Parking space entrance identification method and device

Abstract

The invention relates to a parking space entrance identification method and device, and belongs to the technical field of automatic parking, and the identification method comprises the steps: splicing image information returned by four cameras of a vehicle to obtain an aerial view, and identifying the aerial view through a deep learning algorithm to obtain a first parking space, according to the distance from each angular point to the vehicle, adjusting the angular point serial number corresponding to the angular point to obtain a target angular point serial number corresponding to each angular point, and according to the target angular point serial number of each angular point, determining an entrance of the first parking space; according to the invention, corner detection and corner sorting are carried out on the parking spaces in the parking lot by means of the camera, and the parking space entrance is determined according to the corner sequence number, so that the accuracy and efficiency of parking space entrance identification are effectively improved.

Description

Technical Field

[0001] The present application relates to the technical field of automatic parking, and in particular to a parking space entrance recognition method and device. Background Art

[0002] At present, automatic parking is a powerful part of smart travel and assisted driving. In terms of robustness and safety, automatic parking needs to be divided into visual perception and radar ranging to run simultaneously in order to park safely, accurately and efficiently. The visual algorithm is of course deep learning, and the visual perception algorithm is the main force in the stage of finding parking spaces.

[0003] The existing automatic parking system uses a visual perception algorithm that can identify the four corner points of a parking space. However, the visual perception algorithm cannot determine the direction of the parking space entrance when detecting the corner points of the parking space. The corner points of the parking space need to be drawn into an image, and then the user needs to further determine the entrance to the parking space. Since it involves manual operation, it is easy to cause inaccurate identification of the parking space entrance and low efficiency. Summary of the invention

[0004] To this end, the present invention provides a parking space entrance recognition method and device to solve the problems of inaccurate and low efficiency parking space entrance recognition in existing automatic parking systems.

[0005] In a first aspect, a parking space entrance recognition method is provided, the method comprising:

[0006] Acquire a bird's-eye view of the vehicle; the bird's-eye view includes image data of the vehicle and its surroundings;

[0007] The first parking space is obtained by identifying the bird's-eye view through a deep learning algorithm; the first parking space includes four corner points with corner point serial numbers;

[0008] Adjusting the corner point sequence number corresponding to each corner point according to the distance from each corner point to the vehicle to obtain the target corner point sequence number corresponding to each corner point;

[0009] The entrance of the first parking space is determined according to the target corner point sequence number of each corner point.

[0010] Further, the four corner point numbers of the first parking space are 0, 1, 2, and 3 in counterclockwise order; and the step of adjusting the corner point number corresponding to the corner point according to the distance from each corner point to the vehicle to obtain the target corner point number corresponding to each corner point includes:

[0011] A plane rectangular coordinate system is established based on the bird's-eye view with the rear axle of the vehicle as the origin; the distance from the corner point to the vehicle is the distance from the corner point to the origin;

[0012] If the distance from any corner point of the first parking space to the origin is less than the preset boundary distance, the position of the first parking space is determined according to the corner point coordinates of each corner point in the plane rectangular coordinate system; the position information of the first parking space includes whether the first parking space is on the left side, right side, front side or rear side of the vehicle;

[0013] The corner point sequence number corresponding to each corner point is adjusted according to the position information and the distance from the corner point to the vehicle to obtain the target corner point sequence number corresponding to each corner point.

[0014] Further, adjusting the corner point sequence number corresponding to each corner point according to the position information and the distance from the corner point to the vehicle to obtain the target corner point sequence number corresponding to each corner point includes:

[0015] If the first parking space is on the left side of the vehicle, obtain the two corner points closest to the vehicle and the second closest to the vehicle, set the corner point with the smaller ordinate value of the two corner points as the initial corner point, and set the target corner point sequence number of the initial corner point to 0;

[0016] The other three corner points of the first parking space are set with the initial corner point as the starting point, and the target corner point numbers of the other three corner points are set to 1, 2, and 3 in a counterclockwise order.

[0017] Further, adjusting the corner point sequence number corresponding to each corner point according to the position information and the distance from the corner point to the vehicle to obtain the target corner point sequence number corresponding to each corner point includes:

[0018] If the first parking space is on the right side of the vehicle, obtain the two corner points closest to the vehicle and the second closest to the vehicle, set the corner point with the larger ordinate value as the initial corner point, and set the target corner point sequence number of the initial corner point to 0;

[0019] The other three corner points of the first parking space are set with the initial corner point as the starting point, and the target corner point numbers of the other three corner points are set to 1, 2, and 3 in a counterclockwise order.

[0020] Further, adjusting the corner point sequence number corresponding to each corner point according to the position information and the distance from the corner point to the vehicle to obtain the target corner point sequence number corresponding to each corner point includes:

[0021] If the first parking space is in front of the vehicle, obtain the two corner points closest to the vehicle and the second closest to the vehicle, set the corner point with the smaller horizontal coordinate value of the two corner points as the initial corner point, and set the target corner point sequence number of the initial corner point to 0;

[0022] The other three corner points of the first parking space are set with the initial corner point as the starting point, and the target corner point numbers of the other three corner points are set to 1, 2, and 3 in a counterclockwise order.

[0023] Further, adjusting the corner point sequence number corresponding to each corner point according to the position information and the distance from the corner point to the vehicle to obtain the target corner point sequence number corresponding to each corner point includes:

[0024] If the first parking space is behind the vehicle, obtain the two corner points closest to the vehicle and the second closest to the vehicle, set the corner point with the larger horizontal coordinate value of the two corner points as the initial corner point, and set the target corner point sequence number of the initial corner point to 0;

[0025] The other three corner points of the first parking space are set with the initial corner point as the starting point, and the target corner point numbers of the other three corner points are set to 1, 2, and 3 in a counterclockwise order.

[0026] Further, the step of determining the entrance of the first parking space according to the target corner point sequence number of each corner point includes:

[0027] The entrance formed by the two corner points whose target corner point numbers are 0 and 1 is used as the entrance to the first parking space.

[0028] Furthermore, the determining the position of the first parking space according to the corner point coordinates of each corner point in the plane rectangular coordinate system includes:

[0029] If the horizontal coordinate values ​​of the four corner points of the first parking space are all less than 0, then the first parking space is on the left side of the vehicle;

[0030] If the horizontal coordinate values ​​of the four corner points of the first parking space are all greater than 0, then the first parking space is on the right side of the vehicle;

[0031] If the ordinate values ​​of the four corner points of the first parking space are all greater than 0, and the abscissa values ​​of the four corner points are positive and negative, then the first parking space is in front of the vehicle;

[0032] If the ordinate values ​​of the four corner points of the first parking space are all less than 0, and the abscissa values ​​of the four corner points are positive and negative, then the first parking space is behind the vehicle.

[0033] Furthermore, it also includes:

[0034] Obtaining a bird's-eye view of the vehicle once every preset time period;

[0035] Performing entrance recognition operations on the parking spaces identified in the bird's-eye view one by one within the preset time, until the entrance recognition of all parking spaces is completed, then stopping the current round of recognition and waiting for the next round of recognition,

[0036] Alternatively, the entrance recognition operation is performed on the parking spaces identified in the bird's-eye view one by one until the preset time period is exceeded, then the current round of recognition is stopped and the next round of recognition is performed.

[0037] In a second aspect, a parking space entrance recognition device is provided, the device comprising:

[0038] An acquisition module, used to acquire a bird's-eye view of the vehicle; the bird's-eye view includes image data of the vehicle and its surroundings;

[0039] A corner point recognition module, configured to recognize the bird's-eye view through a deep learning algorithm to obtain a first parking space; the first parking space includes four corner points with corner point serial numbers;

[0040] A corner point number adjustment module, used to adjust the corner point number corresponding to each corner point according to the distance from each corner point to the vehicle to obtain a target corner point number corresponding to each corner point;

[0041] An entrance identification module is used to determine the entrance of the first parking space according to the target corner point sequence number of each corner point.

[0042] The present invention adopts the above technical solution and has at least the following beneficial effects:

[0043] A parking space entrance recognition method and device are provided, wherein image information sent back by four cameras of a vehicle is spliced ​​to obtain a bird's-eye view, and the bird's-eye view is recognized by a deep learning algorithm to obtain a first parking space, wherein the first parking space includes four corner points with corner point numbers, and the corner point numbers corresponding to the corner points are adjusted according to the distance from each corner point to the vehicle to obtain the target corner point numbers corresponding to each corner point, and the entrance of the first parking space is determined according to the target corner point numbers of each corner point; the present invention detects the corner points of parking spaces in a parking lot with the help of cameras and sorts the corner points, and determines the entrance of the parking space according to the corner point numbers, thereby effectively improving the accuracy and efficiency of parking space entrance recognition.

[0044] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] 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 description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0046] Figure 1 is a flow chart of a parking space entrance identification method shown in an exemplary embodiment of the present invention;

[0047] Figure 2 is a schematic block diagram showing an exemplary embodiment of the present invention in which the first parking space is on the left side of the vehicle;

[0048] Figure 3 is a schematic block diagram showing an exemplary embodiment of the present invention in which the first parking space is on the right side of the vehicle;

[0049] Figure 4 is a schematic block diagram showing a first parking space in front of a vehicle according to an exemplary embodiment of the present invention;

[0050] Figure 5 is a schematic block diagram showing a first parking space behind a vehicle according to an exemplary embodiment of the present invention;

[0051] Figure 6 It is a schematic block diagram of a parking space entrance recognition device shown in an exemplary embodiment of the present invention. DETAILED DESCRIPTION

[0052] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0053] Problems with the existing technology: When detecting corner points of parking spaces, the visual perception algorithm sets corner point numbers for the corner points to identify the identified parking spaces. However, the corner point numbers of parking spaces cannot determine the direction of the parking space entrance. To determine the direction of the entrance, other auxiliary means are needed.

[0054] The embodiments of the present application provide a parking space entrance recognition method and device. The point sorting algorithm for parking space line corner point detection based on deep learning can sort the parking space corner point detection results of the visual perception algorithm and output the accurate parking space entrance direction, which provides great convenience and strong support for subsequent parking planning and control.

[0055] The method and device of the present invention are described below through specific embodiments.

[0056] See also Figure 1 , Figure 1 is a flow chart of a parking space entrance identification method shown in an exemplary embodiment of the present invention, see Figure 1 , the method comprising:

[0057] Step S11, obtaining a bird's-eye view of the vehicle; the bird's-eye view includes image data of the vehicle and its surroundings;

[0058] Step S12: obtaining a first parking space by identifying the bird's-eye view through a deep learning algorithm; the first parking space includes four corner points with corner point serial numbers;

[0059] Step S13, adjusting the corner point sequence number corresponding to each corner point according to the distance from each corner point to the vehicle to obtain the target corner point sequence number corresponding to each corner point;

[0060] Step S14: determining the entrance of the first parking space according to the target corner point sequence number of each corner point.

[0061] It should be noted that the technical solution provided in this embodiment can be loaded into the automatic parking system in the form of a small program in specific practice, or can be used as an independent application to realize parking entrance recognition. Applicable scenarios include but are not limited to: automatic parking and other scenarios that require parking entrance recognition.

[0062] Specifically, when the first parking space is obtained by identifying the bird's-eye view through a deep learning algorithm, it is necessary to consider whether the identified parking space meets the parking space requirements. For example, the length and width of the parking space fall within the standard parking space range, and the parking space is identified as a valid parking space, and then the parking space corner point identification is performed; the deep learning algorithm adopts one of the existing recognition algorithms.

[0063] It can be understood that the recognition method provided in this embodiment splices the image information sent back by the four cameras of the vehicle to obtain a bird's-eye view, and recognizes the bird's-eye view through a deep learning algorithm to obtain the first parking space, wherein the first parking space includes four corner points with corner point numbers, and the corner point numbers corresponding to the corner points are adjusted according to the distance from each corner point to the vehicle to obtain the target corner point numbers corresponding to each corner point, and the entrance to the first parking space is determined according to the target corner point numbers of each corner point; the present invention detects the corner points of the parking spaces in the parking lot with the help of cameras and sorts the corner points, and determines the entrance to the parking spaces according to the corner point numbers, thereby effectively improving the accuracy and efficiency of parking space entrance recognition.

[0064] In specific practice, step S11 "obtaining a bird's-eye view of the vehicle" includes: performing AVM stitching based on the image information sent back by the four cameras to obtain a bird's-eye view of the current frame; obtaining a bird's-eye view of the vehicle once every preset time period; performing entrance recognition operations on the parking spaces identified in the bird's-eye view one by one within the preset time period, until the entrances of all parking spaces are identified, then stopping this round of recognition and waiting for the next round of recognition, or, performing entrance recognition operations on the parking spaces identified in the bird's-eye view one by one, until the preset time period is exceeded, then stopping this round of recognition and proceeding to the next round of recognition.

[0065] It should be noted that the preset time is set according to the vehicle speed and the number of parking spaces, and the preset time is shortened as the vehicle speed and the number of parking spaces increase.

[0066] In specific practice, in step S12 "obtaining the first parking space by identifying the bird's-eye view through a deep learning algorithm", it is necessary to consider whether the identified first parking space meets the requirements of a standard parking space, and the deep learning algorithm adopts one of the existing recognition algorithms.

[0067] In specific practice, step S13 "adjusting the corner point number corresponding to the corner point according to the distance from each corner point to the vehicle to obtain the target corner point number corresponding to each corner point" includes: establishing a plane rectangular coordinate system with the rear axle of the vehicle as the origin based on the bird's-eye view; the distance from the corner point to the vehicle is the distance from the corner point to the origin; if the distance from any corner point of the first parking space to the origin is less than the preset boundary distance, then determining the position of the first parking space according to the corner point coordinates of each corner point in the plane rectangular coordinate system; the position information of the first parking space includes whether the first parking space is on the left side, right side, front or rear of the vehicle; adjusting the corner point number corresponding to each corner point according to the position information and the distance from the corner point to the vehicle to obtain the target corner point number corresponding to each corner point.

[0068] It should be noted that the preset boundary distance is generally set to 10m, and can also be set according to specific needs; the four corner points of the first parking space are numbered 0, 1, 2, and 3 in counterclockwise order; each corner point number is assigned by a deep learning algorithm.

[0069] Specifically, the position of the first parking space is determined according to the coordinates of each corner point in the plane rectangular coordinate system, including: if the horizontal coordinate values ​​of the four corner points of the first parking space are all less than 0, the first parking space is on the left side of the vehicle; if the horizontal coordinate values ​​of the four corner points of the first parking space are all greater than 0, the first parking space is on the right side of the vehicle; if the vertical coordinate values ​​of the four corner points of the first parking space are all greater than 0, and the horizontal coordinate values ​​of the four corner points are positive and negative numbers, the first parking space is in front of the vehicle; if the vertical coordinate values ​​of the four corner points of the first parking space are all less than 0, and the horizontal coordinate values ​​of the four corner points are positive and negative numbers, the first parking space is behind the vehicle.

[0070] See also Figure 2 , Figure 2 is a schematic block diagram showing an exemplary embodiment of the present invention in which the first parking space is on the left side of the vehicle, see Figure 2 If the first parking space is on the left side of the vehicle, obtain the two corner points closest to the vehicle and the second closest to the vehicle, set the corner point with the smaller ordinate value as the initial corner point, and set the target corner point sequence number of the initial corner point to 0; take the other three corner points of the first parking space as the starting point, and set the target corner point sequences of the other three corner points to 1, 2, and 3 in a counterclockwise order.

[0071] See also Figure 3 , Figure 3 is a schematic block diagram showing an exemplary embodiment of the present invention in which the first parking space is on the right side of the vehicle, see Figure 3If the first parking space is on the right side of the vehicle, obtain the two corner points closest to the vehicle and the second closest to the vehicle, set the corner point with the larger ordinate value as the initial corner point, and set the target corner point sequence number of the initial corner point to 0; take the initial corner point as the starting point for the other three corner points of the first parking space, and set the target corner point sequences of the other three corner points to 1, 2, and 3 in a counterclockwise order.

[0072] See also Figure 4 , Figure 4 is a schematic block diagram showing an exemplary embodiment of the present invention in which the first parking space is in front of the vehicle, see Figure 4 If the first parking space is in front of the vehicle, obtain the two corner points closest to the vehicle and the second closest to the vehicle, set the corner point with the smaller horizontal coordinate value of the two corner points as the initial corner point, and set the target corner point sequence number of the initial corner point to 0; take the other three corner points of the first parking space as the starting point, and set the target corner point sequences of the other three corner points to 1, 2, and 3 in a counterclockwise order.

[0073] See also Figure 5 , Figure 5 is a schematic block diagram showing an exemplary embodiment of the present invention in which the first parking space is behind the vehicle, see Figure 5 If the first parking space is behind the vehicle, obtain the two corner points closest to the vehicle and the second closest to the vehicle, set the corner point with the larger horizontal coordinate value as the initial corner point, and set the target corner point sequence number of the initial corner point to 0; take the other three corner points of the first parking space as the starting point, and set the target corner point sequences of the other three corner points to 1, 2, and 3 in a counterclockwise order.

[0074] It can be understood that the identification method provided in this embodiment can adjust the serial number of the parking space corner point so as to subsequently identify the parking space entrance efficiently and quickly.

[0075] In specific practice, step S14 "determining the entrance of the first parking space according to the target corner point serial number of each corner point" includes: taking the entrance formed by the two corner points with target corner point serial numbers 0 and 1 as the entrance of the first parking space.

[0076] It is understandable that the recognition method provided in this embodiment can quickly determine the parking space entrance directly through the marked target corner point numbers when automatic parking is required, thereby improving the efficiency of automatic parking.

[0077] See also Figure 6 , Figure 6 is a schematic block diagram of a parking space entrance recognition device according to an exemplary embodiment of the present invention, see Figure 6 , the parking space entrance recognition device 100 includes:

[0078] The acquisition module 101 is used to acquire a bird's-eye view of the vehicle; the bird's-eye view includes image data of the vehicle and its surroundings;

[0079] A corner point recognition module 102 is used to recognize the bird's-eye view image through a deep learning algorithm to obtain a first parking space; the first parking space includes four corner points with corner point serial numbers;

[0080] A corner point number adjustment module 103 is used to adjust the corner point number corresponding to each corner point according to the distance from each corner point to the vehicle to obtain a target corner point number corresponding to each corner point;

[0081] The entrance identification module 104 is used to determine the entrance of the first parking space according to the target corner point sequence number of each corner point.

[0082] It should be noted that the technical solution provided in this embodiment can be applied in specific practice to scenarios including, but not limited to: automatic parking and other scenarios that require identification of parking space entrances.

[0083] Specifically, when the first parking space is obtained by identifying the bird's-eye view through a deep learning algorithm, it is necessary to consider whether the identified parking space meets the parking space requirements. For example, the length and width of the parking space fall within the standard parking space range, and the parking space is identified as a valid parking space, and then the parking space corner point identification is performed; the deep learning algorithm adopts one of the existing recognition algorithms.

[0084] It should be noted that the preset standard opening and closing degree threshold can be set by human experience; it can also be obtained through model learning.

[0085] It can be understood that the recognition device provided in this embodiment splices the image information sent back by the four cameras of the vehicle to obtain a bird's-eye view, and recognizes the bird's-eye view through a deep learning algorithm to obtain the first parking space, wherein the first parking space includes four corner points with corner point numbers, and the corner point numbers corresponding to the corner points are adjusted according to the distance from each corner point to the vehicle to obtain the target corner point numbers corresponding to each corner point, and the entrance to the first parking space is determined according to the target corner point numbers of each corner point; the present invention detects the corner points of the parking spaces in the parking lot with the help of cameras and sorts the corner points, and determines the entrance of the parking spaces according to the corner point numbers, thereby effectively improving the accuracy and efficiency of parking space entrance recognition.

[0086] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

[0087] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0088] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0089] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. A parking space entrance recognition method, characterized in that: The method comprises: Acquire a bird's-eye view of the vehicle; the bird's-eye view includes image data of the vehicle and its surroundings; The first parking space is obtained by identifying the bird's-eye view through a deep learning algorithm; the first parking space includes four corner points with corner point serial numbers; Adjusting the corner point sequence number corresponding to each corner point according to the distance from each corner point to the vehicle to obtain the target corner point sequence number corresponding to each corner point; The entrance of the first parking space is determined according to the target corner point sequence number of each corner point.

2. The identification method according to claim 1, characterized in that: The four corner points of the first parking space are numbered 0, 1, 2, and 3 in counterclockwise order; and adjusting the corner point number corresponding to the corner point according to the distance from each corner point to the vehicle to obtain the target corner point number corresponding to each corner point includes: A plane rectangular coordinate system is established based on the bird's-eye view with the rear axle of the vehicle as the origin; the distance from the corner point to the vehicle is the distance from the corner point to the origin; If the distance from any corner point of the first parking space to the origin is less than the preset boundary distance, the position of the first parking space is determined according to the corner point coordinates of each corner point in the plane rectangular coordinate system; the position information of the first parking space includes whether the first parking space is on the left side, right side, front side or rear side of the vehicle; The corner point sequence number corresponding to each corner point is adjusted according to the position information and the distance from the corner point to the vehicle to obtain the target corner point sequence number corresponding to each corner point.

3. The identification method according to claim 2, characterized in that: The step of adjusting the corner point sequence number corresponding to each corner point according to the position information and the distance from the corner point to the vehicle to obtain the target corner point sequence number corresponding to each corner point includes: If the first parking space is on the left side of the vehicle, obtain the two corner points closest to the vehicle and the second closest to the vehicle, set the corner point with the smaller ordinate value of the two corner points as the initial corner point, and set the target corner point sequence number of the initial corner point to 0; The other three corner points of the first parking space are set with the initial corner point as the starting point, and the target corner point numbers of the other three corner points are set to 1, 2, and 3 in a counterclockwise order.

4. The identification method according to claim 2, characterized in that: The step of adjusting the corner point sequence number corresponding to each corner point according to the position information and the distance from the corner point to the vehicle to obtain the target corner point sequence number corresponding to each corner point includes: If the first parking space is on the right side of the vehicle, obtain the two corner points closest to the vehicle and the second closest to the vehicle, set the corner point with the larger ordinate value as the initial corner point, and set the target corner point sequence number of the initial corner point to 0; The other three corner points of the first parking space are set with the initial corner point as the starting point, and the target corner point numbers of the other three corner points are set to 1, 2, and 3 in a counterclockwise order.

5. The identification method according to claim 2, characterized in that: The step of adjusting the corner point sequence number corresponding to each corner point according to the position information and the distance from the corner point to the vehicle to obtain the target corner point sequence number corresponding to each corner point includes: If the first parking space is in front of the vehicle, obtain the two corner points closest to the vehicle and the second closest to the vehicle, set the corner point with the smaller horizontal coordinate value of the two corner points as the initial corner point, and set the target corner point sequence number of the initial corner point to 0; The other three corner points of the first parking space are set with the initial corner point as the starting point, and the target corner point numbers of the other three corner points are set to 1, 2, and 3 in a counterclockwise order.

6. The identification method according to claim 2, characterized in that: The step of adjusting the corner point sequence number corresponding to each corner point according to the position information and the distance from the corner point to the vehicle to obtain the target corner point sequence number corresponding to each corner point includes: If the first parking space is behind the vehicle, obtain the two corner points closest to the vehicle and the second closest to the vehicle, set the corner point with the larger horizontal coordinate value of the two corner points as the initial corner point, and set the target corner point sequence number of the initial corner point to 0; The other three corner points of the first parking space are set with the initial corner point as the starting point, and the target corner point numbers of the other three corner points are set to 1, 2, and 3 in a counterclockwise order.

7. The identification method according to any one of claims 3 to 6, characterized in that: The step of determining the entrance of the first parking space according to the target corner point sequence number of each corner point includes: The entrance formed by the two corner points whose target corner point numbers are 0 and 1 is used as the entrance to the first parking space.

8. The identification method according to claim 2, characterized in that: The determining the position of the first parking space according to the corner point coordinates of each corner point in the plane rectangular coordinate system includes: If the horizontal coordinate values ​​of the four corner points of the first parking space are all less than 0, then the first parking space is on the left side of the vehicle; If the horizontal coordinate values ​​of the four corner points of the first parking space are all greater than 0, then the first parking space is on the right side of the vehicle; If the ordinate values ​​of the four corner points of the first parking space are all greater than 0, and the abscissa values ​​of the four corner points are positive and negative, then the first parking space is in front of the vehicle; If the ordinate values ​​of the four corner points of the first parking space are all less than 0, and the abscissa values ​​of the four corner points are positive and negative, then the first parking space is behind the vehicle.

9. The identification method according to claim 8, characterized in that: Also includes: Obtaining a bird's-eye view of the vehicle once every preset time period; Performing entrance recognition operations on the parking spaces identified in the bird's-eye view one by one within the preset time, until the entrance recognition of all parking spaces is completed, then stopping the current round of recognition and waiting for the next round of recognition, Alternatively, the entrance recognition operation is performed on the parking spaces identified in the bird's-eye view one by one until the preset time period is exceeded, then the current round of recognition is stopped and the next round of recognition is performed.

10. A parking space entrance identification device, characterized in that: The device comprises: An acquisition module, used to acquire a bird's-eye view of the vehicle; the bird's-eye view includes image data of the vehicle and its surroundings; A corner point recognition module, configured to recognize the bird's-eye view through a deep learning algorithm to obtain a first parking space; the first parking space includes four corner points with corner point serial numbers; A corner point number adjustment module, used to adjust the corner point number corresponding to each corner point according to the distance from each corner point to the vehicle to obtain a target corner point number corresponding to each corner point; An entrance identification module is used to determine the entrance of the first parking space according to the target corner point sequence number of each corner point.

Images