Automatic parking method and device, computer device and storage medium

Abstract

The application relates to an automatic parking method and device, computer equipment and a storage medium. The method comprises the following steps: receiving a parking space matching request, collecting a first all-around-view splicing graph of an environment where a vehicle is located according to the parking space matching request; extracting first image feature points from the first all-around-view splicing graph, and extracting first environment features from the first image feature points; performing feature information matching on the first environment features and pre-stored parking space information; and when the feature information matching is successful, obtaining environment relative position information of a parking space from the pre-stored parking space information matched successfully. The method can improve the parking scene applicability and the parking success rate.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of image processing, in particular to an automatic parking method and device, computer equipment and storage medium. BACKGROUND

[0002] In a real driver parking scene, some parking spaces are exclusive or frequently parked by users, such as community private parking spaces, or commercial parking spaces, etc. The traditional parking system can only search for parking spaces in the driving track when searching for parking spaces, and cannot identify which parking space is needed by the driver. Therefore, the driver needs to observe the display on the vehicle and manually select the target parking space from the multiple searched parking spaces, which greatly restricts the behavior of the driver.

[0003] In addition, due to various reasons such as environmental lighting, obstacle blocking, and reflection of the camera, the traditional parking visual search technology cannot capture effective parking space information, cannot search for parking spaces, and results in parking failure, which leads to poor adaptability of the parking scene. SUMMARY

[0004] Therefore, it is necessary to provide an automatic parking method, device, computer equipment and storage medium capable of improving the adaptability of the parking scene and the success rate of parking.

[0005] An automatic parking method comprises:

[0006] receiving a parking space matching request, collecting a first surround view stitching image of an environment where a vehicle is located according to the parking space matching request;

[0007] extracting a first image feature point from the first surround view stitching image, and extracting a first environment feature from the first image feature point;

[0008] matching the first environment feature with pre-stored parking space information;

[0009] when the feature information matching is successful, obtaining environment relative position information of a parking space from the pre-stored parking space information matched successfully, and selecting the parking space according to the environment relative position information.

[0010] The present application pre-stores parking space information in some commonly used scenes, extracts environment features of the vehicle environment from the collected surround view stitching image of the vehicle environment when receiving a parking space matching requirement, and matches the environment features with the pre-stored parking space information. The environment features are compared to determine whether the environment where the current vehicle is located matches the environment where the pre-stored parking space is located, and the position information of the parking space to be parked is determined from the matched pre-stored parking space information. Therefore, the position of the parking space can also be determined through environment comparison in the case of difficult parking space search, which improves the adaptability and success rate of visual parking

[0011] In one of the embodiments, after obtaining the environmental relative position information of the parking space from the pre-stored parking space information, the method comprises:

[0012] generating a parking instruction and outputting; and / or

[0013] generating a parking space display page according to the environmental relative position information, and sending the parking space display page to a display terminal for display, so that the user can refer to the displayed parking space for driving and parking.

[0014] In one of the embodiments, after obtaining the environmental relative position information of the parking space from the pre-stored parking space information, the method comprises:

[0015] obtaining a current driving position, and obtaining a parking space position according to the environmental relative position information and the current driving position;

[0016] sending the parking space position to a parking control system, so that the parking control system automatically parks according to the parking space position.

[0017] In one of the embodiments, the method of matching the first environmental feature with the pre-stored parking space information comprises:

[0018] calculating a feature matching degree of the first environmental feature and the environmental feature in each of the pre-stored parking space information;

[0019] comparing a highest value in the feature matching degrees with a pre-set matching threshold;

[0020] when the pre-selected feature matching degree exceeds the pre-set matching threshold, determining that the pre-selected feature matching degree corresponds to the pre-stored parking space information, and the feature information of the first environmental feature is matched.

[0021] In one of the embodiments, the method of matching the first environmental feature with the pre-stored parking space information comprises:

[0022] obtaining a pre-stored parking scene identifier, generating a scene selection prompt according to the pre-stored parking scene identifier, and outputting;

[0023] receiving a scene selection instruction corresponding to the scene selection prompt, reading a target parking scene identifier from the scene selection instruction, and finding pre-stored parking space information corresponding to the target parking scene identifier;

[0024] calculating a feature matching degree of the first environmental feature and the pre-stored parking space information found;

[0025] When the matching degree of the found feature exceeds the preset matching threshold, it is determined that the first environment feature matches the feature information of the found pre-stored parking space information.

[0026] In one of the embodiments, before the receiving of the parking space matching request, the method further comprises:

[0027] receiving a parking space memory request, and collecting a second surround view stitching image of an environment where the vehicle is located according to the parking space memory request;

[0028] extracting a second image feature point from the second surround view stitching image, and extracting a second environment feature from the second image feature point, and detecting whether the second environment feature contains a background feature;

[0029] When the second environment feature contains the background feature, generating a parking space environment map according to the environment feature and sending the parking space environment map to a display terminal;

[0030] When receiving selection information returned by the display terminal, extracting an environment relative position of the parking space from the selection information, wherein the selection information is information of the parking space obtained according to the parking space environment map;

[0031] generating pre-stored parking space information according to the environment feature and the environment relative position and storing the pre-stored parking space information.

[0032] In one of the embodiments, the method further comprises:

[0033] When the environment feature does not contain the background feature, generating an environment information acquisition failure prompt, and sending the environment information acquisition failure prompt to the display terminal.

[0034] An automatic parking device, the device comprising:

[0035] an image collection module configured to receive a parking space matching request, and collect a first surround view stitching image of an environment where the vehicle is located according to the parking space matching request;

[0036] a feature extraction module configured to extract a first image feature point from the first surround view stitching image, and extract a first environment feature from the first image feature point;

[0037] a feature matching module configured to perform feature information matching between the first environment feature and pre-stored parking space information;

[0038] a position acquisition module configured to, when the feature information matching is successful, acquire an environment relative position information of a parking space from the pre-stored parking space information for which the feature information matching is successful, and select the parking space according to the environment relative position information.

[0039] A computer device comprises a memory and a processor, the memory stores a computer program, and the processor implements the steps of the above method when executing the computer program.

[0040] A computer readable storage medium, which stores a computer program, the computer program is executed by a processor to implement the steps of the above method.

[0041] The automatic parking method, device, computer device and storage medium described above pre-store parking space information in some common scenarios, when receiving parking space matching demand, extract the environment features of the vehicle environment from the collected surround view stitching map of the vehicle environment, and match the environment features with the pre-stored parking space information, judge whether the environment where the current vehicle is located matches the environment where the pre-stored parking space is located by comparing the environment features, and determine the position information of the parking space to be parked from the matched pre-stored parking space information, so that the parking space position can also be determined by environment comparison in the case of difficult parking space retrieval, improve the applicability and success rate of visual parking, and directly determine the parking space position without the need for the driver to observe the parking space, improve the safety of driving. BRIEF DESCRIPTION OF DRAWINGS

[0042] Figure 1 The application scenario diagram of the automatic parking method in one embodiment is shown in the figure;

[0043] Figure 2 The flowchart of the automatic parking method in one embodiment is shown in the figure;

[0044] Figure 3 The schematic diagram of the environment feature extraction in one embodiment is shown in the figure;

[0045] Figure 4 The schematic diagram of the parking space position in another embodiment is shown in the figure;

[0046] Figure 5 The flowchart of the parking space memory method in another embodiment is shown in the figure;

[0047] Figure 6 The structural block diagram of the automatic parking device in one embodiment is shown in the figure;

[0048] Figure 7 The internal structure diagram of the computer device in one embodiment is shown in the figure. DETAILED DESCRIPTION

[0049] In order to make the purpose, technical scheme and advantages of the present application clearer, the present application is further described in detail below in combination with the 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.

[0050] The automatic parking method provided by the present application can be applied to, for example Figure 1The application environment is shown. Among them, the controller communicates with other vehicle electronic and electrical equipment through CAN (Controller Area Network, Controller Area Network), and the vehicle electronic and electrical system can include IP (Interface Processor, Interface Processor, used for vehicle communication and control), HUT (multimedia host) and camera system, etc. The user can interact with the HUT, send a parking space matching request through the HUT, the HUT sends the parking space matching request to the controller, the controller sends the environment collection instruction to the camera system, the camera system collects the environment map of the environment where the vehicle is located, and generates the surround view stitching map through image processing, and sends the surround view stitching map to the controller. The controller receives the parking space matching request, collects the first surround view stitching map of the environment where the vehicle is located according to the parking space matching request; extracts the first image feature point from the first image feature point, and extracts the first environment feature from the first image feature point; match the first environment feature with the pre-stored parking space information; when the feature information matching is successful, the relative position information of the parking space in the environment is obtained from the pre-stored parking space information matched successfully. The controller can further calculate the real position information of the parking space in the environment according to the relative information, and send the real position information to the IP to control the vehicle parking. Among them, the controller can include a processor and a memory.

[0051] In one embodiment, as shown in Figure 2 An automatic parking method is provided, which is applied to the controller in Figure 1 for example, including the following steps:

[0052] Step 210, receiving a parking space matching request, collecting a first surround view stitching map of the environment where the vehicle is located according to the parking space matching request.

[0053] When the user such as the driver has parking demand, and needs to park into a specific parking space stored in the vehicle system, a parking space matching request can be sent. Specifically, the user can send a parking space matching request by interacting with the HUT (multimedia host), the user can click the parking space matching soft button in the HUT, or make a parking space matching request through voice instruction, and the related system such as the HUT sends the user's parking space matching request to the controller. The user can also send the parking space matching request to the wireless communication system of the vehicle such as the Bluetooth system and the Ethernet system through the terminal Bluetooth or Ethernet module of the personal portable terminal such as the mobile phone and the tablet computer, and the wireless communication system sends the user's parking space matching request to the controller.

[0054] After receiving the parking space matching request, the controller collects a first surround view stitching image of the environment in which the vehicle is located through a camera system of the vehicle. In an embodiment, the camera system can include a plurality of fisheye cameras installed around the vehicle, and the fisheye cameras are used to collect environment images around the vehicle. The collected environment images are sent to the controller, which stitches the environment images collected by the individual fisheye cameras into a whole surround view fisheye image through image compression, correction, and other methods, and then converts the surround view fisheye image into a horizontal real view to obtain the surround view stitching image. In other embodiments, other methods can also be used to obtain the first surround view stitching image.

[0055] In step 220, a first image feature point is extracted from the first surround view stitching image, and a first environment feature is extracted from the first image feature point.

[0056] The controller extracts the first image feature point from the first surround view stitching image. The first image feature point can include a plurality of points that can be used to represent image texture features, shape features, etc., such as inflection points, corner points, points on boundary lines, etc. After the controller extracts the first image feature point, the controller classifies and divides the first image feature point according to the information of the image points and forms different pixel point sets.

[0057] Specifically, before performing image feature point extraction, the controller can convert the first surround view stitching image into a grayscale image and establish a coordinate system in the grayscale image. After the controller extracts the first image feature point, the controller can obtain the grayscale values and coordinate values of the image feature points, and classify the image points according to the grayscale values and coordinate values of the image feature points. In other embodiments, other feature point classification methods can also be used.

[0058] The controller compares the classified feature point set with a pre-stored feature point set to identify the first environment feature parameter. In an embodiment, the FAST feature point detection algorithm or the Harris corner detection algorithm or the SIFT, SURF, etc. algorithm can be used to detect the position of the feature point, and the feature point can be described as a feature point descriptor such as SIFT, SURF descriptor, etc. The feature points are compared and the feature parameters are extracted according to the descriptor.

[0059] The environment feature parameter can include the feature parameter of the parking space and the feature parameter of the background environment of the parking space. The feature parameter of the parking space can include the parking space line, gear line, etc., and the background parameter of the parking space can include the parameter of the wall line, column line, parking indication line, obstacle, etc. Further, the controller can obtain the first environment feature according to the environment feature parameter, such as the length of the parking space line, the length, width, and relative position of the parking space formed by the parking space line, etc. according to the parking space line, and the size of the wall and column, the relative position of the wall and column to the parking space, etc. according to the parameter of the wall line and column line. For example, the controller can obtain the length of the parking space line according to the parameter of the parking space line, and the length, width, and relative position of the parking space formed by the parking space line according to the parameter of the parking space line.Figure 3 As shown in the schematic diagram of the environment feature extraction of an embodiment, the controller can identify the parking space features such as the parking space lines AB and A'B' therefrom, and can identify the background features such as the blocking plates CDE, FGH near the parking space border lines therefrom.

[0060] In step 230, the first environment feature is matched with the pre-stored parking space information.

[0061] The controller acquires the pre-stored parking space information, and matches the extracted first environment feature with the environment features in the pre-stored parking space information, to determine whether the parking space environment where the current vehicle is located is the pre-stored parking space environment scene.

[0062] The pre-stored parking space information is the information of the environment features of the environment where the parking space of each parking scene is located, which can be the information of one scene or multiple scenes. When the pre-stored parking space information only contains the information of one scene, the controller only needs to determine whether the first environment feature matches the pre-stored parking space information. When the pre-stored parking space information contains multiple scenes, the controller needs to match the first environment feature with the pre-stored parking space information of multiple scenes one by one, to determine whether there is pre-stored parking space information matching the first environment feature.

[0063] Due to the parking space light and the possible existence of obstructions on the parking space, etc., the environment features collected by the controller may be missing or different compared to the environment features that can be collected under ideal conditions. When the controller matches the first environment feature with the pre-stored parking space information, the controller can calculate the matching degree of the overall information, and determine whether the feature information matches through the matching degree of the features. Other methods can also be used for feature matching.

[0064] In step 240, when the feature information matching is successful, the environment relative position information of the parking space is acquired from the pre-stored parking space information where the matching is successful, and the parking space is selected according to the environment relative position information.

[0065] When the controller determines that the first environment feature matches the pre-stored parking space information, the controller acquires the parking space identifier of the parking space from the pre-stored parking space information where the matching is successful, and acquires the environment relative position where the parking space corresponding to the parking space identifier is located. When the controller pre-stores the parking space information of each parking scene, the parking space identifier of the specific parking space in each scene is recorded. The environment relative position information can include the relative position of the parking space in all parking spaces, such as the position of the parking space 5 in the 7 parking spaces arranged horizontally, and can also include the relative position distance of the parking space and the background features such as the column, wall surface, and identifier line, etc.

[0066] The controller obtains the relative position between the parking space and the vehicle according to the environment relative position information and the first surround view spliced map, and can further control the vehicle to perform the parking operation.

[0067] In the automatic parking method, the controller pre-stores the parking space information in some common scenes, extracts the environment features of the vehicle environment from the surround view spliced map of the vehicle environment when receiving the parking space matching demand, and matches the environment features with the pre-stored parking space information. Whether the environment where the current vehicle is located matches the environment where the pre-stored parking space is located is determined by comparing the environment features, and the position information of the parking space to be parked is determined from the matched pre-stored parking space information. Therefore, even in the case of difficult parking space retrieval, the parking space position can also be determined by environment comparison, the applicability and success rate of visual parking are improved, and the driver does not need to observe the parking space to directly determine the parking space position, thereby improving the safety of driving.

[0068] In one embodiment, after the step of obtaining the environment relative position information of the parking space from the pre-stored parking space information, the method can further include: generating and outputting a parking instruction; and / or generating a parking space display page according to the environment relative position information, and sending the parking space display page to a display terminal for display.

[0069] After the controller obtains the environment relative position of the parking space, the controller generates a parking instruction and outputs the parking instruction. The controller can directly send the parking instruction to the vehicle control system to control the vehicle to stop, or can send the parking instruction to the HUT system and display a prompt on a multimedia display terminal through voice playback.

[0070] The controller generates a parking space display page according to the environment relative position. Specifically, the controller can determine the specific position of the parking space on the first spliced surround view according to the environment relative position information, mark and display the parking position on the spliced surround view, and generate a parking space display page. For example, the controller can mark the parking space 5 in FIG. 1 with a different color to mark the area where the parking space is located. Figure 4 In other embodiments, other methods can be used to generate a parking space display page. The controller sends the generated parking space display page to a display terminal such as a HUT multimedia host for display, or sends the parking space display page to a user's portable terminal through a wireless network system such as a Bluetooth system or an Ethernet system for display. Therefore, the user can drive and park according to the displayed parking space.

[0071] Further, the controller can determine the position of the parking space relative to the vehicle according to the environment relative position, send the relative position of the parking space to the navigation system, and calculate and generate a parking route by the navigation system. The parking route is displayed on the display terminal by map navigation, or the user is guided to perform the parking operation by voice navigation.

[0072] The method in this embodiment can be applied to scenarios where a driver is present. By matching parking spaces, it automatically selects parking spaces and provides prompts to the user. This eliminates the need for the driver to constantly monitor parking space locations while driving, improving both convenience and safety for the user.

[0073] In one embodiment, after the step of obtaining the environmental relative position information of the parking space from the pre-stored parking space information, the method may further include: obtaining the current driving position, obtaining the parking space position based on the environmental relative position information and the current driving position; and sending the parking space position to the parking control system.

[0074] The controller acquires the current driving position, which is the vehicle's location at the current moment. Based on the current driving position and its relative position to the environment, the controller calculates the parking space location. For example, the controller can calculate the relative distance traveled by the vehicle based on the current driving position and the vehicle's initial position at the time of image acquisition. It then obtains the initial position of the parking space on the first surround-view stitched image corresponding to the vehicle's initial position and calculates the current parking space location based on the initial position and relative distance. The controller sends the calculated parking space location to the parking control system, enabling the parking control system to automatically park the vehicle based on the parking space location.

[0075] like Figure 4 The diagram shows a parking space location in one embodiment. When acquiring the first surround-view stitched image, the controller establishes a coordinate system for the acquisition location. The vehicle's position at point P1 is the image acquisition location. This is achieved by establishing a coordinate system OXY with the vehicle's center position at point P1 as the origin O and the vehicle's direction of travel as the X-axis. For simplicity, the vehicle is generally considered to be traveling along the X-axis. The controller determines the coordinate positions of the parking space lines for each parking space based on the origin position. When the controller successfully matches and identifies parking space 5, it can obtain the coordinate position of parking space 5 at the acquisition time and the vehicle's travel distance S relative to the image acquisition time. Using the vehicle's center position at the current time, as shown in the diagram (P2), as the origin O', a new coordinate system is established. Based on the relative travel distance S, the coordinates of parking space 5 in the OXY coordinate system are converted to the coordinate position in the O'X'Y' coordinate system, allowing the controller to control the vehicle for parking based on these coordinate positions.

[0076] The method in this embodiment can be applied to autonomous driving systems. By automatically matching parking spaces and converting relative positions, the position of the parking space relative to the vehicle at the current moment can be automatically calculated, thereby enabling the autonomous vehicle to drive into the target parking space without requiring the user to perform other operations to select the parking space and control the parking, thus improving parking efficiency.

[0077] In one embodiment, the step of performing feature information matching between the first environment feature and the pre-stored parking space information can include: calculating a feature matching degree between the first environment feature and the environment feature in each pre-stored parking space information; comparing a pre-selected feature matching degree, which is the highest value in the feature matching degrees, with a pre-set matching threshold; and determining that the pre-stored parking space information corresponding to the pre-selected feature matching degree matches the feature information of the first environment feature when the pre-selected feature matching degree exceeds the pre-set matching threshold.

[0078] The controller calculates the feature matching degree between the first environment feature and each pre-stored parking space information one by one. The feature matching degree can be obtained according to the proportion of the matched feature in the first environment feature. The controller can also divide the environment feature into two categories, i.e., parking space features and background features, set different weights for the two categories of features, calculate the feature matching rate of each category of environment features respectively, and then calculate the total feature matching degree according to the feature matching rate and the respective corresponding weight. In other embodiments, the controller can also use other methods to calculate the feature matching degree.

[0079] The controller selects the pre-stored parking space information with the maximum feature matching degree from the plurality of pre-stored parking space information, takes the maximum value of the feature matching degree as a pre-selected feature matching degree, and compares the pre-selected feature matching degree with a pre-set matching threshold. The pre-set matching threshold is the minimum boundary value for determining the feature matching. When the controller determines that the pre-selected feature matching degree exceeds the pre-set matching threshold, it is determined that the pre-stored parking space information corresponding to the maximum value matches the feature information of the first environment feature. When the controller determines that the maximum value of the feature matching degree has not exceeded the pre-set matching threshold, it is determined that there is no parking space information in the pre-stored parking space information that matches the first environment feature.

[0080] In one embodiment, before obtaining the pre-set matching threshold, the controller first detects whether the parking space feature in the first environment feature is complete. The detection can be performed by detecting whether each parking line is connected and closed to each other, for example. When the parking space feature is complete, a pre-set matching threshold with a higher value can be obtained. When the parking space feature is not complete, a pre-set matching threshold with a lower value can be obtained.

[0081] In the above method, considering that the collected environment image may have missing features due to the influence of the shooting environment at the time or other obstructions, feature matching by feature matching degree instead of precise matching can be applied to more application scenarios and improve the scene applicability of parking matching.

[0082] In one embodiment, the step of matching the first environmental feature with the pre-stored parking space information may include: obtaining a pre-stored parking scene identifier; generating and outputting a scene selection prompt based on the pre-stored parking scene identifier; receiving a scene selection instruction corresponding to the scene selection prompt; reading a target parking scene identifier from the scene selection instruction; and searching for pre-stored parking space information corresponding to the target parking scene identifier; calculating the feature matching degree between the first environmental feature and the found pre-stored parking space information; and determining that the first environmental feature matches the feature information of the found pre-stored parking space information when the found feature matching degree exceeds the preset matching threshold.

[0083] The pre-stored parking scene identifier is used to uniquely identify the parking scene to which a parking space belongs. This identification can be done through scene name or code. Parking scenes can include workplace garages, residential garages, etc. The controller obtains the pre-stored parking scene identifier corresponding to each pre-stored parking space and generates scene selection prompts based on these identifiers. The scene selection prompts include scene options for each pre-stored parking scene. The controller can send the scene selection prompts to the vehicle's or user's portable display terminal for display, or it can output them via voice, etc.

[0084] After receiving the scene selection prompt, the user can select a parking scene. After selection, a scene selection command is generated. The corresponding display system or other multimedia host returns the scene selection command to the controller. The controller reads the target parking scene identifier selected by the user from the scene selection command, and then searches for the pre-stored parking space information corresponding to the target parking scene identifier. It calculates the new feature matching degree between the first environmental feature and the found pre-stored parking space. The specific feature matching degree calculation method can refer to the above embodiment, and will not be repeated here. The controller determines whether the first environmental feature matches the pre-stored parking space information based on the comparison result of the feature matching degree and the pre-stored matching threshold.

[0085] In the above method, before feature matching, the user selects the current parking scenario. This allows the system to match only the first environmental features with the pre-stored parking space information of the user-selected scenario, effectively reducing the workload of matching calculations. Furthermore, it enables matching for specific scenarios, improving calculation accuracy.

[0086] In one embodiment, such as Figure 5 The diagram shown is a flowchart of a parking space memory method in one embodiment, which mainly includes the following steps:

[0087] Step 510: Receive parking space memory request, and collect a second surround view mosaic image of the vehicle's environment based on the parking space memory request.

[0088] When the user has a parking space memory requirement for the current parking scenario, a parking space memory request can be issued. Specifically, the user can issue a parking space memory request by interacting with the HUT (multimedia host). The user can click the parking space memory soft button in the HUT or make a parking space memory request through a voice instruction. The related system, such as the HUT, sends the user's parking space memory request to the controller. The user can also send the parking space memory request to the wireless communication system of the vehicle, such as the Bluetooth system or the Ethernet system, through the terminal Bluetooth or Ethernet module of the personal portable terminal, such as the mobile phone or the tablet computer. The wireless communication system then sends the user's parking space memory request to the controller.

[0089] After the controller receives the parking space matching request, the second surround view stitching image of the environment in which the vehicle is located is collected by the camera system of the vehicle. The collection method of the second surround view stitching image can refer to the method described in the above embodiments, which will not be repeated here.

[0090] Step 520: Extract second image feature points from the second surround view stitching image, and extract second environment features from the second image feature points, and detect whether the second environment features contain background features.

[0091] The extraction of the second image feature points and the second environment features by the controller can refer to the extraction method of the first image feature points and the first environment features described in the above embodiments, which will not be repeated here. After the controller extracts the second environment features, it detects whether the second environment features contain background features. The background features refer to the background features of the surrounding environment of the parking space other than the parking space features, which can include wall, column position line, and parking indication line features. Specifically, the controller can classify the extracted environment features by deep learning or other methods, determine the types of the features, and further determine whether the second environment features contain background features.

[0092] In one embodiment, when the environment features do not contain background features, an environment information acquisition failure prompt is generated, and the environment information acquisition failure prompt is sent to the display terminal.

[0093] When the controller detects that the second environment features do not contain background features, it determines that the collected environment features are invalid, generates a memory failure prompt, and prompts the user. The user can further choose to abandon the parking space memory operation or re-perform the parking space memory. When the user chooses to re-perform the parking space memory, the controller performs the step of collecting the surround view stitching image of the current environment of the vehicle according to the re-memory request. When the number of memory failures reaches a preset threshold, a memory invalid prompt is generated and the user is prompted that the current environment cannot perform parking space memory.

[0094] Step 530: When the second environment features contain background features, a parking environment map is generated according to the environment features and sent to the display terminal.

[0095] When the controller detects that the second environment feature contains the background feature, it is determined that the collected environment feature is valid, and the controller obtains the parking space feature parameters and the background feature parameters according to the collected environment feature, generates a parking space environment map according to the parking space feature parameters and the background feature parameters, and the parking space environment map contains the specific size and position information of each parking space and each background object collected.

[0096] The controller sends the parking space environment map to the display terminal for the user to select and determine the parking space to be memorized. The display terminal can be a multimedia terminal on the vehicle or a portable terminal device of the user. The user can select and confirm the target parking space from the multiple parking spaces in the parking space environment map. After the user selects and confirms, the display terminal generates selection information according to the user's selection and sends the selection information to the controller.

[0097] Step 540, when receiving the selection information returned by the display terminal, the relative position of the parking space in the environment is extracted from the selection information, wherein the selection information is the information of the parking space obtained according to the parking space environment map.

[0098] The controller receives the selection information returned by the display terminal, and the terminal extracts the relative position of the selected parking space in the environment from the selection information. The relative position of the environment can be the arrangement position of the parking space in all collected parking spaces, or the coordinate position of the parking space in the environment splicing map, etc.

[0099] Step 550, generate and store the pre-stored parking space information according to the environment feature and the relative position of the environment.

[0100] The controller can number the parking spaces according to the arrangement order of each collected parking space, confirm the parking space number of the parking space according to the relative position of the environment, and record the parking space number of the parking space. The controller obtains the position, size, etc. of the parking space feature and the background feature according to the environment feature, generates the pre-stored parking space information according to the parking space feature parameters and the background feature parameters, and labels the parking space number of the parking space in the pre-stored parking space information. The pre-stored parking space information can be recorded in a standard data format, such as a key-value string, a string array, etc.

[0101] The controller can automatically number and store the pre-stored parking space information, or generate a scene naming prompt to prompt the user to input the scene name or number of the parking memory scene.

[0102] In the above method, when the user wants to remember the parking space, the controller can automatically collect the current parking space environment, and after the user confirms the parking space, the parking space and the environment information are numbered and stored, so that multiple scene information can be stored according to the user's demand, and scene matching and parking space selection can be automatically performed again.

[0103] It should be understood that, although Figures 2-5 The steps in the flowchart of the application are displayed in sequence according to the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, the execution of these steps is not strictly limited in sequence, and these steps can be executed in other orders. Moreover, Figures 2-5 At least part of the steps in the application can include multiple sub-steps or multiple stages, which are not necessarily executed at the same time, but can be executed at different times, and the execution order of these sub-steps or stages is not necessarily sequential, but can be executed alternately or alternately with at least part of other steps or sub-steps or stages of other steps.

[0104] In one embodiment, as shown in Figure 6 An automatic parking device is provided, comprising: an image acquisition module 610, a feature extraction module 620, a feature matching module 630, and a position acquisition module 640, wherein:

[0105] The image acquisition module 610 is configured to receive a parking space matching request and acquire a first surround view stitching image of the environment in which the vehicle is located according to the parking space matching request.

[0106] The feature extraction module 620 is configured to extract first image feature points from the first surround view stitching image and extract first environment features from the first image feature points.

[0107] The feature matching module 630 is configured to perform feature information matching between the first environment features and pre-stored parking space information.

[0108] The position acquisition module 640 is configured to acquire environment relative position information of a parking space from the pre-stored parking space information that is successfully matched when the feature information matching is successful, and select the parking space according to the environment relative position information.

[0109] In one embodiment, the automatic parking device can further comprise:

[0110] The parking instruction module is configured to generate a parking instruction and output the parking instruction.

[0111] The page display module is configured to generate a parking space display page according to the environment relative position information, and send the parking space display page to a display terminal for display.

[0112] In one embodiment, the automatic parking device may further include:

[0113] The location acquisition module is used to obtain the current driving position and obtain the parking space location based on the relative position information of the environment and the current driving position.

[0114] The parking control module is used to send the parking space location to the parking control system.

[0115] In one embodiment, the feature matching module may include:

[0116] The matching calculation unit is used to calculate the feature matching degree between the first environmental feature and the environmental features in each pre-stored parking space information.

[0117] The threshold comparison unit is used to compare the highest value among the feature matching degrees as the pre-selected feature matching degree with the preset matching threshold.

[0118] The matching determination unit is used to determine, when the highest value in the feature matching degree is taken as the pre-selected feature matching degree and the pre-stored parking space information corresponding to the highest value in the feature matching degree exceeds the preset matching threshold, and match it with the feature information of the first environmental feature.

[0119] In one embodiment, the feature matching module may include:

[0120] The scene selection unit is used to obtain the pre-stored parking scene identifier, generate scene selection prompts based on the pre-stored parking scene identifier, and output them.

[0121] The information lookup unit is used to receive the scene selection instruction corresponding to the scene selection prompt, read the target parking scene identifier from the scene selection instruction, and look up the pre-stored parking space information corresponding to the target parking scene identifier.

[0122] The matching calculation unit is used to calculate the feature matching degree between the first environmental features and the found pre-stored parking space information.

[0123] The matching determination unit is used to determine whether the first environmental feature matches the feature information of the pre-stored parking space information when the feature matching degree found exceeds the preset matching threshold.

[0124] In one embodiment, the automatic parking device may further include:

[0125] The memory request module is used to receive parking space memory requests and collect a second surround view mosaic image of the vehicle's environment based on the parking space memory request.

[0126] The feature detection module is used to extract second image feature points from the second panoramic stitched image, extract second environmental features from the second image feature points, and detect whether the second environmental features contain background features.

[0127] The environment map display module is configured to generate a parking space environment map according to the environment features and send the parking space environment map to the display terminal when the background features are contained in the second environment features.

[0128] The information extraction module is configured to extract the environment relative position of the parking space from the selection information returned by the display terminal when the selection information is received, wherein the selection information is the information of the parking space obtained according to the parking space environment map.

[0129] The information storage module is configured to generate and store the pre-stored parking space information according to the environment features and the environment relative position.

[0130] In one embodiment, the automatic parking device can further include:

[0131] The failure prompt module is configured to generate an environment information acquisition failure prompt when the background features are not contained in the environment features, and send the environment information acquisition failure prompt to the display terminal.

[0132] The specific limitations of the automatic parking device can refer to the limitations of the automatic parking method described above, which will not be repeated here. Each module in the above automatic parking device can be realized by software, hardware and their combination in whole or in part. The above modules can be embedded in or independent of the processor in the computer device in hardware form, or can be stored in the memory in the computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0133] In one embodiment, a computer device is provided, which can be a controller, and its internal structure diagram can be as shown in Figure 7 The computer device includes a processor, a memory, a network interface and a database connected by a system bus. The processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operating system and the computer program in the non-volatile storage medium to run. The database of the computer device is configured to store automatic parking data. The network interface of the computer device is configured to communicate with external terminals through network connection. The computer program is executed by the processor to implement an automatic parking method.

[0134] Those skilled in the art can understand that Figure 7 the structure shown in the figure is only a block diagram of part of the structure related to the scheme of the present application, and does not constitute a limitation on the computer device to which the scheme of the present application is applied. The specific computer device can include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0135] In one embodiment, a computer device is provided, comprising a memory and a processor, the memory storing a computer program, and the processor implementing the following steps when executing the computer program: receiving a parking space matching request, collecting a first surround view stitching image of an environment where a vehicle is located according to the parking space matching request; extracting a first image feature point from the first surround view stitching image, and extracting a first environment feature from the first image feature point; performing feature information matching between the first environment feature and pre-stored parking space information; when the feature information matching is successful, obtaining environment relative position information of a parking space from the pre-stored parking space information for which the matching is successful, and selecting the parking space according to the environment relative position information.

[0136] In one embodiment, the processor further implements the following steps when executing the computer program: generating a parking instruction and outputting; and / or generating a parking space display page according to the environment relative position information, and sending the parking space display page to a display terminal for display.

[0137] In one embodiment, the processor further implements the following steps when executing the computer program: obtaining a current driving position, obtaining a parking space position according to the environment relative position information and the current driving position; and sending the parking space position to a parking control system.

[0138] In one embodiment, when the processor implements the step of performing feature information matching between the first environment feature and the pre-stored parking space information, it is further used for: calculating a feature matching degree between the first environment feature and environment features in each pre-stored parking space information, comparing a highest value in the feature matching degrees as a preselected feature matching degree with a pre-set matching threshold; when the highest value in the feature matching degrees as the preselected feature matching degree exceeds the pre-set matching threshold, determining that the pre-stored parking space information corresponding to the highest value in the feature matching degrees as the preselected feature matching degree is matched with the feature information of the first environment feature.

[0139] In one embodiment, when the processor implements the step of performing feature information matching between the first environment feature and the pre-stored parking space information, it is further used for: obtaining a pre-stored parking scene identifier, generating a scene selection prompt according to the pre-stored parking scene identifier and outputting; receiving a scene selection instruction corresponding to the scene selection prompt, reading a target parking scene identifier from the scene selection instruction, and searching for pre-stored parking space information corresponding to the target parking scene identifier; calculating a feature matching degree between the first environment feature and the searched pre-stored parking space information; when the searched feature matching degree exceeds a pre-set matching threshold, determining that the feature information of the first environment feature is matched with the searched pre-stored parking space information.

[0140] In one embodiment, the processor, when executing the computer program, also implements the following steps: receiving a parking space memory request, collecting a second surround view stitching image of an environment in which the vehicle is located according to the parking space memory request; extracting a second image feature point from the second surround view stitching image, and extracting a second environment feature from the second image feature point; detecting whether the second environment feature contains a background feature; when the second environment feature contains the background feature, generating a parking space environment map according to the environment feature and sending the parking space environment map to a display terminal; and when receiving selection information returned by the display terminal, extracting the relative position of the parking space from the selection information, wherein the selection information is information of the parking space obtained according to the parking space environment map.

[0141] In one embodiment, the processor, when executing the computer program, also implements the following steps: when the environment feature does not contain the background feature, generating an environment information acquisition failure prompt, and sending the environment information acquisition failure prompt to the display terminal.

[0142] In one embodiment, a computer readable storage medium is provided, and the computer readable storage medium stores a computer program. The computer program, when executed by a processor, implements the following steps: receiving a parking space matching request, collecting a first surround view stitching image of an environment in which the vehicle is located according to the parking space matching request; extracting a first image feature point from the first surround view stitching image, and extracting a first environment feature from the first image feature point; performing feature information matching between the first environment feature and pre-stored parking space information; when the feature information matching is successful, obtaining the relative position information of the parking space from the pre-stored parking space information corresponding to the highest value of the feature matching degrees, and selecting the parking space according to the relative position information of the parking space.

[0143] In one embodiment, the computer program, when executed by the processor, also implements the following steps: generating a parking instruction and outputting the parking instruction; and / or generating a parking space display page according to the relative position information of the parking space, and sending the parking space display page to a display terminal for display.

[0144] In one embodiment, the computer program, when executed by the processor, also implements the following steps: obtaining a current driving position, obtaining the parking space position according to the relative position information of the parking space and the current driving position, and sending the parking space position to a parking control system.

[0145] In one embodiment, when the computer program, when executed by the processor, implements the step of performing feature information matching between the first environment feature and the pre-stored parking space information, the computer program is also used for: calculating a feature matching degree between the first environment feature and the environment feature in each pre-stored parking space information, comparing a highest value in the feature matching degrees with a pre-set matching threshold as a pre-selected feature matching degree; and when the highest value in the feature matching degrees as the pre-selected feature matching degree exceeds the pre-set matching threshold, determining that the pre-stored parking space information corresponding to the highest value in the feature matching degrees as the pre-selected feature matching degree is matched with the feature information of the first environment feature.

[0146] In one embodiment, the computer program, when executed by the processor, further implements the step of matching the first environment feature with the pre-stored parking space information, and is further configured to: obtain a pre-stored parking scene identifier, generate a scene selection prompt according to the pre-stored parking scene identifier, and output the scene selection prompt; receive a scene selection instruction corresponding to the scene selection prompt, read a target parking scene identifier from the scene selection instruction, and search for pre-stored parking space information corresponding to the target parking scene identifier; calculate a feature matching degree between the first environment feature and the searched pre-stored parking space information; and determine that the first environment feature matches the feature information of the searched pre-stored parking space information when the searched feature matching degree exceeds a preset matching threshold.

[0147] In one embodiment, the computer program, when executed by the processor, further implements the following steps: receiving a parking space memory request, collecting a second surround view stitching image of an environment in which the vehicle is located according to the parking space memory request; extracting a second image feature point from the second surround view stitching image, and extracting a second environment feature from the second image feature point; detecting whether the second environment feature contains a background feature; generating a parking space environment map according to the environment feature and sending the parking space environment map to a display terminal when the second environment feature contains the background feature; and extracting an environment relative position of the parking space from selection information returned by the display terminal when the selection information is received, wherein the selection information is information of the parking space obtained according to the parking space environment map.

[0148] In one embodiment, the computer program, when executed by the processor, further implements the following steps: generating an environment information acquisition failure prompt when the environment feature does not contain the background feature, and sending the environment information acquisition failure prompt to the display terminal.

[0149] 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. The computer program can be stored in a non-volatile computer readable storage medium, and when the computer program is executed, the processes of the above-mentioned embodiments of the methods can be included. Any reference to memory, storage, databases, or other media in the embodiments provided by the present 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 but not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), 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), etc.

[0150] The technical features of the above embodiments can be combined in any way. In order to make the description simple, not all possible combinations of the technical features in the above embodiments are described, but as long as the combinations of the technical features do not exist, they should be considered as the scope of the present application.

[0151] The above embodiments only express several implementation manners of the present application, and the description is more specific and detailed, but it should not be understood as a limitation on the scope of the patent. It should be pointed out that for ordinary skilled in the art, without departing from the concept of the present application, some modifications and improvements can be made, which are all within the scope of the present application. Therefore, the scope of the patent of the present application should be subject to the appended claims.

Claims

1. An automatic parking method, comprising: receiving a parking space matching request, collecting a first surround view stitching image of an environment in which a vehicle is located according to the parking space matching request; extracting first image feature points from the first surround view stitching image, and extracting first environment features from the first image feature points, the first environment features including parking space features and background features; performing feature information matching between the first environment features and pre-stored parking space information; when the feature information matching is successful, obtaining environment relative position information of a parking space from the pre-stored parking space information for which the matching is successful, and selecting a parking space for parking according to the environment relative position information; the feature information matching between the first environment features and the pre-stored parking space information comprises: obtaining a pre-stored parking scene identifier, generating a scene selection prompt according to the pre-stored parking scene identifier, and outputting the scene selection prompt; receiving a scene selection instruction corresponding to the scene selection prompt, reading a target parking scene identifier from the scene selection instruction, and searching for pre-stored parking space information corresponding to the target parking scene identifier; calculating a feature matching degree of the first environment features and the pre-stored parking space information found; wherein the parking space features and the background features correspond to different weights, the feature matching rates of the parking space features and the background features are calculated respectively, and the total feature matching degree is calculated based on the feature matching rates and the weights; when the calculated feature matching degree exceeds a preset matching threshold, it is determined that the feature information matching between the first environment features and the pre-stored parking space information found is successful; wherein the controller detects whether the parking space features in the first environment features are complete before obtaining the preset matching threshold, and the preset matching threshold when the parking space features are complete is greater than the preset matching threshold when the parking space features are not complete.

2. The method of claim 1, wherein, after the environment relative position information of the parking space is obtained from the pre-stored parking space information, comprising: generating a parking instruction and outputting the parking instruction; and / or generating a parking space display page according to the environment relative position information, and sending the parking space display page to a display terminal for display.

3. The method of claim 2, wherein, after the environment relative position information of the parking space is obtained from the pre-stored parking space information, comprising: obtaining a current driving position, obtaining a parking space position according to the environment relative position information and the current driving position; sending the parking space position to a parking control system.

4. The method of claim 1, wherein, before the parking space matching request is received, further comprising: receiving a parking space memory request, collecting a second surround view stitching image of an environment in which a vehicle is located according to the parking space memory request; extracting second image feature points from the second surround view stitching image, and extracting second environment features from the second image feature points, and detecting whether the second environment features include background features; when the second environment features include the background features, generating a parking space environment image according to the environment features and sending the parking space environment image to a display terminal; when receiving selection information returned by the display terminal, extracting environment relative position information of the parking space from the selection information, wherein the selection information is information of a parking space obtained according to the parking space environment image; The pre-stored parking space information is generated according to the environment feature and the environment relative position and stored.

5. The method of claim 4, wherein, The method further comprises: When the background feature is not contained in the environment feature, an environment information acquisition failure prompt is generated, and the environment information acquisition failure prompt is sent to a display terminal.

6. An automatic parking apparatus characterized by comprising: The device comprises: An image acquisition module is configured to receive a parking space matching request, and acquire a first surround view stitching image of an environment in which a vehicle is located according to the parking space matching request. A feature extraction module is configured to extract first image feature points from the first surround view stitching image, and extract first environment features from the first image feature points, wherein the first environment features comprise parking space features and background features. A feature matching module is configured to perform feature information matching between the first environment features and pre-stored parking space information, comprising: acquiring a pre-stored parking scene identifier, generating a scene selection prompt according to the pre-stored parking scene identifier, and outputting the scene selection prompt; receiving a scene selection instruction corresponding to the scene selection prompt, reading a target parking scene identifier from the scene selection instruction, and searching for pre-stored parking space information corresponding to the target parking scene identifier; calculating a feature matching degree between the first environment features and the searched pre-stored parking space information; wherein the parking space features and the background features correspond to different weights, the feature matching rates of the parking space features and the background features are calculated respectively, and the total feature matching degree is calculated based on the feature matching rates and the weights; when the calculated feature matching degree exceeds a preset matching threshold, it is determined that the feature information of the first environment features and the searched pre-stored parking space information is matched; wherein the controller detects whether the parking space features in the first environment features are complete before acquiring the preset matching threshold, and the preset matching threshold when the parking space features are complete is greater than the preset matching threshold when the parking space features are not complete. A position acquisition module is configured to acquire environment relative position information of a parking space from the pre-stored parking space information matched successfully when the feature information matching is successful, and select a parking space according to the environment relative position information.

7. The apparatus of claim 6, wherein, The device further comprises: A parking instruction module is configured to generate a parking instruction and output the parking instruction; and / or A page display module is configured to generate a parking space display page according to the environment relative position information, and send the parking space display page to a display terminal for display. 8.A computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the computer device is configured to perform the method according to any one of claims 1-7. The processor executes the computer program to implement the steps of the method of any one of claims 1 to 5.

9. A computer readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the steps of the method of any one of claims 1 to 5.

Images