Automatic parking method, device and equipment of vehicle and storage medium

Abstract

The invention provides an automatic parking method, device and equipment for a vehicle and a storage medium, and relates to the technical field of automatic driving, and the method comprises the steps: determining an initial position corresponding to a target parking space where the vehicle is to be parked; controlling the vehicle to run to the initial position; acquiring a first image of the surrounding environment of the vehicle when the vehicle is at the initial position; processing the first image to obtain a parking label corresponding to the vehicle; and controlling the vehicle to park in the target parking space according to a parking instruction associated with the parking label. Therefore, the parking control of the vehicle is finer and more accurate, the alignment precision of the vehicle and the target parking space is improved, the safety of parking at the target parking space can be improved, the risk that the vehicle collides with the mechanical parking space or scratches a hub is reduced, and the problem that parking at the mechanical parking space is difficult is solved; the vehicle in the embodiment has very low requirements on computing power and sensors, and can be adapted to various new energy vehicle types.

Description

Technical Field

[0001] The present disclosure relates to the field of autonomous driving technology, and in particular, to an automatic parking method, device, equipment, and storage medium for a vehicle. Background Art

[0002] The width of a conventional mechanical garage deep groove is generally between 2.0 and 2.1 meters. For a vehicle with a width close to 2 meters, after parking in the center, the distances on both sides are about 5 centimeters. Even a slight deviation will cause collision with the entrance or scratching of the wheel hubs. Therefore, mechanical garage spaces are a difficult problem for the automatic parking assistance system.

[0003] For mechanical garage spaces, it is generally considered that the accuracy of the garage space map needs to reach the order of 5 centimeters to park safely. The traditional parking method is to construct a garage space map by geometric modeling using lidar or vision, but the accuracy is usually only in the order of 10 - 20 centimeters, and there is still a certain risk of collision between the vehicle and the mechanical garage space or scratching of the wheel hubs.

[0004] Therefore, how to improve the safety of a vehicle automatically parking in a mechanical garage space is an urgent problem to be solved currently. Summary of the Invention

[0005] The present application provides an automatic parking method, device, and storage medium for a vehicle, aiming to at least solve one of the technical problems in the related art to some extent.

[0006] The first aspect of the embodiments of the present application provides an automatic parking method for a vehicle, including:

[0007] Determine an initial position corresponding to a target garage space to which the vehicle is to be parked;

[0008] Control the vehicle to travel to the initial position;

[0009] Obtain a first image of the environment around the vehicle when the vehicle is at the initial position;

[0010] Process the first image to obtain a parking label corresponding to the vehicle;

[0011] Control the vehicle to park in the target garage space according to a parking instruction associated with the parking label.

[0012] The second aspect of the embodiments of the present application provides an automatic parking device for a vehicle, including:

[0013] A first determination module, configured to determine an initial position corresponding to a target garage space to which the vehicle is to be parked;

[0014] A first control module, configured to control the vehicle to travel to the initial position;

[0015] An acquisition module, configured to acquire a first image of the environment around the vehicle when the vehicle is at the initial position;

[0016] A processing module, configured to process the first image to obtain a parking label corresponding to the vehicle;

[0017] A second control module, configured to control the vehicle to park in the target parking space according to a parking instruction associated with the parking label.

[0018] An embodiment of the third aspect of the present application provides an electronic device, including: at least one processor; and a memory communicatively connected to the at least one processor; wherein, the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the automatic parking method of the vehicle according to the embodiments of the present application.

[0019] An embodiment of the fourth aspect of the present application provides a non-transitory computer-readable storage medium storing computer instructions, where the computer instructions are used to cause the computer to execute the automatic parking method of the vehicle disclosed in the embodiments of the present application.

[0020] In the embodiments of the present disclosure, first, an initial position corresponding to a target parking space to which the vehicle is to be parked is determined, then the vehicle is controlled to travel to the initial position, then a first image of the environment around the vehicle when the vehicle is at the initial position is acquired, then the first image is processed to obtain a parking label corresponding to the vehicle, and then the vehicle is controlled to park in the target parking space according to a parking instruction associated with the parking label. Thus, by means of the parking instruction corresponding to the initial position of the vehicle, the vehicle is controlled to park, making the parking control of the vehicle more precise and accurate, improving the alignment accuracy between the vehicle and the target parking space, further improving the safety of parking in the target parking space, reducing the risk of collision between the vehicle and the mechanical parking space or scratching the wheels, and solving the problem of difficult parking in mechanical parking spaces. In addition, in this embodiment, the requirements for computing power and sensors of the vehicle are very low, and it can be adapted to a variety of new energy vehicle models.

[0021] Additional aspects and advantages of the present disclosure will be given in part in the following description, become apparent in part from the following description, or be learned through the practice of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The above and / or additional aspects and advantages of the present disclosure will become apparent and be readily understood from the following description of the embodiments in conjunction with the drawings, where:

[0023] Figure 1 is a schematic flowchart of an automatic parking method of a vehicle according to a first embodiment of the present disclosure;

[0024] Figure 2It is a schematic diagram of the physical scene of the mechanical garage space provided according to an embodiment of the present disclosure;

[0025] Figure 3 It is a schematic flowchart of the automatic parking method for a vehicle provided according to the second embodiment of the present disclosure;

[0026] Figure 4 It shows a schematic diagram of the relative position between a vehicle and a target parking space;

[0027] Figure 5 It shows a schematic diagram of the relative position with a predicted collision category of a left collision;

[0028] Figure 6 It is a schematic flowchart of the automatic parking method for a vehicle provided according to the third embodiment of the present disclosure;

[0029] Figure 7 It is a schematic diagram of the automatic parking device for a vehicle according to an embodiment of the present disclosure;

[0030] Figure 8 It is a block diagram of an electronic device for implementing the automatic parking method for a vehicle according to an embodiment of the present disclosure. Detailed Description of the Embodiment

[0031] The embodiments of the present disclosure will be described in detail below. Examples of the embodiments are shown in the accompanying drawings, where the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described by referring to the accompanying drawings are exemplary only for explaining the present disclosure and should not be construed as a limitation of the present disclosure. On the contrary, the embodiments of the present disclosure include all changes, modifications, and equivalents falling within the spirit and scope of the appended claims.

[0032] As a possible application scenario, when a vehicle parks into a mechanical parking space, a human driver usually aligns the vehicle in front of the parking space and then only controls the accelerator and brake to reverse the vehicle into the parking space. In actual operation, alignment (axial alignment) is a technical difficulty in parking into a mechanical parking space. A human driver generally parks the vehicle in front of the parking space according to a rough target and then adjusts the vehicle pose by observing the relative position between the vehicle and the parking space in the rearview mirror to achieve precise alignment with the parking space. The automatic parking method for a vehicle proposed in the embodiments of the present disclosure can imitate this process of a human driver, dividing the parking into two stages: an initial alignment stage between the vehicle and the parking space and a fine alignment stage, to solve the problem of difficult parking of a vehicle into a mechanical parking space.

[0033] Among them, in the initial alignment stage, the vehicle can be first parked at an initial position a certain distance in front of the mechanical parking space. At this time, the vehicle and the mechanical parking space are in a basically aligned state, that is, the central axis of the vehicle and the central axis of the mechanical parking space can be nearly aligned. In the fine alignment stage, the vehicle pose can be adjusted to achieve precise alignment with the parking space, and then the vehicle is parked in the parking space.

[0034] It should be noted that the execution subject of the automatic parking method of the vehicle in this embodiment can be the automatic parking device of the vehicle. This device can be implemented in software and / or hardware, and this device can be configured in the vehicle.

[0035] In the embodiments of the present disclosure, the "vehicle-mounted system" will be used as the execution subject to execute the automatic parking method of the vehicle for illustration, and it is not limited here.

[0036] Figure 1 is a schematic flowchart of the automatic parking method of the vehicle provided in the first embodiment of the present disclosure. As Figure 1 shown, the method includes:

[0037] S101: Determine the initial position corresponding to the target parking space where the vehicle is to be parked.

[0038] Among them, the target parking space can be the parking space where the vehicle is to be parked.

[0039] Optionally, the target parking space can be the parking space selected by the user, or it can also be the parking space automatically selected by the vehicle through calculation, which is not limited here.

[0040] As a possible implementation method, in the parking space selection stage, the user can select a parking space as the target parking space on the vehicle's central control screen according to their own preferences. Or, the target parking space can also be autonomously selected by the vehicle through the path planning module. For example, a parking space without parked vehicles, no obstacles nearby, and relatively close to the current position of the vehicle can be used as the target parking space, which is not limited here.

[0041] In the embodiments of the present disclosure, the type of the target parking space can be a mechanical garage parking space, or it can also be a parking space of other types of garages, which is not limited here.

[0042] Among them, the mechanical garage parking space is a special type of garage parking space, and the access and storage of vehicles are realized through mechanical devices. The mechanical garage parking space usually consists of one or more vertically or horizontally moving platforms, and the vehicle can be lifted, moved and placed in a specified position. The parking space of the mechanical garage is usually strictly limited to a deep groove. If the vehicle cannot be parked strictly in the center, it may cause the vehicle to collide with the entrance or scrape the wheel hub.

[0043] As Figure 2 shown,Figure 2 It is a schematic diagram of the physical scene of a mechanical garage parking space.

[0044] Furthermore, after the vehicle obtains the target parking space, it can calculate the initial position associated with the target parking space, and then use this initial position as the initial parking target to plan a reasonable path through the path planning module.

[0045] Among them, the initial position can be the position where the vehicle needs to park before entering the target parking space.

[0046] It can be understood that if the vehicle is directly parked in the target parking space, there may be situations such as parking space collision or wheel rubbing. Therefore, an initial position relatively close to the target parking space can be determined first, so that the vehicle can first park at this initial position, and then further determine whether the vehicle can safely park in the target parking space.

[0047] As a possible implementation method, an initial position can be selected directly in front of the target parking space, so that when the vehicle is at this initial position, it is relatively close to the target parking space and maintains a certain distance from the target parking space. There is no limitation here.

[0048] S102: Control the vehicle to drive to the initial position.

[0049] Optionally, a vehicle control system, such as an automatic driving system or a remote control system, can be used to control the vehicle to drive automatically. Specifically, sensors installed on the vehicle, such as cameras, radars, lidars, etc., can be used to monitor the surrounding environment in real time to avoid collisions and ensure safe driving. The vehicle automatically drives to the set initial position according to the navigation instructions or the instructions of the control system. During the driving process, the vehicle may need to comply with traffic rules and road signs.

[0050] It should be noted that controlling the vehicle to drive automatically to the initial position requires corresponding automatic driving technologies and equipment, and the safety of the driving process must be ensured. In practical applications, traffic regulations and safety regulations also need to be complied with, and sufficient testing and verification need to be carried out to ensure the reliability and accuracy of the system.

[0051] As a possible implementation method, based on the Automatic Parking Assist (APA) system, the vehicle can be controlled to drive from any other position to the initial position.

[0052] It should be noted that the automatic parking assist system is a modern vehicle safety and driving assistance technology designed to help drivers park more easily and accurately when parking. The APA uses on-vehicle sensors (such as ultrasonic or cameras) to detect the surrounding environment and obstacles, and uses specific algorithms and control systems to automatically control the steering, acceleration, and braking of the vehicle to achieve the automatic parking process. The automatic parking assist system can help drivers achieve more precise and efficient parking operations in scenarios such as parallel parking, perpendicular parking, or reverse parking into a garage.

[0053] It can be understood that driving the vehicle to the initial position can be a process of initial alignment. When the vehicle is in the initial position, it is directly in front of the vehicle. However, when the vehicle is parked from the initial position into the target parking space, there is still a risk of collision with the parking space or scratching the wheels. Therefore, in the embodiments of the present disclosure, the vehicle's ability to safely park in the target parking space is further analyzed based on the state of the vehicle at the initial position.

[0054] S103: Obtain a first image of the surrounding environment of the vehicle when the vehicle is in the initial position.

[0055] Among them, the first image can be a single image, or it can also be multiple images, or it can also be a fused image of multiple images, which is not limited here.

[0056] Among them, the first image is used to observe the relative positions of the target parking space and the vehicle.

[0057] It should be noted that multiple camera devices can be pre-set on the vehicle, such as a fish-eye camera. For example, a fish-eye camera can be set respectively at the front (front of the vehicle), rear (rear of the vehicle), left, and right of the vehicle body, so as to obtain the front view image, rear view image, left view image, and right view image of the vehicle correspondingly at the initial position.

[0058] As a possible implementation method, if the vehicle reverses (along the rear direction) into the target parking space, the rear view image can be used as the first image. Or, the rear view image, the left view image, and the right view image can also be used together as the first image. Or, the fused image of the rear view image, the left view image, and the right view image can also be used as the first image, which is not limited here.

[0059] As another possible implementation method, if the vehicle drives forward in the direction of the vehicle head into the target parking space, the front view image can be used as the first image. Or, the front view image, the left view image, and the right view image can also be used together as the first image. Or, the fused image of the front view image, the left view image, and the right view image can also be used as the first image, which is not limited here.

[0060] Optionally, the first image may further include a depth image collected by a depth camera and / or a lidar point cloud image captured by a lidar camera. Alternatively, the first image may further include an image collected by a normal camera and / or an image collected by a high-dynamic range camera, which is not limited herein. Thus, more accurate and reliable perception of the target parking space can be achieved, and the accuracy of parking into the garage can be improved.

[0061] In the embodiments of the present disclosure, any one or more frames of images that can reflect the relative positions of the target parking space and the vehicle may be used as the first image, and the acquisition method of the first image is not limited herein.

[0062] S104: Process the first image to obtain a parking label corresponding to the vehicle.

[0063] Among them, the parking label is used to represent the parking method in the target parking space. It should be noted that since the parking label is determined based on the first image, and the first image is an image of the surrounding environment when the vehicle is in the initial position, and the first image contains information about the target parking space. Therefore, the parking label is generated for the relative position relationship between the target parking space and the vehicle.

[0064] In the embodiments of the present disclosure, the parking labels may be classified into different categories to associate different indication information. For example, the parking labels may be classified into a first category of parking labels, a second category of parking labels, and a third category of parking labels, which is not limited herein.

[0065] As a possible implementation, the first image may be input into a pre-trained neural network model to obtain a parking label corresponding to the vehicle. Among them, the neural network model may be a convolutional neural network, or other network structures may also be selected, which is not limited herein.

[0066] Specifically, the target image features corresponding to the first image may be obtained by first performing feature extraction on the first image. Then, the in-vehicle system may make a decision based on the target image features to predict whether the vehicle will collide with the target parking space when driving into the target parking space, and then generate a corresponding processing strategy, that is, the parking label. For example, if it is predicted that the vehicle will not collide with the target parking space when driving into the target parking space, a first category of parking labels may be generated; if it is predicted that the vehicle may collide with both sides of the target parking space when driving into the target parking space, a second category of parking labels may be generated; if it is predicted that the vehicle may collide with one side of the target parking space when driving into the target parking space, a third category of parking labels may be generated, which is not limited herein.

[0067] As another possible implementation, in the embodiments of the present disclosure, an end-to-end decision-making and planning network that has been pre-trained can be used to generate a parking label corresponding to the vehicle's initial position based on the target image features, which will not be limited herein.

[0068] S105: Control the vehicle to park in the target parking space according to the parking instruction associated with the parking label.

[0069] Among them, the parking instruction can be a strategy for reminding the vehicle how to park. The parking instructions associated with different categories of parking labels may be different. For example, some parking instructions are to terminate parking, some are to directly park in the target parking space, and some are to control the vehicle to drive into the target parking space after aligning with the target parking space, etc., which will not be limited herein.

[0070] Optionally, if the parking label belongs to the first category label, it is determined that the parking instruction is to directly park in the target parking space, and the in-vehicle system can then control the vehicle to park in the target parking space.

[0071] Among them, the first category label indicates that the vehicle parks in the target parking space without collision.

[0072] Specifically, if the parking label belongs to the first category label, it means that the size of the vehicle is suitable for the target parking space, and the vehicle can be parked in the target parking space normally without colliding with the target parking space. The in-vehicle system can directly park the vehicle in the target parking space according to the parking instruction associated with the first category label. After parking in the target parking space, the in-vehicle system can output a signal to terminate parking and give a prompt to the user.

[0073] Or, if the parking label belongs to the second category label, it is determined that the parking instruction is to terminate parking.

[0074] Among them, the second category label indicates that the target parking space is not available for vehicle parking.

[0075] Specifically, if the parking label belongs to the second category label, it means that the size of the vehicle is small and cannot be parked in the target parking space, otherwise it will collide with both sides of the target parking space. At this time, the in-vehicle system can terminate parking according to the parking instruction, output a signal to terminate parking, and give a prompt to the user.

[0076] In the embodiments of the present disclosure, first, an initial position corresponding to the target parking space where the vehicle is to be parked is determined. Then, the vehicle is controlled to drive to the initial position. Next, a first image of the vehicle's surrounding environment when the vehicle is at the initial position is obtained. After that, the first image is processed to obtain a parking label corresponding to the vehicle. Then, according to the parking instruction associated with the parking label, the vehicle is controlled to park in the target parking space. Thus, by means of the parking instruction corresponding to the initial position of the vehicle, the vehicle is controlled to park, making the parking control of the vehicle more precise and accurate, improving the accuracy of aligning the vehicle with the target parking space, and further being able to improve the safety of parking in the target parking space, reducing the risk of the vehicle colliding with or scraping the wheels of the mechanical parking space, and solving the problem of difficult parking in mechanical parking spaces. In addition, in this embodiment, the requirements for computing power and sensors of the vehicle are very low, and it can be adapted to a variety of new energy vehicle models.

[0077] Figure 3 It is a schematic flowchart of the automatic parking method of the vehicle according to the second embodiment of the present disclosure. As Figure 3 shown, the method includes:

[0078] S201: Determine the parking space position of the target parking space.

[0079] Among them, the parking space position may be the coordinate position of the target parking space in the world coordinate system.

[0080] Specifically, the vehicle-mounted system can detect the target parking space to locate the coordinate position of the target parking space in the world coordinate system. For example, positioning technologies such as the Global Positioning System (GPS), Wireless Sensor Network (WSN), or indoor positioning systems (such as Bluetooth, Wi-Fi, or ultrasonic positioning systems) can be used to obtain the accurate position information of the target parking space.

[0081] S202: Determine the distance interval associated with the attribute information of the vehicle.

[0082] Among them, the attribute information of the vehicle may be the vehicle model, the size of the vehicle (such as length, width, and height), etc., which are not limited herein.

[0083] Among them, the vehicle models may include sedans, small cars, micro cars, compact cars, medium-sized cars, mid-large cars, large cars, SUVs (Sports Utility Vehicles), MPVs (Multi-Purpose Vehicles), sports cars, etc., which are not limited herein.

[0084] Among them, the distance interval may be the numerical interval corresponding to the distance between the target parking space and the vehicle.

[0085] It should be noted that the attribute information of different vehicles may be the same or different. For example, for sedans and SUVs, the vehicle sizes are different. Therefore, some parking spaces can be used to park sedans but not SUVs. In the embodiments of the present disclosure, a mapping relationship table between the attribute information of the vehicle and the distance interval can be established in advance, so that the in-vehicle system can directly obtain the distance interval to which the current vehicle belongs by reading the stored mapping relationship table. Alternatively, the corresponding distance interval can also be directly configured in the vehicle.

[0086] For example, if the vehicle model is a sedan, the corresponding distance interval can be set to 45 cm - 55 cm in advance, which is not limited here. It should be noted that the distance intervals corresponding to sedans of different sizes may also be different.

[0087] S203: Determine the initial position corresponding to the position of the parking space based on the distance interval.

[0088] Optionally, the edge line of the target parking space facing the vehicle and the vehicle side line facing the target parking space can be determined first, and then the included angle between the central axis of the target parking space and the central axis of the vehicle can be determined.

[0089] As Figure 4 shown, Figure 4 shows a schematic diagram of the relative positions of a vehicle and a target parking space. Among parking spaces 1, 2, and 3, the vehicle is directly in front of parking space 2, and parking space 2 is the target parking space for the vehicle. Among them, side A is the front side corresponding to parking space 2 and also the outer side corresponding to parking space 2. The vehicle can drive into the target parking space from side A.

[0090] Among them, the central axis of the target parking space is k1, and the central axis of the vehicle is k2. Among them, the acute included angle between k1 and k2 can be used as the included angle between the central axis of the target parking space and the central axis of the vehicle. Among them, A is the outer side among the four sides of the target parking space, that is, the side of the target parking space facing the vehicle, and the edge line corresponding to side A can be used as the edge line of the target parking space facing the vehicle. Figure 4 In [the figure], the rear of the vehicle faces the target parking space, so the edge line on the rear side can be used as the vehicle side line facing the target parking space.

[0091] Furthermore, if the vehicle is at any position directly in front of the target parking space and the distance between the vehicle side line and the parking space edge line is within the distance interval and the included angle is less than the preset threshold, then any position is determined as the initial position corresponding to the position of the parking space.

[0092] Among them, the preset threshold can be the threshold of the included angle. It should be noted that if the included angle is less than the preset threshold, it means that the central axis of the vehicle and the central axis of the target storage location are basically aligned, otherwise it means that there is a large deviation between the central axis of the vehicle and the central axis of the target storage location.

[0093] As Figure 4 shown, h is the distance between the vehicle side line and the storage location edge line. For example, if the vehicle is a sedan and the corresponding distance range is 45 cm - 55 cm, and the preset threshold is 45 degrees. If the vehicle is at position P, h is equal to 50 cm, and the included angle between k1 and k2 is 30 degrees, it means that the included angle of 30 degrees is less than the preset threshold of 45 degrees, and the distance of 50 cm between the vehicle side line and the storage location edge line is within the distance range of 45 cm - 55 cm. Therefore, position P can be used as the initial position corresponding to the storage location, and no limitation is made here.

[0094] S204: Control the vehicle to drive to the initial position.

[0095] S205: Obtain the first image of the vehicle's surrounding environment when the vehicle is at the initial position.

[0096] It should be noted that for the specific implementation methods of steps S204 - S205, reference can be made to the above embodiments and will not be elaborated here.

[0097] S206: Extract features from the first image to obtain the target image features.

[0098] Among them, the target image features can be the image features corresponding to the first image. Among them, the target image features can be embodied in the form of feature vectors, and no limitation is made here.

[0099] Optionally, the first image can be subjected to feature extraction through a feature extraction algorithm, so that the target image features corresponding to the first image can be obtained. Or, the first image can also be input into a backbone network to obtain the target image features corresponding to the first image.

[0100] S207: Input the target image features into a pre-constructed position adjustment decision network to obtain the predicted collision category and the distance adjustment label.

[0101] Among them, the predicted collision category is the predicted result output by the position adjustment decision network on whether the vehicle will collide with the target storage location. For example, the predicted collision category can be "no collision", "both-side collision", "one-side collision", and no limitation is made here.

[0102] It should be noted that "no collision" means that when the vehicle parks into the target parking space, it will not collide with the edge of the target parking space. "Both-side collision" means that when the vehicle parks into the target parking space, the vehicle will collide with both sides of the target parking space. "One-side collision" means that when the vehicle parks into the target parking space, either the left side or the right side of the vehicle will collide with the target parking space.

[0103] Furthermore, "one-side collision" can be further divided into "left-side collision" and "right-side collision", which will not be limited herein.

[0104] Among them, "left-side collision" means that when the vehicle parks into the target parking space, the left side of the vehicle will collide with the target parking space. "Right-side collision" means that when the vehicle parks into the target parking space, the right side of the vehicle will collide with the target parking space.

[0105] As Figure 4 shown, Figure 4 if the left edge line k3 of the target parking space intersects with the left side of the vehicle in the figure, it means that when the vehicle directly parks into the target parking space from the initial position, the left side of the vehicle will collide with the target parking space, and the corresponding predicted collision category is "left-side collision" in "one-side collision".

[0106] Among them, the distance adjustment label can be the recommended offset distance. For example, if the predicted collision category is "left-side collision" and the distance adjustment label is 5 cm, it means that if the vehicle offsets 5 cm to the right, it may not collide with the target parking space, which will not be limited herein.

[0107] It can be understood that if the predicted collision category is "left-side collision", it can be first determined that the left edge line of the target parking space passes through the left side of the vehicle, that is, intersects with the left area of the vehicle on the plane. Furthermore, the distance between the left edge line and the leftmost point of the vehicle can be used as the first distance. That is to say, the offset distance of the vehicle to the right needs to be at least greater than this first distance so that it will not collide with the left side of the target parking space.

[0108] As Figure 5 shown, Figure 5 it is a schematic diagram of the relative position with the predicted collision category of left-side collision. Among them, m2 is the straight line where the left edge line of the target parking space is located. Among them, m2 intersects with the left side of the vehicle. The leftmost point of the vehicle is point e, and the angle corresponding to point e is the angle in front of the left of the vehicle. Among them, the distance from point e to m2 is the distance f between the straight line m1 where point e is located and m2, and the straight lines m1 and m2 are parallel. Therefore, the offset distance of the vehicle to the right needs to be at least greater than this distance f so that the left side of the vehicle will not collide with the left side of the target parking space when the vehicle parks into the target parking space. Therefore, the offset distance corresponding to the distance adjustment label can be set to a value slightly greater than the distance f.

[0109] Among them, the position adjustment decision network is a pre-trained decision network that can reach an available state. It can be used to predict the best offset distance as the distance adjustment label and the collision situation between the vehicle and the target storage location under given input conditions.

[0110] Specifically, the position adjustment decision network can determine the state information of the vehicle's current position, the position information of the target storage location, the environmental conditions, etc. through the target image features, which are not limited here. By learning historical data and training samples, the position adjustment decision network can learn to infer the best offset distance (distance adjustment label) that the vehicle should take at the initial position and the predicted collision category between the vehicle and the target storage location.

[0111] Optionally, the position adjustment decision network can consist of multiple layers, including an input layer, a hidden layer, and an output layer. The input layer receives the target image feature information and passes it to the hidden layer for processing. The hidden layer uses activation functions and weight adjustments to calculate and generate intermediate representations, and then passes them to the output layer. The training of the position adjustment decision network can use supervised learning methods to optimize the parameters and adjust the network weights using known inputs and corresponding correct outputs. Through the backpropagation algorithm, the network can gradually optimize its decision-making ability to make more accurate position adjustment decisions in similar situations.

[0112] S208: Input the target image features into a pre-constructed angle adjustment decision network to obtain an angle adjustment label.

[0113] Among them, the angle adjustment label can be the recommended best angle adjustment amount and the adjusted heading. For example, adjust the heading to the left and adjust 0.4 degrees to the left, which is not limited here.

[0114] Among them, the angle adjustment decision network is a pre-trained decision network that can reach an available state. It can be used to predict the best angle adjustment amount and the adjusted heading as the angle adjustment label under given input conditions.

[0115] Specifically, the angle adjustment decision network can determine the state information of the vehicle's current position, the position information of the target storage location, the environmental conditions, etc. through the target image features, which are not limited here. By learning historical data and training samples, the angle adjustment decision network can learn to infer the best angle adjustment amount and the adjusted heading that the vehicle should take at the initial position.

[0116] Optionally, the angle adjustment decision network can consist of multiple layers, including an input layer, a hidden layer, and an output layer. The input layer receives the target image feature information and passes it to the hidden layer for processing. The hidden layer uses activation functions and weight adjustments to calculate and generate intermediate representations, which are then passed to the output layer. The training of the angle adjustment decision network can use supervised learning methods, using known inputs and corresponding correct outputs to optimize parameters and adjust network weights. Through the backpropagation algorithm, the network can gradually optimize its decision-making ability to make more accurate angle adjustment decisions in similar situations.

[0117] As Figure 5 shown, Figure 5 the r in it is the angle between the left edge line of the target parking space and the left edge line of the vehicle. The angle adjustment amount included in the angle adjustment label can be r, and the adjusted heading can be to adjust the heading to the right, which is not limited here.

[0118] Among them, the parking label includes a predicted collision category, a distance adjustment label, and an angle adjustment label. That is to say, the predicted collision category, the distance adjustment label, and the angle adjustment label can be jointly used as the label content in the parking label.

[0119] Optionally, when the predicted collision category is no collision, the parking label is the first category label.

[0120] Optionally, when the predicted collision category is two-sided collision, the parking label is the second category label.

[0121] Optionally, when the predicted collision category is one-sided collision, the parking label is the third category label.

[0122] S209: Control the vehicle to park in the target parking space according to the parking instruction associated with the parking label.

[0123] It should be noted that for the specific implementation method of step S209, reference can be made to the above embodiments, which will not be elaborated here.

[0124] In the embodiments of the present disclosure, first, the location of the target storage location is determined. Then, a distance range associated with the attribute information of the vehicle is determined. Next, based on the distance range, an initial location corresponding to the storage location is determined. After that, the vehicle is controlled to travel to the initial location. Then, a first image of the vehicle's surrounding environment when the vehicle is at the initial location is obtained, and feature extraction is performed on the first image to obtain target image features. Next, the target image features are input into a pre-constructed position adjustment decision network to obtain a predicted collision category and a distance adjustment label. Then, the target image features are input into a pre-constructed angle adjustment decision network to obtain an angle adjustment label. Finally, based on the parking instruction associated with the parking label, the vehicle is controlled to park in the target storage location. Thus, according to the attribute information of the vehicle, the best position for the in-vehicle system to observe the relative position relationship between the vehicle and the target storage location can be selected, which can ensure the reliability of the target image features. Then, based on the position adjustment decision network and the angle adjustment decision network, the distance adjustment label, the angle adjustment label, and the predicted collision category of the vehicle can be accurately and reliably obtained, improving the safety and reliability of parking in the garage.

[0125] Figure 6 is a schematic flowchart of an automatic parking method for a vehicle according to the third embodiment of the present disclosure. As Figure 6 shown, the method includes:

[0126] S301: Determine an initial location corresponding to the target storage location where the vehicle is to be parked.

[0127] S302: Control the vehicle to travel to the initial location.

[0128] S303: Obtain a first image of the vehicle's surrounding environment when the vehicle is at the initial location.

[0129] S304: Process the first image to obtain a parking label corresponding to the vehicle.

[0130] It should be noted that for the specific implementation methods of steps S301 - S304, reference can be made to the above embodiments and will not be elaborated here.

[0131] S305: If the parking label belongs to the third category label, determine the parking instruction as: Control the vehicle to drive into the target storage location after aligning with the target storage location.

[0132] It should be noted that if the parking label belongs to the third category label, it can be determined that the vehicle may collide with one side of the target storage location according to the predicted collision category in the parking label. Therefore, in order to prevent the vehicle from colliding with one side of the target storage location, the pose of the vehicle can be adjusted so that the vehicle is aligned with the target storage location, and then the vehicle is controlled to drive into the target storage location.

[0133] S306: Based on the movement adjustment tags included in the parking tag, control the vehicle to move to adjust the relative position between the vehicle and the target parking space.

[0134] Among them, the movement adjustment tags at least include a distance adjustment tag and an angle adjustment tag, and the movement adjustment tags reflect the pose adjustment method of the vehicle.

[0135] It should be noted that the angle adjustment tag can be a numerical value for angle adjustment. For example, 0 means no adjustment, a positive number means that the heading needs to be adjusted to the left, and a negative number means that the heading needs to be adjusted to the right. The distance adjustment tag can be a numerical value for distance adjustment. For example, 0 means no adjustment, a positive number means that it needs to shift to the left, and a negative number means that it needs to shift to the right.

[0136] For example, if the rear of the vehicle faces the target parking space, the distance adjustment tag is +4 cm, and the angle adjustment tag is +0.4 degrees, the in-vehicle system can plan a trajectory of driving forward and then reversing. Specifically, the first section of the route is to turn the wheel forward to the left front, and the second section of the route is the corresponding reversing trajectory, so that when the vehicle reverses to the corresponding position in front of the target parking space, the heading is adjusted 0.4 degrees to the left and shifted 4 cm to the left. No limitation is made here.

[0137] It should be noted that the above example is only an illustrative description and does not limit the present disclosure.

[0138] S307: In response to determining that the vehicle is aligned with the target parking space, control the vehicle to drive into the target parking space.

[0139] Optionally, in response to determining that the vehicle moves to the first position corresponding to the target parking space, the in-vehicle system can first obtain a second image of the vehicle's surrounding environment when the vehicle is at the first position, and then process the second image to obtain the target parking tag corresponding to the vehicle.

[0140] Among them, the first position can be the position where the vehicle moves after being adjusted according to the parking indication associated with the third category of tags.

[0141] Among them, the conditions that the first position and the initial position need to meet are the same. Referring to the steps in the above embodiments, if when the vehicle moves to the first position, the included angle between the central axis of the vehicle and the central axis of the target parking space is less than the preset threshold, and the distance between the vehicle side line and the edge line of the parking space is within the distance interval, it means that the first position is the first position corresponding to the target parking space.

[0142] Among them, the second image can be an image of the vehicle's surrounding environment collected by the vehicle when the vehicle is at the first position, and the second image is used to observe the relative position between the target parking space and the vehicle when the vehicle is at the first position.

[0143] Among them, the target parking label may be a parking label determined based on the second image.

[0144] It should be noted that the method for obtaining the second image and the method for processing the second image to obtain the target parking label may both refer to the above embodiments and will not be elaborated here.

[0145] Specifically, if the target parking label is the first category label, it indicates that the vehicle and the target parking space are already in position alignment.

[0146] Specifically, if the target parking label is the second category label, the vehicle parking needs to be terminated, indicating that the vehicle cannot drive into the target parking space.

[0147] Specifically, if the target parking label is the third category label, based on the target parking label, adjust the relative position of the vehicle and the target parking space until the vehicle and the target parking space are aligned, and then control the vehicle to park in the target parking space.

[0148] It should be noted that if the target parking label is the third category label, the relative position of the vehicle and the target parking space needs to be adjusted based on the distance adjustment label and the angle adjustment label included in the target parking label so that the vehicle and the target parking space are aligned. If the vehicle and the target parking space are aligned, the vehicle can be directly controlled to park in the target parking space. If the vehicle cannot be aligned with the target parking space and the target parking space is too small, control the vehicle to terminate parking. If after adjusting the relative position of the vehicle and the target parking space, the parking label corresponding to the vehicle is still the third category label, the above steps need to be repeated until the finally obtained parking label is the first category label to control the vehicle to park in the target parking space, or until the finally obtained parking label is the second category label to control the vehicle to terminate parking.

[0149] In the embodiments of the present disclosure, first determine the initial position corresponding to the target parking space where the vehicle is to be parked, then control the vehicle to drive to the initial position, then obtain the first image of the vehicle's surrounding environment when the vehicle is at the initial position, then process the first image to obtain the parking label corresponding to the vehicle, and then if the parking label belongs to the third category label, determine the parking instruction as: control the vehicle to drive into the target parking space after aligning with the target parking space, and then based on the movement adjustment label included in the parking label, control the vehicle to move to adjust the relative position of the vehicle and the target parking space, and finally in response to determining that the vehicle and the target parking space are in position alignment, control the vehicle to drive into the target parking space. Thus, in the case where the vehicle may collide with one side of the target parking space, the vehicle can be controlled to move based on the movement adjustment label included in the parking label so that the vehicle can be aligned with the target parking space, thereby avoiding collisions between the vehicle and the target parking space, achieving end-to-end adjustment, and parking the vehicle comfortably and accurately into the target parking space, solving the problem of difficult mechanical parking.

[0150] Figure 7 It is a schematic diagram of an automatic parking device for a vehicle according to another embodiment of the present disclosure. As Figure 7 shown, the automatic parking device 700 of the vehicle includes:

[0151] A first determination module 710, configured to determine an initial position corresponding to a target parking space where the vehicle is to be parked;

[0152] A first control module 720, configured to control the vehicle to travel to the initial position;

[0153] An acquisition module 730, configured to acquire a first image of the environment around the vehicle when the vehicle is at the initial position;

[0154] A processing module 740, configured to process the first image to obtain a parking label corresponding to the vehicle;

[0155] A second control module 750, configured to control the vehicle to park in the target parking space according to a parking instruction associated with the parking label.

[0156] Optionally, the second control module 750 is specifically configured to:

[0157] If the parking label belongs to a first category label, determine that the parking instruction is to directly park in the target parking space, and control the vehicle to park in the target parking space, where the first category label indicates that the vehicle parks in the target parking space without collision;

[0158] Or,

[0159] If the parking label belongs to a second category label, determine that the parking instruction is to terminate parking, where the second category label indicates that the target parking space is not available for parking the vehicle.

[0160] Optionally, if the parking label belongs to a third category label, the second control module 750 includes:

[0161] A first determination unit, configured to determine that the parking instruction is to control the vehicle to drive into the target parking space after aligning with the target parking space;

[0162] A first control unit, configured to control the vehicle to move based on a movement adjustment label included in the parking label to adjust the relative position between the vehicle and the target parking space;

[0163] A second control unit, configured to control the vehicle to drive into the target parking space in response to determining that the vehicle is aligned with the target parking space,

[0164] where the movement adjustment label at least includes a distance adjustment label and an angle adjustment label.

[0165] Optionally, the second control unit includes:

[0166] A first acquisition subunit, configured to, in response to determining that the vehicle moves to a first position corresponding to the target parking space, acquire a second image of the environment around the vehicle at the first position;

[0167] A second acquisition subunit, configured to process the second image to obtain a target parking label corresponding to the vehicle;

[0168] A determination subunit, configured to determine that the vehicle is aligned with the target parking space if the target parking label is the first category label.

[0169] Optionally, the second acquisition subunit is further configured to:

[0170] If the target parking label is the third category label, based on the target parking label, adjust the relative position between the vehicle and the target parking space until the vehicle is aligned with the target parking space;

[0171] Control the vehicle to park in the target parking space.

[0172] Optionally, the first determination module 710 includes:

[0173] A second determination unit, configured to determine the parking space position of the target parking space;

[0174] A third determination unit, configured to determine a distance interval associated with the attribute information of the vehicle;

[0175] A fourth determination unit, configured to determine an initial position corresponding to the parking space position based on the distance interval.

[0176] Optionally, the fourth determination unit is specifically configured to:

[0177] Determine the edge line of the parking space facing the vehicle of the target parking space, and the vehicle side line facing the target parking space;

[0178] Determine the included angle between the central axis of the target parking space and the central axis of the vehicle;

[0179] If the vehicle is at any position directly in front of the target parking space, the distance between the vehicle side line and the edge line of the parking space is within the distance interval, and the included angle is less than a preset threshold, then determine the any position as the initial position corresponding to the parking space position.

[0180] Optionally, the processing module 740 is specifically configured to:

[0181] Extract features from the first image to obtain target image features;

[0182] Input the target image features into a pre-constructed position adjustment decision network to obtain a predicted collision category and a distance adjustment label;

[0183] Input the target image features into a pre-constructed angle adjustment decision network to obtain an angle adjustment label;

[0184] Wherein, the parking label includes the predicted collision category, the distance adjustment label, and the angle adjustment label;

[0185] When the predicted collision category is no collision, the parking label is a first category label;

[0186] When the predicted collision category is two-sided collision, the parking label is a second category label;

[0187] When the predicted collision category is one-sided collision, the parking label is a third category label.

[0188] In the embodiments of the present disclosure, first, an initial position corresponding to the target parking space to be parked by the vehicle is determined, then the vehicle is controlled to drive to the initial position, then a first image of the vehicle surrounding environment when the vehicle is at the initial position is obtained, and then the first image is processed to obtain a parking label corresponding to the vehicle. Then, according to the parking instruction associated with the parking label, the vehicle is controlled to park in the target parking space. Thus, by the parking instruction corresponding to the initial position of the vehicle, the vehicle is controlled to park, making the parking control of the vehicle more precise and accurate, improving the alignment accuracy between the vehicle and the target parking space, and further improving the safety of parking in the target parking space, reducing the risk of collision or rubbing of the vehicle wheels with the mechanical parking space, and solving the problem of difficult parking in the mechanical parking space. In addition, in this embodiment, the requirements for computing power and sensors of the vehicle are very low, and it can be adapted to a variety of new energy vehicle models.

[0189] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0190] Figure 8 The block diagram of an exemplary computer device suitable for implementing the embodiments of the present application is shown. Figure 8 The displayed computer device 12 is only an example and should not bring any limitation to the functions and usage scope of the embodiments of the present application.

[0191] Such as Figure 8As shown, the computer device 12 is presented in the form of a general-purpose computing device. The components of the computer device 12 may include, but are not limited to: one or more processors or processing units 16, a system memory 28, and a bus 18 that connects different system components (including the system memory 28 and the processing unit 16).

[0192] The bus 18 represents one or more of several types of bus architectures, including a memory bus or memory controller, a peripheral bus, a graphics acceleration port, a processor, or a local bus using any of the various bus architectures. By way of example, these architectures include, but are not limited to, Industry Standard Architecture (ISA) bus, Micro Channel Architecture (MAC) bus, Enhanced ISA bus, Video Electronics Standards Association (VESA) local bus, and Peripheral Component Interconnection (PCI) bus.

[0193] The computer device 12 typically includes a variety of computer system-readable media. These media can be any available media that can be accessed by the computer device 12, including volatile and non-volatile media, removable and non-removable media.

[0194] The memory 28 may include computer system-readable media in the form of volatile memory, such as random access memory (RAM) 30 and / or cache memory 32. The computer device 12 may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, a storage system 34 can be used for reading and writing on non-removable, non-volatile magnetic media ( Figure 8 not shown, commonly referred to as a "hard disk drive").

[0195] Although Figure 8Not shown in the figure, a disk drive for reading and writing a removable non-volatile disk (such as a "floppy disk") and an optical disk drive for reading and writing a removable non-volatile optical disk (such as a Compact Disc Read Only Memory (hereinafter referred to as: CD-ROM), a Digital Video Disc Read Only Memory (hereinafter referred to as: DVD-ROM) or other optical media) can be provided. In these cases, each drive can be connected to the bus 18 through one or more data medium interfaces. The memory 28 may include at least one program product having a set (such as at least one) of program modules configured to perform the functions of the embodiments of the present application.

[0196] A program / utility 40 having a set (at least one) of program modules 42 can be stored, for example, in the memory 28. Such program modules 42 include, but are not limited to, an operating system, one or more application programs, other program modules, and program data. The implementation of a network environment may be included in each or some combination of these examples. The program modules 42 generally perform the functions and / or methods in the embodiments described in the present application.

[0197] The computer device 12 can also communicate with one or more external devices 14 (such as a keyboard, a pointing device, a display 24, etc.), and can also communicate with one or more devices that enable a user to interact with the computer device 12, and / or communicate with any device that enables the computer device 12 to communicate with one or more other computing devices (such as a network card, a modem, etc.). Such communication can be carried out through the input / output (I / O) interface 22. In addition, the computer device 12 can also communicate with one or more networks (such as a Local Area Network (hereinafter referred to as: LAN), a Wide Area Network (hereinafter referred to as: WAN) and / or a public network, such as the Internet) through the network adapter 20. As shown in the figure, the network adapter 20 communicates with other modules of the computer device 12 through the bus 18. It should be understood that although not shown in the figure, other hardware and / or software modules can be used in combination with the computer device 12, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems, etc.

[0198] The processing unit 16 executes various functional applications and data processing by running the programs stored in the system memory 28, such as implementing the automatic parking method of the vehicle mentioned in the foregoing embodiments.

[0199] Those skilled in the art will readily conceive of other embodiments of the present application after considering the specification and practicing the application disclosed herein. The present application is intended to cover any variations, uses, or adaptations of the present application, which follow the general principles of the present application and include well-known common general knowledge or conventional technical means in the technical field not disclosed in the present application. The specification and examples are only regarded as exemplary, and the true scope and spirit of the present application are pointed out by the following claims.

[0200] It should be understood that the present application is not limited to the exact structures described above and shown in the drawings, and various modifications and changes can be made without departing from its scope. The scope of the present application is only limited by the appended claims.

[0201] It should be noted that in the description of the present application, the terms "first", "second", etc. are only used for descriptive purposes and cannot be construed as indicating or implying relative importance. In addition, in the description of the present application, unless otherwise specified, the meaning of "a plurality of" is two or more.

[0202] Any process or method description shown in the flowchart or described in other ways herein can be understood as representing a module, segment, or part of code including one or more executable instructions for implementing a specific logical function or process. The scope of the preferred embodiments of the present application includes additional implementations, where the functions can be executed in a substantially simultaneous manner or in a reverse order according to the functions involved, rather than in the order shown or discussed, which should be understood by those skilled in the technical field of the embodiments of the present application.

[0203] It should be understood that each part of the present application can be implemented by hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, any one or a combination of the following techniques well-known in the art can be used: discrete logic circuits having logic gate circuits for implementing logical functions on data signals, application-specific integrated circuits having appropriate combinational logic gate circuits, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0204] Those of ordinary skill in the art can understand that all or part of the steps carried by the method of implementing the above embodiments can be completed by instructing relevant hardware through a program, and the program can be stored in a computer-readable storage medium. When the program is executed, it includes one or a combination of the steps of the method embodiments.

[0205] In addition, each functional unit in various embodiments of the present application may be integrated into a processing module, or each unit may exist physically alone, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. When the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium. The above-mentioned storage medium may be a read-only memory, a magnetic disk, an optical disc, or the like.

[0206] In the description of this specification, the descriptions with reference to the terms "one embodiment", "some embodiments", "examples", "specific examples", or "some examples", etc. mean that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0207] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and should not be construed as limiting the present application. Those of ordinary skill in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present application.

Claims

1. An automatic parking method for a vehicle, characterized in that, Including: Determine an initial position corresponding to a target parking space where the vehicle is to be parked; Control the vehicle to travel to the initial position; Obtain a first image of the vehicle's surrounding environment when the vehicle is at the initial position; Process the first image to obtain a parking label corresponding to the vehicle; Control the vehicle to park in the target parking space according to a parking instruction associated with the parking label.

2. The method according to claim 1, wherein The controlling the vehicle to park in the target parking space according to the parking instruction associated with the parking label includes: If the parking label belongs to a first category label, determine that the parking instruction is to directly park in the target parking space, and control the vehicle to park in the target parking space, where the first category label indicates that the vehicle can park in the target parking space without collision; Or, If the parking label belongs to a third category label, determine that the parking instruction is to terminate parking, where the second category label indicates that the vehicle cannot be parked in the target parking space.

3. The method according to claim 2, wherein If the parking label belongs to a third category label, it further includes: Determine that the parking instruction is to control the vehicle to drive into the target parking space after aligning with the target parking space; Based on a movement adjustment label included in the parking label, control the vehicle to move to adjust the relative position between the vehicle and the target parking space; In response to determining that the vehicle is aligned with the target parking space, control the vehicle to drive into the target parking space, where the movement adjustment label at least includes a distance adjustment label and an angle adjustment label.

4. The method according to claim 3, wherein The responding to determining that the vehicle is aligned with the target parking space includes: In response to determining that the vehicle moves to a first position corresponding to the target parking space, obtain a second image of the vehicle's surrounding environment when the vehicle is at the first position; Process the second image to obtain a target parking label corresponding to the vehicle; If the target parking label is the first category label, determine that the vehicle is aligned with the target parking space.

5. The method according to claim 4, wherein After the processing the second image to obtain a target parking label corresponding to the vehicle, it further includes: If the target parking label is the third category label, based on the target parking label, adjust the relative position between the vehicle and the target parking space until the vehicle is aligned with the target parking space; Control the vehicle to park in the target parking space.

6. The method according to claim 1, characterized in that, The determining an initial position corresponding to a target parking space where the vehicle is to be parked includes: Determine the parking position of the target parking space; Determine a distance interval associated with the attribute information of the vehicle; Based on the distance interval, determine an initial position corresponding to the parking position.

7. The method according to claim 6, characterized in that, The based on the distance interval, determining the initial position corresponding to the parking position includes: Determine the edge line of the target parking space facing the vehicle and the side line of the vehicle facing the target parking space; Determine the angle between the central axis of the target parking space and the central axis of the vehicle; When the vehicle is at any position directly in front of the target parking space, if the distance between the vehicle side line and the edge line of the parking space is within the distance range and the included angle is less than a preset threshold, then determine any of the positions as the initial position corresponding to the parking space position.

8. The method according to claim 1, wherein The processing of the first image to obtain the parking label corresponding to the vehicle includes: Performing feature extraction on the first image to obtain target image features; Inputting the target image features into a pre-constructed position adjustment decision network to obtain a predicted collision category and a distance adjustment label; Inputting the target image features into a pre-constructed angle adjustment decision network to obtain an angle adjustment label; Wherein, the parking label includes the predicted collision category, the distance adjustment label, and the angle adjustment label; When the predicted collision category is no collision, the parking label is a first category label; When the predicted collision category is two-side collision, the parking label is a second category label; When the predicted collision category is one-side collision, the parking label is a third category label.

9. An automatic parking device for a vehicle, characterized in that, Including: A first determination module, configured to determine an initial position corresponding to a target parking space to which the vehicle is to be parked; A first control module, configured to control the vehicle to travel to the initial position; An acquisition module, configured to acquire a first image of the surrounding environment of the vehicle when the vehicle is at the initial position; A processing module, configured to process the first image to obtain a parking label corresponding to the vehicle; A second control module, configured to control the vehicle to park in the target parking space according to a parking instruction associated with the parking label.

10. An electronic device, including: At least one processor; And A memory communicatively connected to the at least one processor; wherein, The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method according to any one of claims 1-8.

11. A non-transitory computer-readable storage medium storing computer instructions, wherein, The computer instructions are used to cause the computer to execute the method according to any one of claims 1-8.

Images