Method and apparatus for automatic parking

Abstract

An automatic parking method comprises: when it is determined that a parking area of a target vehicle is a non-standard parking space, obtaining a three-dimensional point cloud map and surrounding environment image information of the surrounding environment of the target vehicle; performing two-dimensional projection on the three-dimensional point cloud map; performing quadrilateral graphic fitting on the two-dimensional projection through a preset maximum concave hull model; performing graphic matching on a preset geometric model of the target vehicle and the fitted quadrilateral graphic; when the matching is successful, performing parking space type estimation on the parking area according to the two-dimensional projection and the surrounding environment image information; performing path planning on the target vehicle according to the parking space type, and parking the target vehicle in the parking area according to the planned path.

Description

Technical Field

[0001] This article relates to intelligent driving technology, and in particular to a method and device for automatic parking. Background Technology

[0002] Currently, automatic parking technology requires clear parking space information for the system to plan its route. For non-standard parking spaces (including parking areas without parking lines and situations where other drivers occupy other parking spaces due to improper parking), standard parking spaces are projected onto the open ground using laser projection, and then the vehicle is parked. Summary of the Invention

[0003] This application provides a method and apparatus for automatic parking. It solves the problem of automatic parking in non-standard parking spaces by employing a non-laser projection method.

[0004] This application provides an automatic parking method, including:

[0005] When the parking area of ​​the target vehicle is determined to be a non-standard parking space, a three-dimensional point cloud map of the surrounding environment of the target vehicle and surrounding environment image information are obtained.

[0006] The three-dimensional point cloud map is projected into a two-dimensional shape; the two-dimensional projection is fitted with a quadrilateral shape using a preset maximum concave hull model.

[0007] Perform graphic matching between the preset geometric model of the target vehicle and the fitted quadrilateral shape;

[0008] Once a match is successful, the parking space type of the waiting area is estimated based on the two-dimensional projection and the surrounding environment image information.

[0009] Based on the parking space type, a path is planned for the target vehicle, and the target vehicle is parked in the waiting area according to the planned path.

[0010] In one exemplary embodiment, the non-standard parking space is determined as follows:

[0011] If a parking area is found to have no parking space lines or partially covered by parking space lines, the parking area is determined to be a non-standard parking area.

[0012] In one exemplary embodiment, the step of fitting a quadrilateral shape to the two-dimensional projection using a preset maximum hull model includes:

[0013] The maximum concave polygon geometry of the two-dimensional projection is drawn using a preset maximum concave hull model.

[0014] The maximum concave polygonal geometry is approximated as a quadrilateral.

[0015] In one exemplary embodiment, the step of performing graphic matching between the preset geometric model of the target vehicle and the fitted quadrilateral graphic includes:

[0016] When the area of ​​the fitted quadrilateral is greater than the area of ​​the geometric model of the target vehicle and the side length of the shortest side of the fitted quadrilateral is greater than the width of the target vehicle, the search nodes are expanded outward in the form of a random search tree with the geometric center of the quadrilateral as the origin and the sides of the quadrilateral as the boundaries.

[0017] For each searched node, perform the following operations until the searched node meets the preset conditions: For each searched node, substitute the coordinates of the geometric center point of the target vehicle's geometric model into the node coordinates. When the geometric center of the target vehicle's geometric model is located at the current node, determine whether the quadrilateral shape intersects with the target vehicle's geometric model. If there is an intersection, expand the search for the next node outwards using a random search tree. The preset conditions include that the quadrilateral shape has no intersection with the target vehicle's geometric model or that all nodes within the boundary have been traversed.

[0018] When it is determined that the quadrilateral shape has no intersection with the geometric model of the target vehicle, then the preset geometric model of the target vehicle and the fitted quadrilateral shape are matched.

[0019] In one exemplary embodiment, after a successful match, the parking space type of the waiting area is estimated based on the two-dimensional projection and surrounding environment image information. This estimation includes:

[0020] The location information of surrounding vehicles of the target vehicle is determined based on the surrounding environment image information;

[0021] The pose of the target vehicle and its current state in the environment are determined based on the location information of the surrounding vehicles.

[0022] The state of the surrounding vehicles is estimated based on the pose of the target vehicle, the state of the current environment, the position information of the surrounding vehicles, and the two-dimensional projection.

[0023] The parking space type of the waiting area is estimated based on the state of the surrounding vehicles.

[0024] In one exemplary embodiment, the route planning for the target vehicle based on the parking space type includes:

[0025] The target vehicle is planned using the hybrid A* algorithm.

[0026] In one exemplary embodiment, it further includes:

[0027] Gravity constraints are used in the path planning process.

[0028] In one exemplary embodiment, before parking the target vehicle into the waiting area according to the planned path, the process includes:

[0029] The vehicle's infotainment system simulates the target vehicle parking in the designated parking area.

[0030] This application provides an automatic parking device, including a memory and a processor, comprising:

[0031] The memory is used to store the program for automatic parking;

[0032] The processor is used to read and execute the program for automatic parking and execute the automatic parking method described above.

[0033] Other features and advantages of this application will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing the application. Other advantages of this application can be realized and obtained by means of the solutions described in the description and the accompanying drawings. Attached Figure Description

[0034] The accompanying drawings are used to provide an understanding of the technical solutions of this application and constitute a part of the specification. They are used together with the embodiments of this application to explain the technical solutions of this application and do not constitute a limitation on the technical solutions of this application.

[0035] Figure 1 This is a schematic diagram of an automatic parking method according to an embodiment of this application;

[0036] Figure 2 This is a schematic diagram of another automatic parking method according to an embodiment of this application;

[0037] Figure 3 This is a functional module diagram of another automatic parking method according to an embodiment of this application;

[0038] Figure 4 This is a schematic diagram of a non-standard parking space with parking lines, according to an embodiment of this application.

[0039] Figure 5 for Figure 4 The diagram shows the matching of non-standard parking spaces.

[0040] Figure 6 This is a schematic diagram of a non-standard parking space with parking lines, according to an embodiment of this application.

[0041] Figure 7This is a schematic diagram of a non-standard parking space without parking lines, according to an embodiment of this application.

[0042] Figure 8 for Figure 7 The diagram shows the matching of non-standard parking spaces.

[0043] Explanation of icon numbers:

[0044] 40 represents the vehicle's geometric model;

[0045] 42 represents a parking space graphic that is obscured;

[0046] 50 represents a polygonal geometric figure;

[0047] 52 represents the vehicle's geometric model;

[0048] 60 represents the vehicle's geometric model;

[0049] 62 represents a parking space graphic that is obscured;

[0050] 70 represents the vehicle's geometric model;

[0051] 80 represents a polygonal geometric figure;

[0052] 82 represents the vehicle's geometric model. Detailed Implementation

[0053] Figure 1 This is a schematic diagram of an automatic parking method according to an embodiment of this application, as shown below. Figure 1 As shown, the automatic parking method of this embodiment includes steps S11-S15:

[0054] S11. When it is determined that the parking area of ​​the target vehicle is a non-standard parking space, obtain a three-dimensional point cloud map and image information of the surrounding environment of the target vehicle.

[0055] S12. Project the three-dimensional point cloud map into a two-dimensional shape; fit the two-dimensional projection into a quadrilateral shape using a preset maximum concave hull model;

[0056] S13. Perform graphic matching between the preset geometric model of the target vehicle and the fitted quadrilateral shape;

[0057] S14. After a successful match, estimate the parking space type for the waiting area;

[0058] S15. Calculate the path for the target vehicle according to the parking space type, and park the target vehicle in the waiting area according to the planned path.

[0059] In one exemplary embodiment, the non-standard parking space is determined as follows:

[0060] If a parking area is found to have no parking space lines or partially covered by parking space lines, the parking area is determined to be a non-standard parking area.

[0061] In one exemplary embodiment, the step of fitting a quadrilateral shape to the two-dimensional projection using a preset maximum hull model includes:

[0062] The maximum concave polygon geometry of the two-dimensional projection is drawn using a preset maximum concave hull model.

[0063] The maximum concave polygonal geometry is approximated as a quadrilateral.

[0064] In one exemplary embodiment, the step of performing graphic matching between the preset geometric model of the target vehicle and the fitted quadrilateral graphic includes:

[0065] When the area of ​​the fitted quadrilateral is greater than the area of ​​the geometric model of the target vehicle and the side length of the shortest side of the fitted quadrilateral is greater than the width of the target vehicle, the search nodes are expanded outward in the form of a random search tree with the geometric center of the quadrilateral as the origin and the sides of the quadrilateral as the boundaries.

[0066] For each searched node, perform the following operations until the searched node meets the preset conditions: For each searched node, substitute the coordinates of the geometric center point of the target vehicle's geometric model into the node coordinates. When the geometric center of the target vehicle's geometric model is located at the current node, determine whether the quadrilateral shape intersects with the target vehicle's geometric model. If there is an intersection, expand the search for the next node outwards using a random search tree. The preset conditions include that the quadrilateral shape intersects with the target vehicle's geometric model and that all nodes within the boundary have been traversed.

[0067] When it is determined that the quadrilateral shape has no intersection with the geometric model of the target vehicle, then the preset geometric model of the target vehicle and the fitted quadrilateral shape are matched.

[0068] The geometric model of the target vehicle can be rectangular.

[0069] In simple terms, a random search tree is based on a single point and then explores outwards like a vine climbing. A random search tree can define a step size to expand to the next node.

[0070] The intersection of the quadrilateral and the geometric model of the target vehicle proves that the two quadrilaterals are not in an inclusion relationship; the absence of intersection proves that the vehicle model is inside the quadrilateral.

[0071] In one exemplary embodiment, after a successful match, the parking space type of the waiting area is estimated based on the two-dimensional projection and surrounding environment image information. This estimation includes:

[0072] The location information of surrounding vehicles of the target vehicle is determined based on the surrounding environment image information;

[0073] The pose of the target vehicle and its current state in the environment are determined based on the location information of the surrounding vehicles.

[0074] The state of the surrounding vehicles is estimated based on the pose of the target vehicle, the state of the current environment, the position information of the surrounding vehicles, and the two-dimensional projection.

[0075] The parking space type of the waiting area is estimated based on the state of the surrounding vehicles.

[0076] The position information of the vehicles surrounding the target vehicle can be reference coordinates relative to the target vehicle's world coordinate system. The current state of the environment includes the target vehicle's heading and its distance relative to the surrounding vehicles.

[0077] Pose is used to consider the target vehicle's position in the surrounding environment and its heading relative to the previous keyframe.

[0078] For example, if a row of vehicles is parked, the current reference parking method, such as reversing into a parking space or parallel parking, can be analyzed based on the pose of the target vehicle, the position information of the surrounding vehicles, and the two-dimensional projection.

[0079] The parking space types here can include horizontal parking spaces and vertical parking spaces.

[0080] In one exemplary embodiment, the route planning for the target vehicle based on the parking space type includes:

[0081] The target vehicle is planned using the hybrid A* algorithm.

[0082] In one exemplary embodiment, the method further includes employing gravity constraints during path planning.

[0083] Gravity constraints are a model that assumes there are attractive and repulsive forces between the target object and obstacles, and plans the path based on the equilibrium point of these forces. Introducing a gravity model into the A* algorithm can address the shortcomings in A*'s path planning.

[0084] In one exemplary embodiment, before parking the target vehicle into the waiting area according to the planned path, the process includes:

[0085] The vehicle's infotainment system simulates the target vehicle parking in the designated parking area.

[0086] The automatic parking method of this application embodiment performs two-dimensional projection of a three-dimensional point cloud map of the surrounding environment of the target vehicle; fits the two-dimensional projection with a quadrilateral shape using a preset maximum concave hull model; performs graphic matching between the preset geometric model of the target vehicle and the fitted quadrilateral shape; when the matching is successful, estimates the parking space type of the waiting area; performs path planning for the target vehicle according to the parking space type, and parks the target vehicle in the waiting area according to the planned path. This method solves the problem of automatic parking for users in non-standard parking spaces by using a non-laser projection method.

[0087] The automatic parking method described in this application is also applicable to standard parking spaces.

[0088] Figure 2 This is a schematic diagram of another automatic parking method according to an embodiment of this application. Figure 2 The parking method shown includes the following steps:

[0089] 1) The driver drives the vehicle to the parking space and makes a parking judgment;

[0090] If it is a standard parking space, the automatic parking procedure will be activated; if it is a non-standard parking space, the non-standard parking procedure will be activated.

[0091] 2) After starting the non-standard parking space parking procedure, create a map of the environment;

[0092] First, based on the vehicle-mounted LiDAR, a 3D point cloud map of the surrounding environment is acquired and then projected into a 2D model. Second, based on the vehicle-mounted camera (such as an AVM (Around View Monitor) system), image information of the surrounding environment is acquired and vehicle targets are detected.

[0093] 3) Draw the virtual parking spaces, including:

[0094] For non-standard parking spaces with parking lines (such as...) Figure 4 , Figure 6 As shown, this may be due to improper parking by car owners occupying other parking spaces, and non-standard parking spaces without parking lines (such as...). Figure 7 As shown in the figure, before a vehicle parks, a two-dimensional projection of the three-dimensional point cloud map is required. Using the maximum concave hull model, a polygonal geometry with the maximum concave hull is drawn according to the two-dimensional projection model, and the outline of this polygonal geometry is approximated to a quadrilateral, such as... Figure 5 , Figure 8 As shown.

[0095] Figure 4In this context, 40 represents the vehicle's geometric model, and 42 represents the parking space. Figure 5 In the diagram, 50 represents a quadrilateral, and 52 represents the geometric model of the vehicle. Figure 6 In the diagram, 60 represents the geometric model of the vehicle, and 62 represents the parking space. Figure 7 The number 70 in the middle represents the geometric model of the vehicle. Figure 8 In the diagram, 80 represents a quadrilateral, and 82 represents the geometric model of the vehicle.

[0096] 4) Matching the vehicle geometry model with the fitted quadrilateral shape, including:

[0097] Calculate the area of ​​the fitted quadrilateral;

[0098] Calculate whether the shortest side of the quadrilateral is greater than the width of the vehicle. If the shortest side of the quadrilateral is greater than the width of the vehicle and the area is greater than the vehicle's geometric model, then it is considered possible to park in the parking space.

[0099] At this point, the quadrilateral's geometric center is taken as the origin, and its sides as the boundaries. The search nodes are expanded outwards using a random search tree. The coordinates of the vehicle's geometric center are input into the node coordinates. When the vehicle's geometric center is located at the current node, it is determined whether the quadrilateral intersects with the vehicle's geometric model.

[0100] If there are no intersections, it can be determined that the vehicle's geometric model is inside the fitted quadrilateral, and the vehicle can park there.

[0101] If there is an intersection, it is determined that the current vehicle geometry model is not inside the fitted quadrilateral, and the model is expanded to the next node. Then, the intersection is checked again. If there are no valid points within the boundary, the vehicle is considered unable to park.

[0102] 5) Parking space type estimation, including:

[0103] The vehicle detects the position information of surrounding vehicles by vehicle target detection, and determines its own pose and the state of its current environment;

[0104] The state of surrounding vehicles is estimated using the vehicle's own pose information, the position information of surrounding vehicles, and the two-dimensional projection information of the vehicle's surrounding environment.

[0105] The parking space type is estimated based on the status of surrounding vehicles.

[0106] 6) Path planning, including:

[0107] After determining in step 4) that the vehicle can be parked in the non-standard parking space and obtaining the parking space type information in step 5), path planning is then performed.

[0108] A hybrid A* algorithm can be used for vehicle parking path planning;

[0109] At the same time, gravity constraints can be used to increase the smoothness of the path during the parking and driving process.

[0110] 7) Virtual parking simulation, including:

[0111] After the route planning is completed, virtual parking simulation can be performed in the vehicle system to enhance parking reliability.

[0112] This application embodiment designs a parking system for standard parking lots and unmarked parking spaces based on onboard sensors (laser and vision). For non-standard parking lots, it establishes a map environment, preprocesses the environment map, draws parking space models, and performs path planning to enable vehicles to automatically park in non-standard parking spaces.

[0113] This application also discloses an automatic parking device, wherein the memory and processor include:

[0114] The memory is used to store the program for automatic parking;

[0115] The processor is used to read and execute the program for automatic parking and execute the automatic parking method described above.

[0116] This application describes several embodiments, but these descriptions are exemplary and not restrictive, and it will be apparent to those skilled in the art that many more embodiments and implementations are possible within the scope of the embodiments described herein. Although many possible combinations of features are shown in the drawings and discussed in the detailed description, many other combinations of the disclosed features are also possible. Unless specifically limited, any feature or element of any embodiment may be used in combination with, or may replace, any feature or element of any other embodiment.

[0117] This application includes and contemplates combinations of features and elements known to those skilled in the art. The embodiments, features, and elements disclosed in this application may also be combined with any conventional features or elements to form a unique inventive scheme as defined by the claims. Any feature or element of any embodiment may also be combined with features or elements from other inventive schemes to form another unique inventive scheme as defined by the claims. Therefore, it should be understood that any feature shown and / or discussed in this application may be implemented individually or in any suitable combination. Therefore, the embodiments are not limited except by the limitations imposed by the appended claims and their equivalents. Furthermore, various modifications and changes may be made within the scope of the appended claims.

[0118] Furthermore, in describing representative embodiments, the specification may have presented methods and / or processes as a specific sequence of steps. However, the method or process should not be limited to the specific order of steps described herein, to the extent that it does not depend on such a specific order. As will be understood by those skilled in the art, other sequences of steps are also possible. Therefore, the specific order of steps set forth in the specification should not be construed as a limitation of the claims. Moreover, the claims concerning the method and / or process should not be limited to the steps performed in the written order, and those skilled in the art will readily understand that these orders can be varied and still remain within the spirit and scope of the embodiments of this application.

[0119] It will be understood by those skilled in the art that all or some of the steps, systems, or apparatuses disclosed above, and their functional modules / units, can be implemented as software, firmware, hardware, or suitable combinations thereof. In hardware implementations, the division between functional modules / units mentioned above does not necessarily correspond to the division of physical components; for example, a physical component may have multiple functions, or a function or step may be performed collaboratively by several physical components. Some or all components may be implemented as software executed by a processor, such as a digital signal processor or microprocessor, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit (ASIC). Such software may be distributed on a computer-readable medium, which may include computer storage media (or non-transitory media) and communication media (or transient media). As is known to those skilled in the art, the term computer storage media includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disc (DVD) or other optical disc storage, magnetic cartridges, magnetic tape, disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and can be accessed by a computer. Furthermore, it is well known to those skilled in the art that communication media typically contain computer-readable instructions, data structures, program modules, or other data in modulated data signals such as carrier waves or other transmission mechanisms, and may include any information delivery medium.

Claims

1. An automatic parking method, characterized in that, include: When the parking area of ​​the target vehicle is determined to be a non-standard parking space, a three-dimensional point cloud map of the surrounding environment of the target vehicle and surrounding environment image information are obtained. The three-dimensional point cloud map is projected into a two-dimensional form. The quadrilateral shape fitting of the two-dimensional projection is performed using a preset maximum hull model, including: drawing the maximum hull polygonal geometry of the two-dimensional projection using the preset maximum hull model; and approximating the maximum hull polygonal geometry as a quadrilateral shape. The process involves matching the preset geometric model of the target vehicle with the fitted quadrilateral shape, including: When the area of ​​the fitted quadrilateral is greater than the area of ​​the geometric model of the target vehicle and the side length of the shortest side of the fitted quadrilateral is greater than the width of the target vehicle, the search nodes are expanded outward in the form of a random search tree with the geometric center of the quadrilateral as the origin and the sides of the quadrilateral as the boundaries. For each searched node, perform the following operations until the searched node meets the preset conditions: For each searched node, substitute the coordinates of the geometric center point of the target vehicle's geometric model into the node coordinates. When the geometric center of the target vehicle's geometric model is located at the current node, determine whether the quadrilateral shape intersects with the target vehicle's geometric model. If there is an intersection, expand the search for the next node outwards using a random search tree. The preset conditions include that the quadrilateral shape has no intersection with the target vehicle's geometric model or that all nodes within the boundary have been traversed. When it is determined that the quadrilateral shape has no intersection with the geometric model of the target vehicle, then the preset geometric model of the target vehicle and the fitted quadrilateral shape are matched. Once a match is successful, the parking space type of the waiting area is estimated based on the two-dimensional projection and the surrounding environment image information. Based on the parking space type, a path is planned for the target vehicle, and the target vehicle is parked in the waiting area according to the planned path.

2. The automatic parking method according to claim 1, characterized in that, The non-standard parking spaces are determined as follows: If a parking area is found to have no parking space lines or partially covered by parking space lines, the parking area is determined to be a non-standard parking area.

3. The automatic parking method according to claim 1, characterized in that, Once a match is successful, the parking space type is estimated for the waiting area based on the two-dimensional projection and surrounding environment image information. This estimation includes: The location information of surrounding vehicles of the target vehicle is determined based on the surrounding environment image information; The pose of the target vehicle and its current state in the environment are determined based on the location information of the surrounding vehicles. The state of the surrounding vehicles is estimated based on the pose of the target vehicle, the state of the current environment, the position information of the surrounding vehicles, and the two-dimensional projection. The parking space type of the waiting area is estimated based on the state of the surrounding vehicles.

4. The automatic parking method according to claim 1, characterized in that, The step of planning the route for the target vehicle based on the parking space type includes: The target vehicle is planned using the hybrid A* algorithm.

5. The automatic parking method according to claim 4, characterized in that, Also includes: Gravity constraints are used in the path planning process.

6. The automatic parking method according to claim 1, characterized in that, Before parking the target vehicle into the waiting area according to the planned route, the process includes: The vehicle's infotainment system simulates the target vehicle parking in the designated parking area.

7. An automatic parking device, comprising a memory and a processor, characterized in that, include: The memory is used to store the program for automatic parking; The processor is configured to read and execute the program for automatic parking, and to execute the method as described in any one of claims 1-6.

Images