions respectively can be comprehensively considered, and compared with the manner of performing smoothing operation globally in the entire contribution degree space in the prior art (only considering a certain contribution degree curve), the accuracy of the target grid point can be improved.

[0099] Optionally, after the target grid point is obtained, a plurality of obstacle boundary lines can be obtained based on the target grid point and the conversion relationship of the coordinate system, wherein one target grid point can correspond to one obstacle curve.

[0100] Optionally, after the plurality of obstacle boundary lines are obtained, a reference obstacle boundary line can be obtained based on the category information of each of the plurality of obstacles and the plurality of obstacle boundary lines.

[0101] The category information of each of the plurality of obstacles can be represented by a label value. For example, the label value of a pedestrian can be 1, the label value of a vehicle can be 2, and the label value of a wall can be 3.

[0102] Since each of the plurality of obstacles corresponds to category information, all feature points in the grid space can also correspond to category information. The confidence space can be obtained in the same way as steps S2241-S2244. The label value space Ω-ρ-θ can be obtained based on the confidence space and the category information of all feature points, and the reference obstacle boundary line can be obtained.

[0103] Optionally, the method for obtaining the reference obstacle boundary line based on the label value space H-ρ-θ and the plurality of obstacle boundary lines can be as follows: for each (θ * ,ρ * ) corresponding to an obstacle boundary line, the average of the label values of the four nearest neighbor grid points of (θ * ,ρ * ) in the label value space H-ρ-θ can be calculated. The average is rounded according to the "rounding" rule, and the integer value corresponding to the rounding is taken as the label value corresponding to the obstacle boundary line, so as to obtain the category information of the obstacle boundary line, and the obstacle boundary line with the category information of the wall is taken as the reference obstacle boundary line.

[0104] S230: Obtain boundary information of the target obstacle boundary line based on the position information and the reference obstacle boundary line.

[0105] As one way, the target obstacle boundary line can be obtained based on the position information of the target parking space boundary line and the reference obstacle boundary line. The target obstacle boundary line can be a reference obstacle boundary line whose distance from the target parking space boundary line is within a preset distance value range. The boundary information of the target obstacle boundary line can be obtained based on the distance between the target parking space boundary line and the target obstacle boundary line and the slope relationship between the target parking space boundary line and the target obstacle boundary line.

[0106] Optionally, the target obstacle boundary line can be obtained based on the position information of the target parking space boundary line and the distance between the reference obstacle boundary line. If the distance between the position information of the target parking space boundary line and the reference obstacle boundary line is less than the boundary distance threshold, it can be indicated that the reference obstacle boundary line is the target obstacle boundary line, that is, the reference obstacle boundary line near the target parking space needs to be considered when determining the type of the target parking space.

[0107] The distance between the position information of the target parking space boundary line and the reference obstacle boundary line can be understood as the distance between the centroids of the target parking space boundary line and the reference obstacle boundary line. The centroid of the target parking space boundary line can be understood as the center of the target parking space boundary line, and the centroid of the reference obstacle boundary line can be obtained based on the following method:

[0108] After obtaining the category information of the K obstacle target parking space boundary lines, the distances of all obstacles from the discrete points to each obstacle boundary line can be judged one by one. When the distance between a certain discrete point and a certain obstacle boundary line is less than a preset threshold, the discrete point can be attributed to the obstacle boundary line. After traversing all obstacle boundary lines and all discrete points, the obstacle points that do not meet the attribution condition can be deleted, so that each obstacle boundary line can be associated with a limited number of discrete points. Thus, the centroid of each obstacle boundary line, that is, the mean value of the X and Y axes, can be obtained from the discrete points in each obstacle boundary line and the coordinates of the associated discrete points.

[0109] Optionally, after obtaining the target obstacle boundary line, the target obstacle boundary line can be arranged in the order of the centroid relative to the self-vehicle counterclockwise rotation, and the intersection points can be calculated two by two. According to the intersection points, the target obstacle boundary line can be fitted to the cutting state of the target parking space.

[0110] Optionally, the position relationship between the target obstacle boundary line and the target parking space boundary line can be obtained based on the distance between the target parking space boundary line and the target obstacle boundary line according to the method in step S120.

[0111] As shown in FIG. 10, for example, there are three target obstacle boundary lines, which are parallel or intersect with the target parking space boundary line.

[0112] Optionally, the position change trend of the target obstacle boundary line and the target parking space boundary line can be obtained based on the slope relationship between the target parking space boundary line and the target obstacle boundary line. If at least one of the x component and the y component of the intersection point coordinates of the target parking space boundary line and the target obstacle boundary line is greater than a preset component threshold, or the slope difference between the target parking space boundary line and the target obstacle boundary line is less than or equal to a tolerance threshold, it can be determined that the target parking space boundary line and the target obstacle boundary line belong to a parallel relationship, that is, no intersection.

[0113] In the embodiment of the present application, whether the target parking space boundary line and the target obstacle boundary line are parallel is determined by the size relationship between the slope difference and the tolerance threshold, which can reduce the position change trend judgment error caused by sensor error, and the higher the sensor accuracy, the lower the tolerance threshold.

[0114] S240: Obtain the parking space type of the target parking space based on the position information and the boundary information, the parking space type being a narrow parking space or a non-narrow parking space.

[0115] S250: If the target parking space is a narrow parking space, perform automatic parking in a head-in manner.

[0116] The automatic parking method provided in the embodiment can confirm whether the target parking space is a narrow parking space after obtaining the information of the obstacle itself and the position information of the target parking space, and perform automatic parking in a head-in manner when the target parking space is a narrow parking space, thereby reducing the parking path length and the parking frequency, and further reducing the parking error, improving the accuracy and safety of parking. In the embodiment, the stability of obstacle boundary line detection can be improved by improving the Hough transform to obtain multiple obstacle boundary lines, thereby improving the accuracy of the parking space type judgment of the target parking space.

[0117] Referring to FIG. 11, an automatic parking method provided in an embodiment of the present application includes:

[0118] S310: Obtain the echo time difference corresponding to the ultrasonic wave.

[0119] The echo time difference can refer to the time difference between the time when the ultrasonic sensing system (such as an ultrasonic sensor, an ultrasonic radar, etc.) of the vehicle sends a transmission wave and the time when the reflected wave is received.

[0120] As a way, after the APA function of the vehicle is started, the ultrasonic sensing system of the vehicle can start to obtain the echo time difference corresponding to the ultrasonic wave.

[0121] S320: Obtain the parking space information of the vehicle based on the echo time difference.

[0122] As a way, the distance between the vehicle and the obstacle can be obtained based on the echo time difference, and the parking space information can be obtained based on the distance between the vehicle and the obstacle.

[0123] As an example, as shown in the left part of FIG. 12, the vehicle can send a transmitting wave through the ultrasonic sensing system, and receive a reflected wave, wherein the parking spaces 2-1, 2-5 and 2-6 are occupied by other vehicles, and thus, as shown in the right part of FIG. 12, the echo time difference between the transmitting wave transmitted to the parking spaces 2-1, 2-5 and 2-6 and the corresponding reflected wave is smaller than the echo time difference of the parking spaces 2-2, 2-3 and 2-4.

[0124] In a possible implementation, if the parking space information indicates that the available space for the vehicle during the parking process is limited, the automatic parking is performed in a head-in parking manner.

[0125] The available space can be understood as a space that can be temporarily used by the vehicle during the parking process, and the available space can be a passageway, a parking space that is not occupied by a vehicle, or the like. The available space being limited can be understood as that the area of the available space is smaller than a preset area, or the available space is in an aggregated distribution, which can be understood as that the connectivity between the available spaces is high, for example, as shown in FIG. 12, the available spaces are a passageway and a connected region formed by the parking spaces 2-2, 2-3 and 2-4.

[0126] As a manner, if the parking space information indicates that the available space for the vehicle during the parking process is limited, the automatic parking can be directly performed in a head-in parking manner.

[0127] In the embodiments of the present application, the automatic parking is directly performed based on the parking strategy, which can accelerate the speed of the automatic parking and save the time of the user.

[0128] As another manner, if the parking space information indicates that the available space for the vehicle during the parking process is limited, the recommendation information can be sent to the user, the recommendation information can be used to recommend the user to perform the automatic parking in a head-in parking manner, and if a confirmation operation of the user indicating an agreement to the recommendation information is received, the automatic parking is performed in a head-in parking manner.

[0129] The confirmation operation of the user indicating an agreement to the recommendation information can be voice information indicating an agreement of the user, or a touch operation of the user on the central control display screen indicating an agreement, or a gesture of the user indicating an agreement, and the like.

[0130] In the embodiments of the present application, when the parking space information indicates that the available space for the vehicle during the parking process is limited, the recommendation information is sent to the user, and after the confirmation operation of the user indicating an agreement to the recommendation information is received, the automatic parking is performed, which can improve the selection right of the user, increase the interactivity between the user and the vehicle, and improve the use experience of the user.

[0131] Optionally, if the parking space information indicates that the available space for the vehicle is not limited, the automatic parking is performed in a head-in parking manner.

[0132] As a manner, the target parking space can be determined to be a narrow parking space based on the steps S110-S130 or the steps S210-S250, and the parking strategy can be generated after the parking space available for the vehicle is determined to be limited based on the steps S310-S330.

[0133] Optionally, as shown in FIG. 13, if the target parking space is a narrow parking space and the parking space information indicates that the parking space available for the vehicle is limited, the automatic parking can be performed in a manner of parking with the front end of the vehicle.

[0134] Optionally, as shown in FIG. 14, if the target parking space is a non-narrow parking space or the parking space information indicates that the parking space available for the vehicle is not limited, the automatic parking can be performed in a manner of parking with the rear end of the vehicle.

[0135] In the embodiments of the present application, when the target parking space is a narrow parking space and the parking space available for the vehicle is limited, the automatic parking is performed in a manner of parking with the front end of the vehicle, thereby reducing the length of the parking path and the number of parking cranks, and further reducing the parking error, and improving the accuracy and safety of the parking.

[0136] Optionally, the parking space information obtained by the visual perception system in the step S110 can further include the arrangement type of the parking space, such as vertical, parallel or diagonal arrangement, wherein, in the vertical and parallel arrangement types, the target parking space can be a rectangular parking space; and in the diagonal arrangement type, the target parking space can be a non-rectangular parking space, such as a parallelogram parking space. When the target parking space is a rectangular parking space, the first parking strategy can be to plan a parking path in a manner of parking with the front end of the vehicle, and the second parking strategy can be to plan a parking path in a manner of parking with the rear end of the vehicle; and when the target parking space is a non-rectangular parking space, the first parking strategy can be to plan a parking path in the manner of parking with the front end of the vehicle based on the determined front end parking angle, and the second parking strategy can be to plan a parking path in the manner of parking with the rear end of the vehicle based on the determined rear end parking angle. The front end parking angle and the rear end parking angle can be determined based on the position information of the vehicle and the position information of the target parking space.

[0137] In the embodiments of the present application, different parking strategies can be obtained for target parking spaces of different shapes, thereby improving the accuracy of parking for target parking spaces of different shapes, and increasing the adaptability and flexibility of the automatic parking method proposed in the present application.

[0138] Optionally, if the target parking space is a narrow parking space and the parking space information indicates that the parking space available for the vehicle is limited, the user can be pushed with suggestion information indicating to change the target parking space, and when an operation indicating that the user agrees to the suggestion information is received, the information of all the parking spaces and obstacles themselves is displayed to the user in the form of an image, so that the user can reselect the target parking space, and then the scheme proposed in the present application is executed based on the new target parking space.

[0139] In the embodiment of the present application, when the type of the target parking space is a narrow parking space and the parking space information indicates that the available parking space is limited, the user is pushed the suggestion information indicating to change the target parking space, which can increase the interaction between the vehicle and the user and improve the user experience.

[0140] Optionally, whether to perform automatic parking in the original target parking space in the way of parking with the front end of the vehicle or to replan the parking strategy based on the new target parking space can be determined based on the limited degree of the available parking space. If the limited degree of the available parking space is equal to or higher than a preset limited degree, the user is pushed the suggestion information indicating to change the target parking space; if the limited degree of the available parking space is lower than the preset limited degree, the parking strategy corresponding to the original target parking space can be obtained.

[0141] Optionally, when it is determined whether the available parking space is limited based on the area of the available parking space, the limited degree of the available parking space can be obtained based on a first preset mapping relationship, which can indicate the mapping relationship between the area of the available parking space and the limited degree of the available parking space. For example, when the area of the available parking space is greater than 0 square meters and less than 2 square meters, the limited degree of the available parking space is high; when the area of the available parking space is greater than or equal to 2 square meters and less than 4 square meters, the limited degree of the available parking space is medium; and when the area of the available parking space is greater than or equal to 4 square meters, the limited degree of the available parking space is low.

[0142] Optionally, when it is determined whether the available parking space is limited based on the distribution of the available parking space, the limited degree of the available parking space can be obtained based on a second preset mapping relationship, which can indicate the mapping relationship between the distribution of the available parking space and the limited degree of the available parking space.

[0143] The automatic parking method provided in the embodiment can obtain the parking space information of the vehicle based on the echo time difference of the ultrasonic wave after obtaining the echo time difference of the ultrasonic wave, and perform automatic parking in the way of parking with the front end of the vehicle if the parking space information indicates that the available parking space of the vehicle is limited during parking. In this way, automatic parking can be performed in the way of parking with the front end of the vehicle if the parking space information indicates that the available parking space of the vehicle is limited during parking, so as to reduce the length of the parking path and the number of parking cranks, and further reduce the parking error and improve the accuracy and safety of parking.

[0144] Referring to FIG. 15, the automatic parking device 600 provided in the present application includes:

[0145] The information acquisition unit 610 is configured to acquire the information of the obstacle itself and the position information of the target parking space.

[0146] The parking space type acquisition unit 620 is configured to determine whether the target parking space is a narrow parking space based on the information of the obstacle itself and the position information of the target parking space, the narrow parking space being a parking space in which at least two parking boundary lines are constrained by a target obstacle.

[0147] The parking strategy generation unit 630 is configured to perform automatic parking in a head-in manner if the target parking space is a narrow parking space.

[0148] As a manner, the parking space type acquisition unit 620 is specifically configured to obtain boundary information of a target obstacle boundary line based on the information of the obstacle itself and the position information, the target obstacle boundary line being an obstacle boundary line that has a parking space constraint on the target parking space; and determine whether the target parking space is a narrow parking space based on the position information and the boundary information.

[0149] Optionally, the parking space type acquisition unit 620 is specifically configured to perform smoothing operation on the information of the obstacle itself to obtain a reference obstacle boundary line, the reference obstacle boundary line being a boundary line formed by an obstacle of a specified category; and obtain the boundary information of the target obstacle boundary line based on the position information and the reference obstacle boundary line.

[0150] Optionally, the information of the obstacle itself includes position information of a plurality of obstacles in a first coordinate system and category information of the plurality of obstacles, and the parking space type acquisition unit 620 is specifically configured to perform smoothing operation on the position information of the plurality of obstacles in the first coordinate system to obtain a plurality of obstacle boundary lines; and obtain the reference obstacle boundary line based on the category information of the plurality of obstacles and the plurality of obstacle boundary lines.

[0151] Optionally, the parking space type acquisition unit 620 is specifically configured to obtain a two-dimensional sinusoidal curve of each of the plurality of obstacles in a second coordinate system based on the position information of each of the plurality of obstacles in the first coordinate system, so as to obtain a plurality of two-dimensional sinusoidal curves; obtain a grid space of the plurality of two-dimensional sinusoidal curves based on a value range and a definition range of the plurality of two-dimensional sinusoidal curves; perform discretization processing on the plurality of two-dimensional sinusoidal curves in the grid space to obtain a plurality of feature points; and obtain the plurality of obstacle boundary lines based on the plurality of feature points.

[0152] Optionally, the grid space includes a plurality of grid points, and the parking space type acquisition unit 620 is specifically configured to obtain a contribution value of each of the plurality of grid points based on the plurality of feature points; obtain a target grid point based on the contribution value of each of the plurality of grid points, the target grid point being a grid point whose corresponding contribution value ranks in the first K positions or a grid point whose corresponding contribution value is greater than a preset value, the contribution value representing a probability that an obstacle is located at the corresponding grid point; and obtain the plurality of obstacle boundary lines based on the target grid point.

[0153] Optionally, the parking space type acquisition unit 620 is specifically configured to obtain a three-dimensional distribution curve of each of the feature points in the second coordinate system based on position information of each of the feature points in the second coordinate system, the three-dimensional distribution curve representing an energy distribution of the corresponding feature point in the second coordinate system; obtain a reference grid point corresponding to each of the feature points, the reference grid point being a vertex of a grid in which the feature point is located; obtain a contribution degree curve of the reference grid point corresponding to each of the feature points based on a distance between each of the feature points and the corresponding reference grid point and the three-dimensional distribution curve corresponding to each of the feature points, to obtain a contribution degree curve of the reference grid point corresponding to each of the plurality of feature points, the contribution degree curve representing a contribution degree distribution of the reference grid point of the corresponding feature point in the second coordinate system; and obtain the contribution value of each of the plurality of grid points based on the contribution degree curve of the reference grid point corresponding to each of the plurality of feature points.

[0154] Optionally, the position information includes position information of a target parking space boundary line, and the parking space type acquisition unit 620 is specifically configured to obtain a target obstacle boundary line based on the position information of the target parking space boundary line and the reference obstacle boundary line, the target obstacle boundary line being a reference obstacle boundary line whose distance from the target parking space boundary line is within a preset distance value range; and obtain boundary information of the target obstacle boundary line based on a distance between the target parking space boundary line and the target obstacle boundary line and a slope relationship between the target parking space boundary line and the target obstacle boundary line.

[0155] Optionally, the position information includes position information of a target parking space boundary line, and the boundary information includes a position relationship and a position change trend between the target obstacle boundary line and the target parking space boundary line, and the parking space type acquisition unit 620 is specifically configured to obtain a target parking space area based on the position information of the target parking space boundary line; obtain an area of a constraint space formed by the target obstacle boundary line based on the position relationship and the position change trend between the target obstacle boundary line and the target parking space boundary line; obtain a constraint feature based on the target parking space area and the area of the constraint space; and determine whether the target parking space is a narrow and long parking space based on the constraint feature.

[0156] Optionally, the parking space type obtaining unit 620 is specifically configured to: obtain an intersection area based on the target parking space area and the constraint space area, the intersection area being an area that belongs to both the target parking space and the constraint space; and obtain the constraint feature based on a ratio of the intersection area to the constraint space area.

[0157] Optionally, the parking space type obtaining unit 620 is specifically configured to: if the ratio of the intersection area to the constraint space area is greater than a first preset value and less than a second preset value, determine that the constraint feature is a double-side constraint; if the ratio of the intersection area to the constraint space area is greater than or equal to the second preset value and less than a third preset value, determine that the constraint feature is a three-side constraint; and if the ratio of the intersection area to the constraint space area is a fourth preset value, determine that the constraint feature is no constraint or a single-side constraint.

[0158] Optionally, the parking space type obtaining unit 620 is specifically configured to: if the constraint feature is the double-side constraint or the three-side constraint, determine that the target parking space is a narrow parking space; and if the constraint feature is the single-side constraint or the no constraint, determine that the target parking space is a non-narrow parking space.

[0159] As a manner, the parking strategy generating unit 430 is specifically configured to: if the parking space type of the target parking space is the non-narrow parking space, perform automatic parking in a manner of tail-in.

[0160] As a manner, the parking strategy generating unit 430 is specifically configured to: if the target parking space is the narrow parking space, send recommendation information to a user, the recommendation information being used to recommend to the user to perform automatic parking in a manner of head-in; and if a confirmation operation representing agreement to the recommendation information is received from the user, perform automatic parking in the manner of head-in.

[0161] Referring to FIG. 16, an automatic parking device 800 provided by the present application includes:

[0162] An information obtaining unit 810 is configured to obtain an echo time difference corresponding to ultrasonic waves, the echo time difference representing a distance between a vehicle and an obstacle;

[0163] A parking space information obtaining unit 820 is configured to obtain parking space information of the vehicle based on the echo time difference;

[0164] A parking strategy generating unit 830 is configured to perform automatic parking in a manner of head-in if the parking space information represents that a borrowable space of the vehicle during parking is limited.

[0165] As a manner, the parking space information obtaining unit 820 is specifically configured to obtain the distance between the vehicle and the obstacle based on the echo time difference; and obtain the parking space information based on the distance between the vehicle and the obstacle.

[0166] As a manner, the parking strategy generating unit 830 is specifically configured to perform automatic parking in a manner of parking at the back end if the parking space information represents that the available parking space of the vehicle is not limited.

[0167] Next, a vehicle provided by the present application will be described in combination with FIG. 17.

[0168] Referring to FIG. 17, based on the automatic parking method and device described above, the present application further provides another vehicle 100 that can perform the automatic parking method described above. The vehicle 100 comprises a processor 102, a memory 104 and a data acquisition module 106. The memory 104 stores a program that can perform the content of the foregoing embodiments, and the processor 102 can execute the program stored in the memory 104.

[0169] The processor 102 can comprise one or more processing cores. The processor 102 connects various parts in the vehicle 100 through various interfaces and lines, executes instructions, programs, code sets or instruction sets stored in the memory 104, and calls data stored in the memory 104, to perform various functions and process data of the vehicle 100. Optionally, the processor 102 can be implemented in at least one of the following hardware forms: a network processor (NPU), a digital signal processor (DSP), a field-programmable gate array (FPGA), and a programmable logic array (PLA). The processor 102 can be integrated with a combination of one or more of the following: a central processing unit (CPU), a graphics processor (GPU), a network processing unit (NPU), and a modem. The CPU is mainly used to process operating systems, user interfaces, and application programs; the GPU is used to render and draw display content; the NPU is used to process multimedia data such as videos and images; and the modem is used to process wireless communication. It can be understood that the modem can also not be integrated into the processor 102, but can be implemented by a separate communication chip.

[0170] The memory 104 can include a random access memory (RAM), and can also include a read-only memory (ROM) and a double data rate (DDR). The memory 104 can be used to store instructions, programs, codes, code sets or instruction sets. The memory 104 can include a program storage area and a data storage area, where the program storage area can store instructions for implementing an operating system, instructions for implementing at least one function (such as a touch function, a sound playing function, an image playing function, etc.), instructions for implementing each of the method embodiments described below, etc. The data storage area can also store data created by the vehicle 100 in use (such as a phone book, audio and video data, chat record data), etc.

[0171] The data acquisition module 106 can include, but is not limited to, a camera, a radar, a level, a light sensor, a motion sensor, a pressure sensor, an infrared thermal sensor, a distance sensor, an acceleration sensor, and other sensors. Among them, the camera can acquire images of the environment around the vehicle 100, and the radar can acquire ultrasonic information around the vehicle 100.

[0172] Among them, the pressure sensor can detect the pressure generated by pressing on the vehicle 100. That is, the pressure sensor detects the pressure generated by the contact or pressing between the user and the vehicle 100, for example, the pressure generated by the contact or pressing between the user's hand and the vehicle 100. Therefore, the pressure sensor can be used to determine whether contact or pressing occurs between the user and the vehicle 100, and the size of the pressure.

[0173] Among them, the acceleration sensor can detect the size of the acceleration in each direction (generally three axes), and can detect the size and direction of gravity when at rest, which can be used for applications such as vehicle 100 posture calibration (such as a magnetometer), vibration recognition related functions (such as a pedometer, a knock), etc. In addition, the vehicle 100 can also be configured with a gyroscope, a barometer, a hygrometer, a thermometer and other sensors, which will not be described here.

[0174] The embodiment of the present application provides a computer readable storage medium. The computer readable storage medium stores program code, and the program code can be called by a processor to execute the method described in the above method embodiments.

[0175] The computer-readable storage medium can be an electronic storage, such as a flash memory, EEPROM (Electrically Erasable Programmable Read-Only Memory), EPROM, hard disk, or ROM. Alternatively, the computer-readable storage medium includes a non-transitory computer-readable storage medium. The computer-readable storage medium has storage space for storing program codes for performing any of the method steps described above. The program codes can be read from or written to one or more computer program products. The program codes can be compressed in an appropriate form, for example.

[0176] In summary, the automatic parking method, device and vehicle provided in the application can determine whether the target parking space is a narrow parking space based on the information of the obstacle itself and the position information of the target parking space after the information of the obstacle itself and the position information of the target parking space are obtained, and perform automatic parking in the manner of head-in parking if the target parking space is a narrow parking space. In this way, it can be determined whether the target parking space is a narrow parking space after the information of the obstacle itself and the position information of the target parking space are obtained, and automatic parking is performed in the manner of head-in parking if the target parking space is a narrow parking space, so that the length of the parking path and the number of parking times are reduced, and the parking error is reduced, and the accuracy and safety of parking are improved.

[0177] Finally, it should be noted that: the above examples are only used to illustrate the technical solutions of the application, and not to limit them; although the application has been described in detail with reference to the foregoing examples, those skilled in the art will understand that they can still modify the technical solutions recorded in the foregoing examples, or make equivalent replacement for part of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the application.

Claims

1. An automatic parking method characterized by, The method comprises: obtaining information of the obstacles and position information of the target parking space; determining whether the target parking space is a narrow parking space based on the information of the obstacles and the position information of the target parking space, the narrow parking space being a parking space in which at least two parking boundary lines are constrained by target obstacles; if the target parking space is a narrow parking space, performing automatic parking in a front-end parking manner.

2. The method of claim 1, wherein, The determination of whether the target parking space is a narrow parking space based on the information of the obstacles and the position information of the target parking space comprises: obtaining boundary information of a target obstacle boundary line based on the information of the obstacles and the position information, the target obstacle boundary line being an obstacle boundary line that constrains the parking space of the target parking space; determining whether the target parking space is a narrow parking space based on the position information and the boundary information.

3. The method of claim 2, wherein, The obtaining of the boundary information of the target obstacle boundary line based on the information of the obstacles and the position information comprises: performing smoothing operation on the information of the obstacles to obtain a reference obstacle boundary line, the reference obstacle boundary line being a boundary line formed by obstacles of a specified category; obtaining the boundary information of the target obstacle boundary line based on the position information and the reference obstacle boundary line.

4. The method of claim 3, wherein, The information of the obstacles comprises position information of a plurality of obstacles in a first coordinate system and category information of the plurality of obstacles, and the smoothing operation on the information of the obstacles to obtain the reference obstacle boundary line comprises: performing smoothing operation on the position information of the plurality of obstacles in the first coordinate system to obtain a plurality of obstacle boundary lines; obtaining the reference obstacle boundary line based on the category information of the plurality of obstacles and the plurality of obstacle boundary lines.

5. The method of claim 4, wherein, The smoothing operation on the position information of the plurality of obstacles in the first coordinate system to obtain the plurality of obstacle boundary lines comprises: obtaining two-dimensional sinusoidal curves of the plurality of obstacles in a second coordinate system based on the position information of the plurality of obstacles in the first coordinate system, to obtain a plurality of two-dimensional sinusoidal curves; obtaining a grid space of the plurality of two-dimensional sinusoidal curves based on the value range and the definition range of the plurality of two-dimensional sinusoidal curves; performing discretization processing on the plurality of two-dimensional sinusoidal curves in the grid space to obtain a plurality of feature points; obtaining the plurality of obstacle boundary lines based on the plurality of feature points.

6. The method of claim 5, wherein, The grid space comprises a plurality of grid points, and the obtaining of the plurality of obstacle boundary lines based on the plurality of feature points comprises: obtaining a contribution value of each of the plurality of grid points based on the plurality of feature points; obtaining a target grid point based on the contribution value of each of the plurality of grid points, the target grid point being a grid point whose corresponding contribution value ranks in the top K positions or a grid point whose corresponding contribution value is greater than a preset value, the contribution value representing a probability that an obstacle is located at the corresponding grid point; obtaining the plurality of obstacle boundary lines based on the target grid point.

7. The method of claim 6, wherein, The obtaining of the contribution value of each of the plurality of grid points based on the plurality of feature points comprises: Based on the position information of each feature point in the second coordinate system, a three-dimensional distribution curve of each feature point in the second coordinate system is obtained, and the three-dimensional distribution curve represents the energy distribution of the corresponding feature point in the second coordinate system; A reference grid point corresponding to each feature point is obtained, and the reference grid point is a vertex of a grid where the feature point is located; Based on the distance between each feature point and the corresponding reference grid point and the three-dimensional distribution curve corresponding to each feature point, a contribution degree curve of the reference grid point corresponding to each feature point is obtained, so as to obtain the contribution degree curves of the reference grid points corresponding to the plurality of feature points, and the contribution degree curve can represent the contribution degree distribution of the reference grid point of the corresponding feature point in the second coordinate system; Based on the contribution degree curves of the reference grid points corresponding to the plurality of feature points, the contribution values of the plurality of grid points are obtained.

8. The method of claim 3, wherein, The position information includes position information of a target parking space boundary line, and the boundary information includes position relationship and position change trend of the target obstacle boundary line and the target parking space boundary line, and the target parking space is determined to be a narrow and long parking space based on the position information and the boundary information, including: Based on the position information of the target parking space boundary line and the reference obstacle boundary line, a target obstacle boundary line is obtained, and the target obstacle boundary line is a reference obstacle boundary line with a distance from the target parking space boundary line within a preset distance value range; Based on the distance between the target parking space boundary line and the target obstacle boundary line and the slope relationship between the target parking space boundary line and the target obstacle boundary line, boundary information of the target obstacle boundary line is obtained.

9. The method of claim 2, wherein, The position information includes position information of a target parking space boundary line, and the boundary information includes position relationship and position change trend of the target obstacle boundary line and the target parking space boundary line, and the target parking space is determined to be a narrow and long parking space based on the position information and the boundary information, including: Based on the position information of the target parking space boundary line, the area of the target parking space is obtained; Based on the position relationship and position change trend of the target obstacle boundary line and the target parking space boundary line, the area of the constraint space formed by the target obstacle boundary line is obtained; Based on the target parking space area and the constraint space area, a constraint feature is obtained; Based on the constraint feature, it is determined whether the target parking space is a narrow and long parking space.

10. The method of claim 9, wherein, Based on the target parking space area and the constraint space area, an intersection area is obtained, and the intersection area is an area belonging to both the target parking space and the constraint space; Based on the ratio of the intersection area and the constraint space area, the constraint feature is obtained. Based on the ratio of the intersection area and the constraint space area, the constraint feature is obtained, including:

11. The method of claim 10, wherein, If the ratio of the intersection area and the constraint space area is greater than a first preset value and less than a second preset value, it is determined that the constraint feature is a double-side constraint; If the ratio of the intersection area and the constraint space area is greater than or equal to a second preset value and less than a third preset value, it is determined that the constraint feature is a three-side constraint; ​ If a ratio of the intersection area and the constraint space area is a fourth preset value, it is determined that the constraint feature is unconstrained or single-side constrained.

12. The method of claim 9, wherein, The method further includes: If the constraint feature is double-side constrained or three-side constrained, it is determined that the target parking space is a narrow parking space. If the constraint feature is single-side constrained or unconstrained, it is determined that the target parking space is a non-narrow parking space.

13. The method of any one of claims 1-12, wherein, The method further includes: If the target parking space is a non-narrow parking space, automatic parking is performed in a manner of tail-in.

14. The method of any one of claims 1-12, wherein, The method further includes: If the target parking space is a narrow parking space, recommendation information is sent to a user, the recommendation information being used to recommend to the user to perform automatic parking in a manner of head-in. If a confirmation operation representing agreement of the user to the recommendation information is received, automatic parking is performed in a manner of head-in.

15. An automatic parking method characterized by, The method includes: An echo time difference corresponding to an ultrasonic wave is obtained. Parking space information of the vehicle is obtained based on the echo time difference. If the parking space information represents that a borrowable space of the vehicle is limited during parking, automatic parking is performed in a manner of head-in.

16. The method of claim 15, wherein, The method further includes: If the parking space information represents that the borrowable space of the vehicle is not limited, automatic parking is performed in a manner of tail-in. The apparatus includes:

17. The method according to claim 15 or 16, characterized in that, An information obtaining unit is configured to obtain information of an obstacle itself and position information of a target parking space. A parking space type obtaining unit is configured to determine whether the target parking space is a narrow parking space based on the information of the obstacle itself and the position information of the target parking space, the narrow parking space being a parking space in which at least two parking boundary lines are constrained by a target obstacle.

18. An automatic parking apparatus characterized by comprising: A parking strategy generating unit is configured to perform automatic parking in a manner of head-in if the target parking space is a narrow parking space. The apparatus includes: An information obtaining unit is configured to obtain an echo time difference corresponding to an ultrasonic wave, the echo time difference representing a distance between a vehicle and an obstacle. A parking space information obtaining unit is configured to obtain parking space information of the vehicle based on the echo time difference.

19. An automatic parking apparatus characterized by comprising: A parking strategy generating unit is configured to perform automatic parking in a manner of head-in if the parking space information represents that a borrowable space of the vehicle is limited during parking. One or more processors and a memory are included; One or more programs are stored in the memory and configured to be executed by the one or more processors, the one or more programs being configured to perform the method of any one of claims 1-14 or 15-17. A computer readable storage medium has program codes stored therein, wherein the method of any one of claims 1-14 or 15-17 is performed when the program codes are run.

20. A vehicle characterized by ​ ​ 21. A computer-readable storage medium, characterized in that, ​

Images