Spatial parking space determination method and device

Abstract

The invention relates to a spatial parking space determination method and device, the method is applied to a vehicle, and the method comprises the steps that a grounding point set corresponding to obstacles around a vehicle body of the vehicle is acquired, and the grounding point set comprises a plurality of grounding points, making contact with the ground, of the obstacles around the vehicle body; mapping the grounding point set to a preset plane where the vehicle is located to obtain an obstacle point map corresponding to the vehicle; obstacle points in the obstacle point map are converted into an obstacle grid map of a preset size corresponding to the vehicle, and a target obstacle grid map is obtained; and matching a space parking space corresponding to the vehicle from the target obstacle grid map. Therefore, the spatial parking space of the vehicle can be identified efficiently and accurately.

Description

Technical Field

[0001] The present application relates to the field of vehicle technology, and in particular to a method and device for determining a parking space. Background Art

[0002] Spatial parking spaces refer to non-marked parking spaces, which usually do not have fixed sizes and layouts like traditional marked parking spaces to make it easy for drivers to park. Therefore, in the vehicle's automatic parking system, the detection of spatial parking spaces is of great significance to the automatic parking system, which can provide users with spatial parking spaces that are easy to park in garages or parking lots without clear markings.

[0003] At present, in the automatic parking system of a vehicle, the space available for parking of the vehicle is generally determined by direct distance measurement. The distance to surrounding obstacles is measured by sensors, and the continuous measurement data is used to determine whether there is parking space.

[0004] However, the above method is relatively simple, mainly based on the distance measurement results of obstacles and the pre-set rule algorithm to calculate whether the vehicle can be parked in the area. It is highly dependent on the detection results of the sensor. When the detection distance is inaccurate, it is easy to cause false detection or missed detection, resulting in inaccurate detection results of the space parking space, reducing the user experience. Summary of the invention

[0005] The present application provides a method and device for determining a spatial parking space to solve the technical problem in the prior art that, when determining the spatial parking space of a vehicle, the spatial parking space is strongly dependent on the detection results of the sensor. When the detection result distance is inaccurate, it is easy to cause false detection or missed detection, thereby causing inaccurate detection results of the spatial parking space and reducing the user experience.

[0006] In a first aspect, the present application provides a method for determining a parking space, which is applied to a vehicle, and the method comprises:

[0007] Acquire a grounding point set corresponding to obstacles around a vehicle body, the grounding point set including a plurality of grounding points where the obstacles around the vehicle body contact the ground;

[0008] Mapping the ground point set to a preset plane where the vehicle is located to obtain an obstacle point map corresponding to the vehicle;

[0009] Converting obstacle points in the obstacle point map into an obstacle grid map of a preset size corresponding to the vehicle to obtain a target obstacle grid map;

[0010] A spatial parking space corresponding to the vehicle is matched from the target obstacle grid map.

[0011] As an optional implementation, a plurality of image acquisition modules and a plurality of distance detection modules are pre-installed on the vehicle body, and the step of obtaining a set of grounding points corresponding to obstacles around the vehicle body includes:

[0012] By means of the image acquisition module, a first grounding point set of obstacles around the vehicle body is acquired;

[0013] By means of the distance detection module, a second grounding point set of obstacles around the vehicle body is collected;

[0014] The first grounding point set and the second grounding point set are determined as grounding point sets corresponding to obstacles around a body of the vehicle.

[0015] As an optional implementation, the collecting, by the image collection module, a first grounding point set of obstacles around the vehicle body includes:

[0016] Obstacle images of obstacles around the vehicle body are collected through the image acquisition module; the obstacle images are input into a pre-trained semantic segmentation model to obtain grounding point lines between the obstacles around the vehicle body and the ground output by the semantic segmentation model; and the first grounding point set is extracted from the grounding point lines.

[0017] or,

[0018] Inputting the obstacle image into a pre-trained key point detection model to obtain a first grounding point set of obstacles around the vehicle body output by the key point detection model;

[0019] The collecting of a second grounding point set of obstacles around the vehicle body by the distance detection module includes:

[0020] Obtaining the obstacle coordinates of the obstacles around the vehicle body relative to the vehicle body through the distance detection module;

[0021] The grounding point represented by the obstacle coordinates is included in the second grounding point set.

[0022] As an optional implementation, mapping the ground point set to a preset plane where the vehicle is located to obtain an obstacle point map corresponding to the vehicle includes:

[0023] Using the inverse perspective transformation principle, projecting the first grounding point in the first grounding point set onto the preset plane to obtain a first obstacle point map including a plurality of first obstacle points;

[0024] Determine, from the first obstacle point map, an obstacle area including a plurality of the first obstacle points, wherein a size of the obstacle area is smaller than a preset size;

[0025] For each of the obstacle areas, determining the distance between each of the first obstacle points in the obstacle area and the vehicle body;

[0026] retaining the first obstacle point with the smallest distance among the multiple first obstacle points included in the obstacle area, to obtain a second obstacle point map;

[0027] By using the inverse perspective transformation principle, the second ground point in the second ground point set is projected into the second obstacle point map to obtain the obstacle point map corresponding to the vehicle.

[0028] As an optional implementation, mapping the ground point set to a preset plane where the vehicle is located to obtain an obstacle point map corresponding to the vehicle includes:

[0029] Mapping the ground point set to a preset plane where the vehicle is located to obtain an initial obstacle point map corresponding to the vehicle;

[0030] Establishing a global coordinate system based on the current position of the vehicle, and acquiring a plurality of running grounding point sets collected at different positions during the running of the vehicle, wherein the running grounding point sets include a plurality of running grounding points, and the running grounding points are represented by running grounding point coordinates relative to the vehicle body;

[0031] Converting the operating grounding point coordinates of the plurality of operating grounding points into target grounding point coordinates in a global coordinate system;

[0032] According to the target touchdown point coordinates, the operating touchdown points included in the operating touchdown point set are converted to the initial obstacle point map by using the inverse perspective transformation principle to obtain the obstacle point map corresponding to the vehicle.

[0033] As an optional implementation, the obstacle grid map is centered on the vehicle, the grids in the obstacle grid map include obstacle attributes and non-obstacle attributes, and the obstacle points in the obstacle point map are converted to the obstacle grid map of a preset size corresponding to the vehicle to obtain a target obstacle grid map, including:

[0034] Determining obstacle coordinates of the obstacle points in the obstacle point map relative to the vehicle;

[0035] For each obstacle point in the obstacle point map, determining a target grid of the obstacle point in the obstacle grid map according to obstacle coordinates corresponding to the obstacle point;

[0036] The attributes of the target grid corresponding to each obstacle point are updated to obstacle attributes to obtain the target obstacle grid map.

[0037] As an optional implementation, matching the spatial parking space corresponding to the vehicle from the target obstacle grid map includes:

[0038] Determining a spatial area for matching the spatial parking space from the target obstacle grid map;

[0039] Generating an outer bounding box of the vehicle according to the size of the vehicle;

[0040] According to the outer boundary box, a plurality of the spatial parking spaces are matched from the spatial area.

[0041] As an optional implementation manner, determining the spatial area for matching the spatial parking space from the target obstacle grid map includes:

[0042] Taking the vehicle as the center, dividing the target obstacle grid map into at least two obstacle grid sub-areas;

[0043] For each obstacle grid sub-region, the attributes of the grids between the target grid in the obstacle grid sub-region and the preset boundary corresponding to the target grid are updated to obstacle attributes to obtain an updated obstacle grid map;

[0044] Determine a non-obstacle grid having a non-obstacle attribute from the updated obstacle grid map;

[0045] Extracting the boundary of the non-obstruction grid, and dividing the boundary of the non-obstruction grid into a real boundary and a virtual boundary; wherein the real boundary is the boundary between the non-obstruction grid and the target grid, and the virtual boundary is the boundary of the updated obstacle grid map;

[0046] An area consisting of the real boundary and the virtual boundary in the updated obstacle grid map is determined as a spatial area for matching the spatial parking space.

[0047] As an optional implementation, matching a plurality of the spatial parking spaces from the spatial region according to the outer boundary box includes:

[0048] From the spatial region, matching a target region whose size is greater than or equal to the outer bounding box;

[0049] Using the target area as a parking space corresponding to the vehicle;

[0050] Taking the boundary of the target area as the real boundary and fusing it with the real boundary of the space area to obtain a new space area;

[0051] The new spatial area is used as the spatial area, and the steps of matching a target area whose size is greater than or equal to the outer boundary box from the spatial area and using the target area as the spatial parking space corresponding to the vehicle are re-executed until the target area cannot be matched, thereby obtaining multiple spatial parking spaces matched from the spatial area.

[0052] As an optional implementation manner, matching, from the spatial region, a target region whose size is greater than or equal to the outer bounding box, includes:

[0053] Determine, from the real boundaries included in the spatial region, a target real boundary for matching with the outer boundary box, and determine an inner point of the target real boundary; wherein the inner point is a point in the target real boundary that is closest to the vehicle;

[0054] Based on the inner point, aligning the outer bounding box with the real boundary of the object;

[0055] Perform interference detection on the outer boundary box and the virtual boundary in the space area to obtain a detection result;

[0056] When the detection result indicates that there is no interference relationship between the outer boundary box and any of the virtual boundaries in the spatial region, the region currently corresponding to the outer boundary box is determined as the target region.

[0057] As an optional implementation manner, determining, from the real boundaries included in the spatial region, the target real boundary for matching with the outer boundary box includes:

[0058] In the case where there is a first target real boundary whose length is greater than the length of the outer boundary box in the real boundaries included in the determined spatial region, determining the first target real boundary as the target real boundary;

[0059] In determining the real boundaries included in the spatial region, if there is no first target real boundary whose length is greater than the length of the outer boundary box, and there is a second target real boundary whose length is greater than the width of the outer boundary box, determining the second target real boundary as the target real boundary;

[0060] The step of aligning the outer boundary box with the real boundary of the target based on the inner point comprises:

[0061] In a case where the target real boundary is the first target real boundary, aligning the long side of the outer boundary box with the target real boundary;

[0062] In a case where the target real boundary is the second target real boundary, the broad side of the outer boundary box is aligned with the target real boundary.

[0063] In a second aspect, the present application provides a device for determining a parking space, which is applied to a vehicle, and the device includes:

[0064] An acquisition module, used to acquire a set of grounding points corresponding to obstacles around a vehicle body, wherein the set of grounding points includes a plurality of grounding points where the obstacles around the vehicle body contact the ground;

[0065] A mapping module, used to map the set of grounding points onto a preset plane where the vehicle is located, to obtain an obstacle point map corresponding to the vehicle;

[0066] A conversion module, used to convert the obstacle points in the obstacle point map into an obstacle grid map of a preset size corresponding to the vehicle to obtain a target obstacle grid map;

[0067] A matching module is used to match the spatial parking space corresponding to the vehicle from the target obstacle grid map.

[0068] In a third aspect, the present application provides a vehicle, comprising a processor and a memory, wherein the processor is used to execute a spatial parking space determination program stored in the memory to implement the spatial parking space determination method described in any one of the first aspects.

[0069] In a fourth aspect, the present application provides a storage medium storing one or more programs, and the one or more programs can be executed by one or more processors to implement the method for determining a spatial parking space described in any one of the first aspects.

[0070] The technical solution provided by the embodiment of the present application is to obtain a set of grounding points corresponding to obstacles around the vehicle body, the set of grounding points including multiple grounding points where the obstacles around the vehicle body contact the ground, map the set of grounding points to the preset plane where the vehicle is located, obtain an obstacle point map corresponding to the vehicle, convert the obstacle points in the obstacle point map to an obstacle grid map of a preset size corresponding to the vehicle, obtain a target obstacle grid map, and match the spatial parking space corresponding to the vehicle from the target obstacle grid map. This technical solution obtains an obstacle point map around the vehicle by collecting a set of grounding points between obstacles around the vehicle body and the ground, and maps the set of grounding points, thereby converting the obstacle point map to a target obstacle grid map of the vehicle. By constructing an obstacle grid map corresponding to obstacles around the vehicle, it can comprehensively and accurately identify obstacles around the vehicle and the size of the free area that can be used for parking. Afterwards, the spatial parking space of the vehicle can be more accurately matched from the target obstacle grid map of the vehicle, thereby realizing efficient and accurate identification of the spatial parking space of the vehicle. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0072] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the prior art texts are briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0073] One or more embodiments are exemplarily described by pictures in the corresponding drawings, and these exemplified descriptions do not constitute limitations on the embodiments. Elements with the same reference numerals in the drawings represent similar elements, and unless otherwise stated, the figures in the drawings do not constitute proportional limitations.

[0074] Figure 1 This is an example diagram of a method for determining a parking space in the prior art provided by the present application;

[0075] Figure 2 A flow chart of an embodiment of a method for determining a parking space provided in an embodiment of the present application;

[0076] Figure 3 A flowchart of another method for determining a parking space provided in an embodiment of the present application;

[0077] Figure 4 A schematic diagram of the structure of an obstacle point map provided in an embodiment of the present application;

[0078] Figure 5 A schematic diagram of the structure of an obstacle grid map provided in an embodiment of the present application;

[0079] Figure 6 A flowchart of another method for determining a parking space provided in an embodiment of the present application;

[0080] Figure 7 A schematic diagram of a grounding point line provided in an embodiment of the present application;

[0081] Figure 8 A schematic diagram of a grounding point set provided in an embodiment of the present application;

[0082] Fig. 9 A schematic diagram of the installation of an image acquisition module and a distance detection module provided in an embodiment of the present application;

[0083] Fig.10 A is a schematic diagram of a first obstacle point before fusion provided in an embodiment of the present application;

[0084] Fig.10 B is a schematic diagram of a first obstacle point after fusion provided in an embodiment of the present application;

[0085] Fig.11 A is a schematic diagram of a target obstacle grid map provided in an embodiment of the present application;

[0086] Fig.11 B is a schematic diagram of updating an obstacle grid map provided in an embodiment of the present application;

[0087] Fig.12 A is a schematic diagram of the boundary of a non-obstruction grid provided in an embodiment of the present application;

[0088] Fig.12 B is a schematic diagram of a space region provided in an embodiment of the present application;

[0089] Fig.13 A schematic diagram of a matching space parking space provided in an embodiment of the present application;

[0090] Fig.14 A is a schematic diagram of determining multiple parking spaces provided in an embodiment of the present application;

[0091] Fig.14 B is a schematic diagram of a new space area provided in an embodiment of the present application;

[0092] Fig.15 A block diagram of an embodiment of a device for determining a parking space provided in an embodiment of the present application;

[0093] Fig.16A schematic diagram of the structure of a vehicle provided in an embodiment of the present application. DETAILED DESCRIPTION

[0094] In order to make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be clearly and completely described in conjunction with the drawings in the embodiments of the present application. Obviously, the embodiments described in the text are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0095] The disclosure below provides many different embodiments or examples to implement different structures of the present invention. In order to simplify the disclosure of the present invention, the parts and settings of specific examples are described below. Of course, they are only examples, and the purpose is not to limit the present invention. In addition, the present invention can repeat reference numbers and / or letters in different examples. This repetition is for the purpose of simplification and clarity, and does not itself indicate the relationship between the various embodiments and / or settings discussed.

[0096] In order to facilitate understanding of the method for determining a space parking space provided by the present application, the method for determining a space parking space in the prior art is first described below:

[0097] In the related art, the automatic parking system may include the following methods when identifying the parking spaces around the vehicle:

[0098] First, direct distance measurement method, this method directly measures the distance between the vehicle and the surrounding obstacles through the distance sensor, and then uses the continuously measured data to determine whether there is a parking space. For example, the parking distance between two vehicles is determined through the continuously measured data, and whether there is a parking space between the two vehicles is determined based on the parking distance.

[0099] Among them, this method is relatively simple, mainly based on the measured obstacle distance results and the pre-set rule algorithm to calculate whether the location can park the vehicle. However, this method is highly dependent on the detection results of the sensor. If the detection distance is inaccurate, it is easy to cause false detection or missed detection. At the same time, this method can only detect the part very close to the vehicle, and the detection range is small.

[0100] Second, the probabilistic map method uses map representation and data statistics to divide the environment into a series of small grids, each grid has an occupancy probability, and estimates the parking spaces around the vehicle by the grid occupancy degree, for example, see Figure 1 , is an example diagram of a method for determining a parking space in the prior art provided by the present application, such as Figure 1As shown, in this method, the surrounding environment of the vehicle is divided into a series of grids, and then the spatial parking space of the vehicle is determined according to the occupancy degree of the grids.

[0101] Among them, this method can detect space within a larger range to estimate the spatial parking space in the current scene. However, this method strongly relies on the accuracy of the occupancy map, and it cannot calculate the accurate spatial parking space in the 2D occupancy map.

[0102] In order to solve the technical problem that in the prior art, when determining the spatial parking space of a vehicle, the spatial parking space is strongly dependent on the detection results of the sensor, and when the detection result distance is inaccurate, it is easy to cause false detection or missed detection, thereby causing inaccurate detection results of the spatial parking space and reducing user experience, the present application provides a method and device for determining a spatial parking space, which can obtain an obstacle point map around the vehicle by collecting a set of grounding points of obstacles around the vehicle body and the ground, and mapping the set of grounding points, thereby converting the obstacle point map into a target obstacle grid map of the vehicle, and by constructing an obstacle grid map corresponding to the obstacles around the vehicle, it can comprehensively and accurately identify the obstacles around the vehicle and the size of the free area that can be used for parking. Afterwards, the spatial parking space of the vehicle can be more accurately matched from the target obstacle grid map of the vehicle, thereby realizing efficient and accurate identification of the spatial parking space of the vehicle.

[0103] The method for determining a parking space provided by the present application is further explained below with reference to specific embodiments in conjunction with the accompanying drawings. The embodiments do not constitute a limitation on the embodiments of the present invention.

[0104] See also Figure 2 , is a flow chart of an embodiment of a method for determining a parking space provided in an embodiment of the present application. As an embodiment, Figure 2 The process shown can be applied to vehicles. Figure 2 As shown, the process may include the following steps:

[0105] Step 201: Acquire a set of grounding points corresponding to obstacles around a vehicle body, wherein the set of grounding points includes a plurality of grounding points where the obstacles around the vehicle body contact the ground.

[0106] Step 202: Map the above-mentioned ground point set to the preset plane where the vehicle is located to obtain an obstacle point map corresponding to the vehicle.

[0107] The following is a unified description of step 201 and step 202:

[0108] The above-mentioned obstacles around the vehicle body refer to obstacles around the location of the vehicle, which may include objects around the vehicle body that hinder the operation of the vehicle, such as other vehicles, trees, pedestrians, roadblocks, curbs, road signs, etc.

[0109] The above-mentioned grounding point set refers to a set composed of multiple grounding points, and the above-mentioned grounding points are the points where the obstacles around the above-mentioned vehicle body contact the ground. They can be all the points where the obstacles around the vehicle body contact the ground, or they can be some of the points where the obstacles around the vehicle body contact the ground. The embodiments of the present application do not limit this.

[0110] The above-mentioned preset plane refers to the plane corresponding to the road surface where the vehicle is located, which may be a BEV (Bird's Eye View) plane of the road surface.

[0111] In an embodiment of the present application, in order to accurately obtain the free area around the vehicle that can be used as a parking space, the execution entity of the embodiment of the present application can collect a set of ground points corresponding to obstacles around the vehicle body, and map the set of ground points to a preset plane where the vehicle is located (for example, a bird's-eye view plane of the road where the vehicle is located) to obtain an obstacle point map corresponding to the vehicle. The obstacle point map can accurately reflect the relative relationship between the obstacles around the vehicle body and the free area around the vehicle body.

[0112] As a possible implementation method, an object detection module may be pre-installed around the vehicle body, and obstacles around the vehicle body may be detected by the object detection module, so as to obtain a set of grounding points where the obstacles around the vehicle body contact the ground based on the detection results.

[0113] As an exemplary implementation, in order to improve the accuracy of the grounding point in the grounding point set, the execution subject of the embodiment of the present application can detect obstacles around the vehicle body through different object detection modules, and determine different grounding points according to the detection results, so as to fuse the different grounding points to obtain the above grounding point set. Among them, the above different object detection modules may include an image acquisition module and a distance detection module, the above image acquisition module may be a surround view camera, a camera, etc., and the above distance detection module may be a radar sensor, such as an ultrasonic radar.

[0114] As for how to obtain the ground point set corresponding to the obstacles around the vehicle body through different object detection modules, it can be seen in the following Figure 3 The process shown is explained and will not be described in detail here.

[0115] As an optional implementation, when mapping the grounding point set to the preset plane where the vehicle is located, the grounding point set may be projected onto the preset plane where the vehicle is located by using the inverse perspective transformation principle.

[0116] As an optional implementation, when obtaining an obstacle point map by mapping a ground point set to a preset plane, since the ground points collected by the object detection module each time when obtaining the ground point set of obstacles around the vehicle body are not dense, in order to improve the coverage of the obstacle point map, the execution subject of the embodiment of the present application can collect the ground point set of obstacles around the vehicle body multiple times. Afterwards, the ground point sets collected multiple times can be mapped to the preset plane where the vehicle is located to obtain the above obstacle point map.

[0117] As an exemplary implementation, the above-mentioned ground point set may be mapped onto a preset plane where the vehicle is located to obtain an initial obstacle point map corresponding to the vehicle.

[0118] Afterwards, a global coordinate system can be established at the current position of the vehicle, and multiple operating grounding point sets collected at different positions during the operation of the vehicle can be obtained. The above-mentioned operating grounding point set may include multiple operating grounding points, and the above-mentioned operating grounding points are represented by the operating grounding point coordinates relative to the vehicle body, that is, the coordinate system corresponding to the operating grounding point is the same as the coordinate system corresponding to the grounding point included in the above-mentioned grounding point set, and both are coordinate systems established with the preset point of the vehicle body as the origin.

[0119] Among them, since the vehicle body is movable, the coordinates of the above-mentioned running grounding point relative to the vehicle body and the coordinates of the grounding point relative to the vehicle body may be the same, but in reality they are not the same obstacles around the vehicle body. Therefore, when the vehicle starts to collect the grounding point set, a global coordinate system is established to distinguish different obstacles around the vehicle body during the operation of the vehicle.

[0120] Based on this, the operating grounding point coordinates of the above-mentioned multiple operating grounding points can be converted into the target grounding point coordinates in the global coordinate system, and according to the target grounding point coordinates, the operating grounding points included in the above-mentioned operating grounding point set can be converted to the above-mentioned initial obstacle point map using the inverse perspective transformation principle to obtain the obstacle point map corresponding to the vehicle. For example, the following can be obtained: Figure 4 Obstacle point map shown, see Figure 4 , is a schematic diagram of the structure of an obstacle point map provided in an embodiment of the present application. Figure 4 As shown, vehicle 41 is the vehicle for which the parking space is to be determined, that is, the executor of the embodiment of the present application, and vehicles 42, 43, and a vertical line 44 representing a curb are obstacles around the vehicle bodies. The set of grounding points corresponding to the above-mentioned obstacles around the vehicle bodies is represented by an "X" symbol 45.

[0121] Step 203: convert the obstacle points in the obstacle point map into an obstacle grid map of a preset size corresponding to the vehicle to obtain a target obstacle grid map.

[0122] Step 204: Match the spatial parking space corresponding to the vehicle from the target obstacle grid map.

[0123] The following is a unified description of step 203 and step 204:

[0124] The above obstacle grid map refers to an obstacle grid map of a preset size set in advance for the vehicle. The preset size of the obstacle grid map can be determined according to the size of the vehicle, for example, see Figure 5 , is a schematic diagram of the structure of an obstacle grid map provided in an embodiment of the present application. Figure 5 The obstacle grid map may be a grid map centered on the vehicle, with a side length of 8 meters and a size of 0.3 meters*0.3 meters for each grid.

[0125] Among them, Figure 5 As shown, the above obstacle grid map may include obstacle attributes and non-obstacle attributes. The grid corresponding to the obstacle attribute indicates that there are obstacles in the area corresponding to the grid (for example, Figure 5 The grid corresponding to the non-obstacle attribute indicates that the area corresponding to the grid is an idle area, where there are no obstacles (e.g. Figure 5 blank grid shown).

[0126] In an embodiment of the present application, after determining the obstacle point map currently corresponding to the vehicle, the obstacle points in the above obstacle point map can be converted into an obstacle grid map of a preset size corresponding to the vehicle, thereby obtaining a target obstacle grid map that represents the correspondence between the obstacles around the vehicle and the vehicle body.

[0127] As an optional implementation method, the grid in the above-mentioned obstacle grid map may include obstacle attributes and non-obstacle attributes. The above-mentioned obstacle point map and obstacle grid map both have a corresponding relationship with the vehicle. Based on the corresponding relationship, the obstacle points used to represent the grounding points in the obstacle point map can be mapped to the above-mentioned obstacle grid map, and the attributes of the grid of the mapped obstacle points can be updated to the obstacle attributes, thereby obtaining the target obstacle grid map.

[0128] Afterwards, the spatial parking space corresponding to the vehicle can be matched from the above-mentioned target obstacle grid map according to the preset matching rules.

[0129] As for how to convert the obstacle points in the obstacle point map to the target obstacle grid map obtained in the obstacle grid map corresponding to the vehicle, and how to match the space parking space corresponding to the vehicle from the target obstacle grid map, it can be seen below. Figure 6 The process shown is explained and will not be described in detail here.

[0130] The technical solution provided by the embodiment of the present application is to obtain a set of grounding points corresponding to obstacles around the vehicle body, the set of grounding points including multiple grounding points where the obstacles around the vehicle body contact the ground, map the set of grounding points to the preset plane where the vehicle is located, obtain an obstacle point map corresponding to the vehicle, convert the obstacle points in the obstacle point map to an obstacle grid map of a preset size corresponding to the vehicle, obtain a target obstacle grid map, and match the spatial parking space corresponding to the vehicle from the target obstacle grid map. This technical solution obtains an obstacle point map around the vehicle by collecting a set of grounding points between obstacles around the vehicle body and the ground, and maps the set of grounding points, thereby converting the obstacle point map to a target obstacle grid map of the vehicle. By constructing an obstacle grid map corresponding to obstacles around the vehicle, it can comprehensively and accurately identify obstacles around the vehicle and the size of the free area that can be used for parking. Afterwards, the spatial parking space of the vehicle can be more accurately matched from the target obstacle grid map of the vehicle, thereby realizing efficient and accurate identification of the spatial parking space of the vehicle.

[0131] See also Figure 3 , which is a flow chart of an embodiment of another method for determining a spatial parking space provided in an embodiment of the present application. Figure 3 The process shown in Figure 2 Based on the process shown in the figure, it describes how to obtain the ground point set corresponding to the obstacles around the vehicle body based on the pre-installation of multiple image acquisition modules and multiple distance detection modules on the vehicle body. Figure 3 As shown, the process may include the following steps:

[0132] Step 301: collect a first grounding point set of obstacles around the vehicle body through an image acquisition module.

[0133] The above-mentioned image acquisition module may be a device for acquiring images, which may be a camera, etc. Among them, in order to more accurately obtain the grounding point set of obstacles around the vehicle body, the above-mentioned image acquisition module may be a surround-view camera.

[0134] In an embodiment of the present application, multiple image acquisition modules can be pre-installed around the vehicle body (for example, a surround-view camera is installed on the front, rear, left and right of the vehicle body), and the image acquisition modules are used to determine the grounding point set of obstacles around the vehicle body (for ease of distinction, hereinafter referred to as the first grounding point set).

[0135] As an optional implementation method, the execution subject of the embodiment of the present application can collect obstacle images of obstacles around the vehicle through an image acquisition module, and input the obstacle images into a pre-trained semantic segmentation model to obtain the grounding point lines between the obstacles around the vehicle body and the ground output by the semantic segmentation model. Figure 7, is a schematic diagram of a grounding point line provided in an embodiment of the present application. Figure 7 As shown, the grounding point line 71 is the grounding point line where obstacles around the vehicle body (such as vehicles and curbs) touch the ground.

[0136] As an exemplary implementation, when training the semantic segmentation model, a training sample input by a technician may be obtained, and the true value label in the training sample may be obtained by manual annotation, wherein the grounded part may be annotated as a line, and then the annotated obstacle grounding line is expanded and used as the true value label. Afterwards, the obtained training sample may be used to train the preset initial model to obtain the semantic segmentation model.

[0137] Afterwards, the first grounding point set may be extracted from the grounding point lines. The grounding point lines refer to lines corresponding to the contact parts of obstacles around the vehicle body and the ground.

[0138] As an exemplary implementation, when extracting the first grounding point set from the grounding point line, two endpoints of the grounding point line may be extracted, and the grounding points included in the grounding point line may be extracted at preset intervals to obtain the first grounding point set.

[0139] As another optional implementation, an obstacle image of obstacles around the vehicle body can be collected by an image acquisition module, and the obstacle image can be input into a pre-trained key point detection model to obtain a first ground point set output by the key point detection model. Figure 8 , is a schematic diagram of a grounding point set provided in an embodiment of the present application. Figure 8 As shown, the grounding point 81 is the grounding point where the obstacles (vehicles, curbs, etc.) around the vehicle body are collected and contact the ground.

[0140] As an exemplary implementation, the execution subject of the embodiment of the present application may obtain a ground point data set of a surround view image collected in advance by a user, and use the ground point data set of the surround view image to train a preset initial model to obtain a key point detection model.

[0141] Step 302: Collect a second grounding point set of obstacles around the vehicle sound through the distance detection module.

[0142] The above-mentioned distance detection module refers to a detection device used to detect the distance of obstacles around the vehicle body, which can be a distance sensor, such as an infrared sensor, an ultrasonic radar, etc.

[0143] In an embodiment of the present application, multiple distance detection modules can be pre-installed around the vehicle body (for example, multiple ultrasonic radars can be installed around the vehicle body), and the distance detection modules can be used to determine a set of grounding points of obstacles around the vehicle body (hereinafter referred to as the second grounding point set for ease of distinction).

[0144] As an optional implementation method, the obstacle coordinates of obstacles around the vehicle body relative to the vehicle body (for example, coordinates relative to a preset position of the vehicle body (door, roof or taillight, etc.)) can be collected by the above-mentioned multiple distance detection modules. Afterwards, the grounding points represented by the above-mentioned obstacle coordinates can be included in the second grounding point set.

[0145] Step 303: determine the first grounding point set and the second grounding point set as grounding point sets corresponding to obstacles around the vehicle body.

[0146] In the embodiment of the present application, in order to improve the accuracy of the grounding point set, multiple image acquisition modules and distance detection modules can be pre-installed on the vehicle body, for example, see Fig. 9 , is a schematic diagram of the installation of an image acquisition module and a distance detection module provided in an embodiment of the present application. Fig. 9 As shown, multiple image acquisition modules 91 (such as surround-view cameras) and multiple distance detection modules 92 (such as ultrasonic radars) can be installed around the body of the vehicle, wherein images 93 are all images captured by the image acquisition modules in the image acquisition area 94, and area 95 is the detection range of the distance detection module.

[0147] Based on this, the execution subject of the embodiment of the present application can respectively collect a first grounding point set of obstacles around the vehicle body through multiple image acquisition modules, and collect a second grounding point set of obstacles around the vehicle body through multiple distance detection modules, and use the first grounding point set and the second grounding point set as grounding point sets of obstacles around the vehicle body. By fusing the grounding point sets collected by the image acquisition module and the distance detection module, a more accurate grounding point set can be obtained.

[0148] In addition, based on the above-mentioned first ground point set and second ground point set, the executing entity of the embodiment of the present application can respectively map the above-mentioned first ground point set and second ground point set to the preset plane where the vehicle is located to obtain an obstacle point map corresponding to the vehicle.

[0149] As an implementation method, the principle of inverse perspective transformation can be used to project the first grounding point in the above-mentioned first grounding point set onto the above-mentioned preset plane to obtain a first obstacle point map including multiple first obstacle points.

[0150] Afterwards, the first obstacle points that are closer in the first obstacle point map can be merged. As an exemplary implementation, an obstacle area containing multiple first obstacle points can be determined from the above-mentioned first obstacle point map, and the size of the above-mentioned obstacle area is smaller than the preset size. Afterwards, for each obstacle area, the distance between each first obstacle point in the obstacle area and the vehicle body can be determined, and the first obstacle point with the smallest distance among the multiple first obstacle points included in the obstacle area is retained to obtain a second obstacle point map. The above-mentioned preset size refers to a pre-set smaller range of sizes, such as 10cm*10cm.

[0151] For example, see Fig.10 , is a schematic diagram of a first obstacle point fusion provided in an embodiment of the present application. Fig.10 A is a schematic diagram of a first obstacle point before fusion provided in an embodiment of the present application. Fig.10 The vehicle 1000 in A may include an image acquisition module 1001 and a distance detection module 1002, and the first obstacle point 1003 and the first obstacle point 1004 are the grounding points of the vehicle 1005, the vehicle 1006, or the curb 1007 captured by the image acquisition module 1001 and the distance detection module 1002, respectively. Fig.10 By integrating the first obstacle point in A, we can get Fig.10 The first obstacle point shown in B, Fig.10 B is a schematic diagram of a first obstacle point fusion provided in an embodiment of the present application. Fig.10 As shown in B, Fig.10 Based on A, the two first obstacle points 1004 close to the two first obstacle points 1003 are deleted to obtain the final second obstacle point map.

[0152] Finally, the inverse perspective transformation principle can be used to project the grounding points in the second grounding point set (hereinafter referred to as second grounding points for ease of description) into the second obstacle point map to obtain the obstacle point map corresponding to the vehicle.

[0153] The technical solution provided in the embodiment of the present application collects a first grounding point set of obstacles around the vehicle body through an image acquisition module, collects a second grounding point set of obstacles around the vehicle body through a distance detection module, and determines the first grounding point set and the second grounding point set as the grounding point set corresponding to the obstacles around the vehicle body. This technical solution collects the grounding point set of obstacles around the vehicle body using an image acquisition module and a distance detection module respectively, and merges the grounding point sets collected respectively, thereby determining the grounding point set of obstacles around the vehicle body through a multi-dimensional detection scheme, thereby achieving a more accurate determination of the grounding point set of obstacles around the vehicle body.

[0154] See also Figure 6, which is a flow chart of an embodiment of another method for determining a spatial parking space provided in an embodiment of the present application. Figure 6 The process shown in Figure 2 Based on the process shown in FIG. 1 , it describes how to convert the obstacle points in the obstacle point map into the obstacle grid map of the preset size corresponding to the vehicle. Figure 6 As shown, the process may include the following steps:

[0155] Step 601: Determine the obstacle coordinates of the obstacle point in the obstacle point map relative to the vehicle;

[0156] Step 602: for each obstacle point in the obstacle point map, determine the target grid of the obstacle point in the obstacle grid map according to the obstacle coordinates corresponding to the obstacle point.

[0157] Step 603: Update the attributes of the target grid corresponding to each obstacle point to obstacle attributes, and obtain a target obstacle grid map.

[0158] The following is a unified description of steps 601 to 603:

[0159] The above-mentioned obstacle coordinates refer to the coordinates of the obstacle points in the obstacle point map. Since the obstacle points in the obstacle point map are the points where the obstacles around the vehicle body contact the ground, the coordinates of the obstacle points can be the coordinates in a coordinate system established with the vehicle (for example, the preset position of the vehicle) as the center.

[0160] The obstacle grid map is a grid map of a preset size centered on the vehicle, which is pre-set for the vehicle. For example Figure 5 As shown, the obstacle grid map may include multiple grids, each grid may include obstacle attributes and non-obstacle attributes, the obstacle attributes represent that there are obstacles in the area corresponding to the grid, and the non-obstacle attributes represent that there are no obstacles in the area corresponding to the grid.

[0161] In the embodiment of the present application, when the obstacle points in the obstacle point map are converted into the above-mentioned obstacle grid map, the obstacle coordinates of the obstacle points in the obstacle point map relative to the vehicle can be determined.

[0162] Afterwards, since the obstacle grid map is also a grid map built with the vehicle as the center, its size is the actual size as that of the obstacle point map. Therefore, for each obstacle point in the obstacle point map, the target grid of the obstacle point in the obstacle grid map can be determined according to the obstacle coordinates corresponding to the obstacle point.

[0163] Finally, in order to distinguish the obstacle area from the idle area in the obstacle grid map, the attributes of the target grid corresponding to each obstacle point can be updated to the obstacle attributes, and after the attributes of the target grids corresponding to all obstacle points in the obstacle point map are updated to the obstacle attributes, the target obstacle grid map is obtained.

[0164] Among them, when the obstacle coordinates of the obstacle point are outside the obstacle grid map area, it means that the obstacle point is far away from the vehicle position and the obstacle point can be directly discarded.

[0165] Step 604: Determine the spatial area of ​​the spatial parking space for matching from the target obstacle grid map.

[0166] The above-mentioned spatial area refers to an area in the target obstacle grid map that can be used to match the spatial parking space, that is, an area where there are no obstacles around the vehicle body.

[0167] In an embodiment of the present application, in order to match a spatial parking space that can be used for parking in the area where the vehicle is currently located, the execution entity of the embodiment of the present application can first determine a spatial area that can be used to match the spatial parking space from the above-mentioned target obstacle grid map.

[0168] As an optional implementation method, since in actual applications, the spatial parking spaces of a vehicle can generally be divided into a left parking space located on the left side of the vehicle and a second parking space located on the right side of the vehicle, the executing entity of the embodiment of the present application can divide the target obstacle grid map into at least two obstacle grid sub-areas with the vehicle as the center.

[0169] As an exemplary implementation, the target obstacle grid map may be divided into an obstacle grid sub-area located on the left side of the vehicle and an obstacle grid sub-area located on the right side of the vehicle, with the vehicle as the center.

[0170] As another exemplary implementation, the target obstacle grid map may be divided into four obstacle grid sub-areas located in front, behind, left, and right of the vehicle, respectively, with the vehicle as the center.

[0171] Based on this, for each obstacle grid sub-region, the properties of the grid between the target grid in the obstacle grid sub-region and the preset boundary corresponding to the target grid are updated to obstacle properties to obtain an updated obstacle grid map. The preset boundary corresponding to the target grid refers to the boundary of the obstacle grid sub-region where the target grid is located in the target obstacle grid map.

[0172] For example, see Fig.11 A is a schematic diagram of a target obstacle grid map provided in an embodiment of the present application, such as Fig.11As shown in A, the target obstacle grid map can be divided into obstacle grid sub-areas located on the left and right sides of the vehicle respectively. The shadow grids in the above obstacle grid sub-areas are target grids with obstacle attributes. After that, the target grid and the grids between the preset boundaries corresponding to the target grid are updated to obstacle attributes, and the following can be obtained: Fig.11 Update the obstacle grid map shown in B. Fig.11 B is a schematic diagram of updating an obstacle grid map provided in an embodiment of the present application.

[0173] Afterwards, a non-obstacle grid (e.g. Fig.11 ).

[0174] Afterwards, the boundary of the non-obstruction grid can be extracted and divided into a real boundary and a virtual boundary, wherein the real boundary is the boundary between the non-obstruction grid and the target grid, and the virtual boundary is the boundary of the updated obstacle grid map.

[0175] Finally, the area consisting of the real boundary and the virtual boundary in the updated obstacle grid map can be determined as the spatial area for matching the spatial parking space.

[0176] Continue with Fig.11 Take the updated obstacle grid map shown in B as an example, see Fig.12 A is a schematic diagram of the boundary of a non-obstructive grid provided in an embodiment of the present application. Fig.12 As shown in A, the dotted line 1201 is the boundary of the extracted non-obstruction grid. Afterwards, the boundary of the non-obstruction grid can be divided into a real boundary and a virtual boundary, and the following is obtained: Fig.12 The spatial region shown in B, see Fig.12 B is a schematic diagram of a spatial region provided in an embodiment of the present application, such as Fig.12 As shown in B, Fig.12 In the boundary of the non-obstruction grid obtained by A, the boundary between the non-obstruction grid and the target grid can be used as the real boundary 1202 (the solid line part in the figure), and the boundary of the updated obstacle grid map can be used as the virtual boundary 1203 of the non-obstruction grid (the dotted line part in the figure).

[0177] Step 605: Generate an outer boundary box of the vehicle according to the size of the vehicle.

[0178] Step 606: Match multiple parking spaces from the spatial area according to the outer boundary box.

[0179] The following is a unified description of step 605 and step 606:

[0180] The outer boundary box is a rectangular box that completely contains the vehicle. It can be generated according to the size of the vehicle and is used to represent the area occupied by the vehicle.

[0181] In one embodiment, in order to match the spatial parking space of a vehicle in a spatial area, the execution entity of the embodiment of the present application can generate an outer boundary box that can characterize the area occupied by the vehicle according to the size of the vehicle, thereby matching the spatial parking space of the vehicle from the spatial area according to the outer boundary box.

[0182] As an optional implementation method, since the size of the parking space of the vehicle in actual applications is larger than the size of the vehicle, when generating the outer boundary box of the vehicle, each boundary of the vehicle can be expanded outward by a preset length (for example, 20 cm) according to the size of the vehicle itself, so as to obtain the outer boundary box corresponding to the vehicle.

[0183] In one embodiment, when matching multiple parking spaces from a spatial region according to an outer boundary box, a target region whose size is greater than or equal to the outer boundary box may be matched from the spatial region, and the target region is used as the parking space corresponding to the vehicle.

[0184] As an optional implementation, when matching a target area whose size is greater than or equal to the outer boundary box in the spatial region, the target real boundary for matching with the outer boundary box may be determined from the real boundaries included in the spatial region.

[0185] As an exemplary implementation, it may be determined whether there is a first target real boundary whose length is greater than the length of the outer boundary box in the real boundaries included in the above-mentioned spatial region.

[0186] Optionally, when it is determined that the first target real boundary exists, the first target real boundary may be determined as the target real boundary.

[0187] Optionally, when it is determined that there is no first target real boundary and there is a second target real boundary whose length is greater than the width of the outer boundary box, the second target real boundary may be determined as the target real boundary.

[0188] After determining the target real boundary, the inner point of the target real boundary can be determined, and based on the inner point, the outer boundary box is aligned with the target real boundary. The inner point is the point in the target real boundary that is closest to the vehicle.

[0189] As an exemplary implementation, when the target real boundary is the first target real boundary mentioned above, the long side of the outer boundary box can be aligned with the target real boundary; when the target real boundary is the second target real boundary mentioned above, the wide side of the outer boundary box can be aligned with the target real boundary.

[0190] After aligning the outer boundary box with the real boundary of the target, the outer boundary box and the virtual boundary in the space area can be interfered with to obtain a detection result. Optionally, when the detection result indicates that there is no interference relationship between the outer boundary box and any virtual boundary in the space area, the area corresponding to the outer boundary box is determined as the target area. The interference detection here can be used to detect whether the outer boundary box of the vehicle overlaps with the virtual boundary. Since the virtual boundary is the boundary of the obstacle grid map, if the two overlap, it means that if the vehicle is parked in the area where the current outer boundary box is located, there may be obstacles (such as curbs) that make it impossible to park the vehicle completely at that position.

[0191] Continue with Fig.12 Take the spatial region shown in B as an example, see Fig.13 , is a schematic diagram of a matching space parking space provided in an embodiment of the present application. Fig.13 As shown, after determining the outer boundary box 1301 of the vehicle, the target real boundary 1302 and the target real boundary 1303 can be matched. After that, the inner point and the outer point (boundary in the obstacle grid map) can be determined from the target real boundary 1302 and the target real boundary 1303, and after aligning the outer boundary box 1304 with the above-mentioned target real boundary 1302 and the target real boundary 1303, respectively, and performing inner point interference detection, the space parking space 1305 and the space parking space 1306 can be determined.

[0192] Afterwards, in order to match multiple spatial parking spaces, the boundary of the target area can be used as a real boundary and merged with the real boundary in the spatial area to obtain a new spatial area, which is an idle area that can continue to match spatial parking spaces.

[0193] Finally, the new spatial area can be used as the spatial area, and the above steps of matching a target area whose size is greater than or equal to the outer boundary box from the above spatial area and using the target area as the spatial parking space corresponding to the vehicle can be re-executed until the target area cannot be matched, thereby obtaining multiple spatial parking spaces matched from the spatial area.

[0194] Continue with Fig.13 For example, see Fig.14 , which is a schematic diagram of determining a new spatial area provided in an embodiment of the present application. Fig.14 A is a schematic diagram of determining multiple parking spaces provided in an embodiment of the present application, such as Fig.14 As shown in A, from the spatial region 1400, the spatial parking space 1401 and the spatial parking space 1402 are matched. Then, the boundaries of the above two spatial parking spaces are used as the real boundaries and merged with the real boundaries in the spatial region to obtain the following: Fig.14 The new space region shown in B is Fig.14B is a schematic diagram of a new spatial area provided in an embodiment of the present application.

[0195] The technical solution provided by the embodiment of the present application is to determine the obstacle coordinates of the obstacle point in the obstacle point map relative to the vehicle, determine the target grid of the obstacle point in the obstacle grid map according to the obstacle coordinates corresponding to the obstacle point for each obstacle point in the obstacle point map, update the attributes of the target grid corresponding to each obstacle point to the obstacle attribute, obtain the target obstacle grid map, determine the spatial area of ​​the spatial parking space for matching from the target obstacle grid map, generate the outer boundary box of the vehicle according to the size of the vehicle, and match multiple spatial parking spaces from the spatial area according to the outer boundary box. This technical solution projects the obstacle points in the obstacle point map of the vehicle into a preset obstacle grid map according to the obstacle coordinates relative to the vehicle, and determines the spatial area that can match the spatial parking space in the projected target obstacle grid map, so as to match multiple spatial parking spaces according to the outer boundary box of the vehicle according to the preset matching algorithm, and by matching the target obstacle grid map and the size of the outer boundary box of the vehicle, a spatial parking space with a more accurate size can be matched, thereby achieving more accurate determination of multiple spatial parking spaces.

[0196] See also Fig.15 , is a block diagram of an embodiment of a device for determining a parking space provided in an embodiment of the present application. As an embodiment, Fig.15 The device shown can be applied to a vehicle. Fig.15 As shown, the device may include:

[0197] An acquisition module 1501 is used to acquire a set of grounding points corresponding to obstacles around a vehicle body, wherein the set of grounding points includes a plurality of grounding points where the obstacles around the vehicle body contact the ground;

[0198] A mapping module 1502 is used to map the set of grounding points onto a preset plane where the vehicle is located, to obtain an obstacle point map corresponding to the vehicle;

[0199] A conversion module 1503, configured to convert the obstacle points in the obstacle point map into an obstacle grid map of a preset size corresponding to the vehicle, to obtain a target obstacle grid map;

[0200] The matching module 1504 is used to match the spatial parking space corresponding to the vehicle from the target obstacle grid map.

[0201] like Fig.16As shown, it is a schematic diagram of the structure of a vehicle provided in an embodiment of the present application, including a processor 161, a communication interface 162, a memory 163 and a communication bus 164, wherein the processor 161, the communication interface 162, and the memory 163 communicate with each other through the communication bus 164.

[0202] Memory 163, for storing computer programs;

[0203] In one embodiment of the present application, the processor 161 is used to execute the program stored in the memory 163 to implement the method for determining a parking space provided by any one of the aforementioned method embodiments, including:

[0204] Acquire a grounding point set corresponding to obstacles around a vehicle body, the grounding point set including a plurality of grounding points where the obstacles around the vehicle body contact the ground;

[0205] Mapping the ground point set to a preset plane where the vehicle is located to obtain an obstacle point map corresponding to the vehicle;

[0206] Converting obstacle points in the obstacle point map into an obstacle grid map of a preset size corresponding to the vehicle to obtain a target obstacle grid map;

[0207] A spatial parking space corresponding to the vehicle is matched from the target obstacle grid map.

[0208] An embodiment of the present application further provides a storage medium having a computer program stored thereon. When the computer program is executed by a processor, the steps of the method for determining a spatial parking space provided in any one of the aforementioned method embodiments are implemented.

[0209] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the present embodiment.

[0210] Through the text of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a general hardware platform, and of course, by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the relevant technology can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiment.

[0211] It should be understood that the terms used in the text are only for the purpose of the specific example implementation of the text, and are not intended to be limited. Unless the context clearly indicates otherwise, the singular forms "one", "an" and "said" as used in the text can also be represented to include plural forms. The terms "include", "comprise", "contain", and "have" are inclusive, and therefore specify the existence of the stated features, steps, operations, elements and / or parts, but do not exclude the existence or addition of one or more other features, steps, operations, elements, parts, and / or their combinations. The method steps, processes, and operations of the text in the text are not interpreted as necessarily requiring them to be performed in the specific order of the text or description, unless the execution order is clearly indicated. It should also be understood that additional or alternative steps can be used.

[0212] The foregoing is merely a specific embodiment of the present invention, which enables those skilled in the art to understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features claimed herein.

Claims

1. A method for determining a parking space, characterized in that: Applied to a vehicle, the method comprises: Acquire a grounding point set corresponding to obstacles around a vehicle body, wherein the grounding point set includes a plurality of grounding points where the obstacles around the vehicle body contact the ground; Mapping the ground point set to a preset plane where the vehicle is located to obtain an obstacle point map corresponding to the vehicle; Converting the obstacle points in the obstacle point map into an obstacle grid map of a preset size corresponding to the vehicle to obtain a target obstacle grid map; A spatial parking space corresponding to the vehicle is matched from the target obstacle grid map.

2. The method according to claim 1, characterized in that The vehicle body is pre-installed with a plurality of image acquisition modules and a plurality of distance detection modules, and the step of obtaining a set of grounding points corresponding to obstacles around the vehicle body includes: By means of the image acquisition module, a first grounding point set of obstacles around the vehicle body is acquired; By means of the distance detection module, a second grounding point set of obstacles around the vehicle body is collected; The first grounding point set and the second grounding point set are determined as grounding point sets corresponding to obstacles around a body of the vehicle.

3. The method according to claim 2, characterized in that The collecting of a first grounding point set of obstacles around the vehicle body by the image collection module includes: Obstacle images of obstacles around the vehicle body are collected by the image collection module; Inputting the obstacle image into a pre-trained semantic segmentation model to obtain grounding point lines between the obstacles around the vehicle body and the ground output by the semantic segmentation model; extracting the first grounding point set from the grounding point lines; or, Inputting the obstacle image into a pre-trained key point detection model to obtain a first grounding point set of obstacles around the vehicle body output by the key point detection model; The collecting of a second grounding point set of obstacles around the vehicle body by the distance detection module includes: Obtaining the obstacle coordinates of the obstacles around the vehicle body relative to the vehicle body through the distance detection module; The grounding point represented by the obstacle coordinates is included in the second grounding point set.

4. The method according to claim 2, characterized in that: Mapping the grounding point set to the preset plane where the vehicle is located to obtain an obstacle point map corresponding to the vehicle includes: Using the inverse perspective transformation principle, projecting the first grounding point in the first grounding point set onto the preset plane to obtain a first obstacle point map including a plurality of first obstacle points; Determine, from the first obstacle point map, an obstacle area including a plurality of the first obstacle points, wherein a size of the obstacle area is smaller than a preset size; For each of the obstacle areas, determining the distance between each of the first obstacle points in the obstacle area and the vehicle body; retaining the first obstacle point with the smallest distance among the multiple first obstacle points included in the obstacle area, to obtain a second obstacle point map; By using the inverse perspective transformation principle, the second ground point in the second ground point set is projected into the second obstacle point map to obtain the obstacle point map corresponding to the vehicle.

5. The method according to claim 1, characterized in that Mapping the grounding point set to the preset plane where the vehicle is located to obtain an obstacle point map corresponding to the vehicle includes: Mapping the ground point set to a preset plane where the vehicle is located to obtain an initial obstacle point map corresponding to the vehicle; Establishing a global coordinate system based on the current position of the vehicle, and acquiring a plurality of running grounding point sets collected at different positions during the running of the vehicle, wherein the running grounding point sets include a plurality of running grounding points, and the running grounding points are represented by running grounding point coordinates relative to the vehicle body; Converting the operating grounding point coordinates of the plurality of operating grounding points into target grounding point coordinates in a global coordinate system; According to the target touchdown point coordinates, the operating touchdown points included in the operating touchdown point set are converted to the initial obstacle point map by using the inverse perspective transformation principle to obtain the obstacle point map corresponding to the vehicle.

6. The method according to claim 1, characterized in that The obstacle grid map is centered on the vehicle, the grids in the obstacle grid map include obstacle attributes and non-obstacle attributes, and the obstacle points in the obstacle point map are converted to an obstacle grid map of a preset size corresponding to the vehicle to obtain a target obstacle grid map, including: Determining obstacle coordinates of an obstacle point in the obstacle point map relative to the vehicle; For each obstacle point in the obstacle point map, determining a target grid of the obstacle point in the obstacle grid map according to obstacle coordinates corresponding to the obstacle point; The attributes of the target grid corresponding to each obstacle point are updated to obstacle attributes to obtain the target obstacle grid map.

7. The method according to claim 6, characterized in that The step of matching a space parking space corresponding to the vehicle from the target obstacle grid map includes: Determining a spatial area for matching the spatial parking space from the target obstacle grid map; Generating an outer bounding box of the vehicle according to the size of the vehicle; According to the outer boundary box, a plurality of the spatial parking spaces are matched from the spatial area.

8. The method according to claim 7, characterized in that The step of determining a spatial area for matching the spatial parking space from the target obstacle grid map includes: Taking the vehicle as the center, dividing the target obstacle grid map into at least two obstacle grid sub-areas; For each obstacle grid sub-region, the attributes of the grids between the target grid in the obstacle grid sub-region and the preset boundary corresponding to the target grid are updated to obstacle attributes to obtain an updated obstacle grid map; Determine a non-obstacle grid having a non-obstacle attribute from the updated obstacle grid map; Extracting the boundary of the non-obstruction grid, and dividing the boundary of the non-obstruction grid into a real boundary and a virtual boundary; wherein the real boundary is the boundary between the non-obstruction grid and the target grid, and the virtual boundary is the boundary of the updated obstacle grid map; An area consisting of the real boundary and the virtual boundary in the updated obstacle grid map is determined as a spatial area for matching the spatial parking space.

9. The method according to claim 8, characterized in that The step of matching a plurality of the spatial parking spaces from the spatial region according to the outer boundary box includes: From the spatial region, matching a target region whose size is greater than or equal to the outer bounding box; Using the target area as the space parking space corresponding to the vehicle; Taking the boundary of the target area as the real boundary and fusing it with the real boundary of the space area to obtain a new space area; The new spatial area is used as the spatial area, and the steps of matching a target area with a size greater than or equal to the outer boundary box from the spatial area and using the target area as the spatial parking space corresponding to the vehicle are re-executed until the target area cannot be matched, thereby obtaining multiple spatial parking spaces matched from the spatial area.

10. The method according to claim 9, characterized in that The matching, from the spatial region, of a target region whose size is greater than or equal to the outer bounding box comprises: Determine, from the real boundaries included in the spatial region, a target real boundary for matching with the outer boundary box, and determine an inner point of the target real boundary; wherein the inner point is a point in the target real boundary that is closest to the vehicle; Based on the inner point, aligning the outer bounding box with the real boundary of the object; Perform interference detection on the outer boundary box and the virtual boundary in the space area to obtain a detection result; When the detection result indicates that there is no interference relationship between the outer boundary box and any of the virtual boundaries in the spatial region, the region currently corresponding to the outer boundary box is determined as the target region.

11. The method according to claim 10, characterized in that The determining, from the real boundaries included in the spatial region, a target real boundary for matching with the outer boundary box comprises: In the case where, in the real boundaries included in the determined spatial region, there is a first target real boundary whose length is greater than the length of the outer boundary box, determining the first target real boundary as the target real boundary; In determining the real boundaries included in the spatial region, if there is no first target real boundary whose length is greater than the length of the outer boundary box, and there is a second target real boundary whose length is greater than the width of the outer boundary box, determining the second target real boundary as the target real boundary; The step of aligning the outer boundary box with the real boundary of the target based on the inner point comprises: In a case where the target real boundary is the first target real boundary, aligning the long side of the outer boundary box with the target real boundary; In a case where the target real boundary is the second target real boundary, the broad side of the outer boundary box is aligned with the target real boundary.

12. A device for determining a parking space, characterized in that: Applied to a vehicle, the device comprises: An acquisition module, used to acquire a set of grounding points corresponding to obstacles around a vehicle body, wherein the set of grounding points includes a plurality of grounding points where the obstacles around the vehicle body contact the ground; A mapping module, used to map the set of grounding points onto a preset plane where the vehicle is located, to obtain an obstacle point map corresponding to the vehicle; A conversion module, used to convert the obstacle points in the obstacle point map into an obstacle grid map of a preset size corresponding to the vehicle to obtain a target obstacle grid map; A matching module is used to match the spatial parking space corresponding to the vehicle from the target obstacle grid map.

Images