Parking space detection method and device, parking system, vehicle, storage medium and product

Abstract

The invention relates to the technical field of automatic parking, in particular to a parking space detection method and device, a driving system, a controller, a vehicle, a storage medium and a product. The method comprises the following steps: receiving multi-frame time sequence parking space data, wherein the multi-frame time sequence parking space data comprises current frame time sequence parking space data; generating a parking space feature data cluster according to the multi-frame time sequence parking space data; and determining target parking space feature data according to the current time sequence frame parking space data and the parking space feature data cluster. The parking space feature data cluster is generated through the multi-frame time sequence parking space data, and the target parking space feature data is determined in combination with the current time sequence frame parking space data and the parking space feature data cluster, so that the data precision of parking space detection in an automatic parking process is improved; and the problems of redundant parking in the garage, parking deviation and the like caused by parking space data errors are solved.

Description

Technical Field

[0001] The present application relates to the field of automatic parking technology, and in particular to a parking space detection method, device, parking system, controller, vehicle, storage medium and product, wherein the storage medium is a computer-readable storage medium and the product is a computer program product. Background Art

[0002] In automated parking systems, the accuracy and stability of parking space recognition continue to play a crucial role, profoundly impacting the decision-making module's ability to achieve more reliable and efficient results. However, existing parking space data processing and prediction technologies primarily rely on models or rule-based data. This approach can only analyze overall parking space data and cannot provide precise data on parking spaces. This leads to issues such as insufficient parking space data accuracy and poor parking stability. Summary of the Invention

[0003] The embodiments of the present application provide a parking space detection method, device, parking system, controller, vehicle, storage medium and product, which can generate a parking space feature data cluster through multi-frame time-series parking space data, and determine the target parking space feature data by combining the current time-series frame parking space data and the parking space feature data cluster, thereby improving the data accuracy of parking space detection during automatic parking and solving problems such as unnecessary parking and off-center parking caused by parking space data errors.

[0004] To achieve the above objectives, according to a first aspect of the present application, a parking space detection method is provided, comprising: Receive multiple frames of time-series parking space data, wherein the multiple frames of time-series parking space data include current frame time-series parking space data; generate parking space feature data clusters according to the multiple frames of time-series parking space data; and determine target parking space feature data according to the current frame time-series parking space data and the parking space feature data clusters.

[0005] In some embodiments, the parking space data includes parking space feature data, and the parking space feature data includes at least one of parking space corner point coordinate data, parking space direction data, and object type coordinate point data in the parking space.

[0006] In some embodiments, generating a cluster of parking space feature data based on the multiple frames of time-series parking space data includes: Acquire first parking space feature data, where the first parking space feature data is the parking space feature data in the current time series frame data; update the first parking space feature data capacity pool according to the first parking space feature data; and generate a first parking space feature data cluster based on the first parking space feature data capacity pool, where the parking space feature data cluster includes the first parking space feature data cluster.

[0007] In some embodiments, the first parking space feature data capacity pool includes a first storage capacity and a preset storage capacity, the first storage capacity being the storage quantity of the first parking space feature data currently stored in the first parking space feature data capacity pool, and the preset storage capacity being the maximum storage capacity of the first parking space feature data capacity pool; Updating the first parking space feature data capacity pool according to the first parking space feature data includes: When the first storage capacity is less than the preset storage capacity, the first parking space feature data is stored in the first parking space feature data capacity pool; when the first storage capacity is greater than or equal to the preset storage capacity, the first parking space feature data capacity pool is updated according to the first parking space feature data.

[0008] In some embodiments, storing the first parking space feature data into the first parking space feature data capacity pool includes: determining whether the first parking space feature data is key frame data; In a case where the first parking space feature data is not the key frame data, storing the first parking space feature data into the first parking space feature data capacity pool; In the case that the first parking space feature data is the key frame data, the first parking space feature data is enhanced to generate a first parking space feature data set, and the first parking space feature data set is stored in the first parking space feature data capacity pool, wherein the first parking space feature data set includes the first parking space feature data, and the amount of data in the first parking space feature data set is equal to the difference between the preset storage capacity and the first storage capacity.

[0009] In some embodiments, the enhancing the first parking space feature data to generate a first parking space feature data set includes: generating the first parking space feature data set according to the first parking space feature data, a preset distance, and a first parking space feature data capacity pool.

[0010] In some embodiments, generating the first parking space feature data set according to the first parking space feature data, a preset distance, and a first parking space feature data capacity pool includes: Determining whether first target parking space characteristic data exists within a circle with the first parking space characteristic data as the center and the preset distance as the radius, wherein the first target parking space characteristic data is any other first parking space characteristic data existing in the first parking space characteristic data capacity pool; In a case where the first target parking space feature data exists within the circle, the first parking space feature data set is generated according to the first parking space feature data and the first target parking space feature data.

[0011] In some embodiments, generating the first parking space feature data set according to the first parking space feature data and the first target parking space feature data includes: A virtual first parking space feature data is obtained between the first parking space feature data and the first target parking space feature data, and the virtual first parking space feature data is stored in the first parking space feature data set until the first parking space feature data set is filled.

[0012] In some embodiments, the method further comprises: When the first target parking space feature data does not exist within the circle, or when the first target parking space feature data exists within the circle and the amount of the virtual first parking space feature data is less than the difference between the preset storage capacity and the first storage capacity: copy the first parking space feature data and store it in the first parking space feature data set until the first parking space feature data set is filled.

[0013] In some embodiments, updating the first parking space feature data capacity pool according to the first parking space feature data includes: Generate the first parking space feature data cluster based on the first parking space feature data capacity pool, and the parking space feature data cluster includes a cluster center; obtain the second distance between the first parking space feature data and the cluster center, the first parking space feature data farthest from the cluster center in the first parking space feature data capacity pool, and the third distance between the cluster center and the third parking space feature data; when the third distance is greater than the second distance, store the first parking space feature data in the first parking space feature data capacity pool, and delete the first parking space feature data farthest from the cluster center.

[0014] In some embodiments, the parking space feature data cluster includes a target cluster center point, and the target cluster center point is the cluster center point with the most parking space feature data in the parking space feature data cluster. Determining the target parking space feature data based on the current time series frame parking space data and the parking space feature data cluster includes: Determine whether the parking space data of the current time frame meets the confidence requirement based on the parking space data of the current time frame and the target cluster center point; if the parking space data of the current time frame meets the confidence requirement, set the current time frame data as the target parking space feature data; if the parking space data of the current time frame does not meet the confidence requirement, set the parking space feature data corresponding to the target cluster center point as the target parking space feature data.

[0015] In some embodiments, determining whether the parking space data of the current time sequence frame meets the confidence requirement based on the parking space data of the current time sequence frame and the target cluster center point includes: Obtain a fourth distance between the parking space data of the current time sequence frame and the target cluster center point; when the fourth distance is less than or equal to the set distance threshold, the parking space data of the current time sequence frame meets the confidence requirement; when the fourth distance is greater than the set distance threshold, the parking space data of the current time sequence frame does not meet the confidence requirement.

[0016] According to a second aspect of the present application, a parking space detection device is provided, comprising: A data receiving module, the data receiving module is used to receive multiple frames of time-series parking space data, the multiple frames of time-series parking space data including the current frame of time-series parking space data; A data processing module, connected to the data receiving module, configured to generate a parking space feature data cluster based on the multiple frames of time-series parking space data; A data confirmation module is connected to the data processing module and is used to determine target parking space feature data based on the current time sequence frame parking space data and the parking space feature data cluster.

[0017] According to a third aspect of the present application, there is provided an automatic parking system, comprising: The parking control module and the second aspect are provided in an embodiment. The parking control module is communicatively connected to the parking space detection device. The parking space detection device is configured to detect parking space data to determine target parking space characteristic data. The parking control module is configured to control the vehicle to automatically park based on the target parking space characteristic data. In some embodiments, the automatic parking system further includes a parking space data acquisition module, which is connected to the parking space detection device and is configured to acquire the parking space data using a visual sensor.

[0018] In some embodiments, the automatic parking system also includes a target posture calculation module, which is connected to the parking space detection device and is used to receive the target position feature data sent by the parking space detection device and perform parking position calculation based on the target position feature data; the target posture calculation module is connected to the parking control module, and the parking control module is also used to control the vehicle to achieve automatic parking based on the parking space posture calculation result.

[0019] According to the fourth aspect of the present application, a controller is provided, comprising a memory and a processor; the memory stores a computer program, and the processor is used to run the computer program in the memory to execute any one of the parking space detection methods provided in the embodiments of the first aspect of the present application.

[0020] According to the fifth aspect of the present application, a vehicle is provided, comprising the controller described in the fourth aspect of the version application, or the automatic parking system described in the third aspect of the present application.

[0021] According to the sixth aspect of the present application, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, any parking space detection method provided in the embodiment of the first aspect of the present application is implemented.

[0022] According to the seventh aspect of the present application, a computer program product is provided, comprising a computer program or instructions, which, when executed by a processor, implements any one of the parking space detection methods provided in the embodiments of the first aspect of the present application.

[0023] To sum up, this application generates parking space feature data clusters through multi-frame time-series parking space data, and determines the target parking space feature data by combining the current time-series frame parking space data and the parking space feature data clusters, thereby improving the data accuracy of parking space detection during automatic parking, and solving problems such as unnecessary parking and off-center parking caused by parking space data errors. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] To more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present application. Those skilled in the art can also derive other drawings based on these drawings without inventive effort.

[0025] In order to more completely understand the present application and its beneficial effects, the following description will be given in conjunction with the accompanying drawings, wherein the same drawing numbers represent the same parts in the following description.

[0026] Figure 1 is a flow chart of a parking method provided in an exemplary embodiment of the present application; Figure 2 is a flow chart of another parking method provided in an exemplary embodiment of the present application; Figure 3 This is a schematic diagram of observation results of cluster centers of parking space feature data provided in an exemplary embodiment of the present application; Figure 4 This is a schematic diagram of a parking space feature data cluster center point observation clustering result provided in an exemplary embodiment of the present application; Figure 5 This is a schematic diagram of a SMOTE algorithm for parking space feature key frame data provided in an exemplary embodiment of the present application; Figure 6is a schematic diagram of a parking space detection device provided in an exemplary embodiment of the present application; Figure 7 is a schematic diagram of an automatic parking system provided in an exemplary embodiment of the present application; Figure 8 It is a schematic structural diagram of a controller provided in an exemplary embodiment of the present application. DETAILED DESCRIPTION

[0027] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the embodiments described are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present application.

[0028] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this application belongs; the terms used herein are only for the purpose of describing specific embodiments and are not intended to limit this application; the terms "including" and "having" and any variations thereof in the specification and claims of this application and the above-mentioned figure descriptions are intended to cover non-exclusive inclusions.

[0029] In the description of the embodiments of this application, the technical terms "first" and "second" are used only to distinguish different objects and should not be understood to indicate or imply relative importance or implicitly specify the quantity, specific order, or primary and secondary relationship of the indicated technical features. In the description of the embodiments of this application, the meaning of "plurality" is more than two, unless otherwise clearly and specifically defined.

[0030] References herein to "embodiments" mean that a particular feature, structure, or characteristic described in connection with the embodiments may be included in at least one embodiment of the present application. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor does it constitute an independent or alternative embodiment that is mutually exclusive of other embodiments. It is understood, both explicitly and implicitly, by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0031] In the description of the embodiments of this application, the term "and / or" is simply a description of the association relationship between associated objects, indicating that three relationships can exist. For example, A and / or B can represent the following three situations: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the character " / " in this document generally indicates that the associated objects are in an "or" relationship.

[0032] In the description of the embodiments of the present application, the term "multiple" refers to more than two (including two). Similarly, "multiple groups" refers to more than two groups (including two groups), and "multiple pieces" refers to more than two pieces (including two pieces).

[0033] In the description of the embodiments of this application, unless otherwise specified or limited, technical terms such as "connected" and "connection" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, an indirect connection through an intermediate medium, or a connection through a network. For those of ordinary skill in the art, the specific meanings of the above terms in the embodiments of this application can be understood according to specific circumstances.

[0034] The present application embodiment provides a parking space detection method, such as Figure 1 As shown, the specific process of the parking space detection can be as follows:

[0035] S101: Receive multiple frames of time-series parking space data, where the multiple frames of time-series parking space data include the current frame of time-series parking space data.

[0036] During automated parking, parking space data is typically collected by the vehicle's automated parking system's video sensors. These sensors typically capture multiple frames of continuous, time-series parking space data, including the current frame's time-series parking space data. Time-series parking space data represents the time-series parking space data captured by the video sensors, sorted by acquisition time. The current frame's time-series parking space data refers to the latest parking space data to be processed, typically the last frame's time-series parking space data.

[0037] The present invention is applicable to parking lot scenarios of various types and sizes, including indoor parking lots, outdoor parking lots, multi-story parking lots, etc., and parking spaces include normal parking spaces, mechanical parking spaces, inclined parking spaces, irregular parking spaces, etc.

[0038] In the actual parking space data processing process, the video sensor first sends the collected continuous multi-frame time-series parking space data to the parking space data acquisition module in the automatic parking system for data processing. The parking space data acquisition module decomposes the continuous multi-frame time-series parking space data into frame-by-frame time-series parking space data, stores the time-series parking space data, and processes the parking space-related features.

[0039] In some embodiments, the parking space data generally includes parking space feature data, and the parking space feature data includes at least one of parking space corner point coordinate data, parking space direction data, and object type coordinate point data in the parking space.

[0040] For example, parking space data generally include parking space data such as the location, direction, length, width, shape, corner points and interior of the parking space. During the parking process, attention is usually paid to parking space feature data, which include parking space corner point coordinate data. Parking space corner point coordinate data include the four corner point coordinates of the parking space and other corner point coordinate data; parking space direction data include parking space orientation, parking space angle, included angle, or angle formed with the vehicle during parking; object type coordinate point data in the parking space include some special objects inside the parking space, such as parking wheel stop rods, wheel stop piers, parking locks, limit blocks, internal protrusions or other obstacles in the garage.

[0041] After storing the time-series parking data, the parking space data acquisition module identifies and extracts parking space feature data for each received frame. It analyzes and categorizes these features, identifies the coordinate points within the time-series parking data, and determines whether each type of parking space feature data within the time-series parking data represents a key frame within that type of parking space feature data. The criticality of a frame can be determined based on the corresponding conditions when the vehicle body and the parking space are perpendicular or parallel, forming a specific geometric relationship, or when the visual sensor captures key parking space data, such as observing the two corner points at the bottom of the parking space.

[0042] Optionally, the parking space data acquisition module may include one or more of a post-processing module, a perception fusion module, or a mapping module. The specific module depends on the architecture design of the automatic parking system. These data processing modules have been widely used in existing automatic parking systems and will not be introduced in detail here.

[0043] Optionally, determining whether parking space data of different feature types are key frame data usually includes performing missing value processing and / or outlier processing on the parking space feature data, that is, determining whether the parking space feature data is a missing value or an outlier. If the parking space feature data is neither a missing value nor an outlier, it can be determined that the parking space feature data is key frame data; otherwise, it is non-key frame data.

[0044] S102: Generate parking space feature data clusters based on the multiple frames of time-series parking space data:

[0045] After the parking space data acquisition module identifies and extracts the parking space feature data from each frame of time-series parking space data received, the parking space feature data is sent to the parking space detection device for data detection.

[0046] After receiving multiple frames of time-series parking space data, the parking space detection device generates a cluster of parking space feature data based on the multiple frames of time-series parking space data.

[0047] Specifically, after receiving the parking space data of the current frame, the parking space data acquisition module will process the parking space feature data of the number of parking spaces in the current frame, and send the parking space feature data of the number of parking spaces in the current frame to the parking space detection device for data processing.

[0048] After receiving the parking space characteristic data from the parking space data acquisition module, the parking space detection device stores each type of parking space characteristic data in data pools of different sizes. Each type of parking space characteristic data corresponds to a data pool, and the preset storage capacity of the data pool for different types of parking space characteristic data varies. Optionally, the preset storage capacity is typically the maximum capacity of the data pool.

[0049] A data pool is a collection or warehouse that stores and manages large amounts of data. In fields such as data processing, analytics, and machine learning, data pools are used to centrally store data from various sources for unified management and efficient access. Data in a data pool may include structured data (such as database tables), semi-structured data (such as JSON or XML files), and unstructured data (such as text, images, or audio files). Data pool management typically involves data cleansing, integration, backup, and security.

[0050] In some embodiments, these parking space feature data include first parking space feature data, and the parking space detection device will obtain the first parking space feature data, wherein the first parking space feature data is the parking space feature data in the current time series frame data; the parking space detection device updates the first parking space feature data capacity pool according to the first parking space feature data, wherein the first parking space feature data capacity pool stores the parking space feature data; the parking space detection device will also generate a parking space feature data cluster based on the first parking space feature data capacity pool, and the parking space feature data cluster includes the first parking space feature data cluster.

[0051] The parking space detection device will perform clustering calculations on each type of parking space feature data to generate parking space feature data clusters corresponding to each type of parking space feature data. First, it will obtain each type of parking space feature data and update each type of parking space feature data to the corresponding parking space feature data capacity pool. Then, based on the updated parking space feature data capacity pool, it will generate parking space feature data clusters corresponding to each type of parking space feature data.

[0052] A cluster is a collection of similar objects formed during cluster analysis. Clustering is an unsupervised learning method used to divide a dataset into multiple groups or clusters, so that objects within the same cluster are similar, while objects in different clusters are quite different. Therefore, each type of parking space feature data corresponds to a separate cluster.

[0053] Exemplarily, the parking space feature data corresponding to a type of parking space feature data is first parking space feature data, wherein the parking space feature data includes the first parking space feature data, and the parking space feature data capacity pool corresponding to the first parking space feature data is the first parking space feature data capacity pool. The parking space detection device updates the first parking space feature data capacity pool based on the first parking space feature data, and performs clustering calculation using the updated first parking space feature data capacity pool to generate a first parking space feature data cluster corresponding to the first parking space feature data.

[0054] In some embodiments, the first parking space feature data capacity pool includes a first storage capacity and a preset storage capacity, the first storage capacity being the storage quantity of the first parking space feature data currently stored in the first parking space feature data capacity pool, and the preset storage capacity being the maximum storage capacity of the first parking space feature data capacity pool; updating the first parking space feature data capacity pool according to the first parking space feature data includes: When the first storage capacity is less than the preset storage capacity, the first parking space characteristic data is stored in the first parking space characteristic data capacity pool; when the first storage capacity is greater than or equal to the preset storage capacity, the first parking space characteristic data capacity pool is updated according to the first parking space characteristic data.

[0055] Specifically, the parking space feature data capacity pool corresponding to the parking space feature data will be set with a maximum data storage capacity. Correspondingly, when the parking space feature data stored in the parking space feature data capacity pool has not reached the maximum storage capacity, the parking space feature data stored in the parking space feature data capacity pool has a corresponding storage quantity; for the first parking space feature data capacity pool, its maximum data storage capacity can be set to a preset storage capacity, and the storage capacity corresponding to the first parking space feature data stored in the first parking space feature data capacity pool is set to the first storage capacity.

[0056] In the process of updating the first parking space feature data capacity pool based on the first parking space feature data, it is usually determined by comparing the first storage capacity of the first parking space feature data stored in the first parking space feature data capacity pool with the preset storage capacity to confirm the relationship between the first storage capacity and the preset storage capacity, so as to determine how to update the data in the first parking space feature data capacity pool.

[0057] Optionally, when the first storage capacity is less than a preset storage capacity, the first parking space characteristic data is stored in a first parking space characteristic data capacity pool.

[0058] Optionally, when the first storage capacity is greater than or equal to the preset storage capacity, the first parking space feature data capacity pool is updated according to the first parking space feature data.

[0059] In some embodiments, storing the first parking space feature data into the first parking space feature data capacity pool includes: determining whether the first parking space feature data is key frame data; if the first parking space feature data is not key frame data, storing the first parking space feature data into the first parking space feature data capacity pool; if the first parking space feature data is key frame data, performing enhancement processing on the first parking space feature data to generate a first parking space feature data set, and storing the first parking space feature data set into the first parking space feature data capacity pool, wherein the first parking space feature data set includes the first parking space feature data, and the amount of data in the first parking space feature data set is equal to the difference between the preset storage capacity and the first storage capacity.

[0060] Specifically, when the first storage capacity is less than the preset storage capacity, the data capacity of the first parking space feature data stored in the first parking space feature data capacity pool does not reach the basic data capacity for clustering calculation. Therefore, it is necessary to continue to store the received first parking space feature data in the first parking space feature data capacity pool until the first storage capacity is equal to the preset storage capacity, that is, the first parking space feature data capacity pool meets the basic data capacity for clustering calculation.

[0061] However, it is necessary to determine whether the received first parking space characteristic data is key frame data. If the first parking space characteristic data is not key frame data, the first parking space characteristic data is directly stored in the first parking space characteristic data capacity pool until the data capacity in the first parking space characteristic data capacity pool reaches a preset storage capacity.

[0062] If the first parking space feature data is key frame data, it is necessary to use the Synthetic Minority Over-sampling Technique (SMOTE) algorithm to enhance the first parking space feature data, such as Figure 5 As shown, the enhancement process performs virtual point derivation on the data near the coordinate point of the first parking space feature data to generate a data set containing the virtual point data and the first parking space feature data, namely, the first parking space feature data set. All parking space feature data in the first parking space feature data set is then stored in the first parking space feature data capacity pool. The first parking space feature data set includes the first parking space feature data and the first parking space feature data corresponding to the virtual point derived using the SMOTE algorithm. The amount of data in the first parking space feature data set is generally equal to the difference between the preset storage capacity and the first storage capacity.

[0063] Optionally, the judgment of whether the first parking space feature data is key frame data is usually performed in the parking space data acquisition module, and the judgment method usually includes missing value processing and outlier processing, that is, judging whether the first parking space feature data is a missing value or an outlier. If the first parking space feature data is neither a missing value nor an outlier, it can be determined that the first parking space feature data is key frame data.

[0064] In some embodiments, the first parking space feature data set is generated according to the first parking space feature data, a preset distance, and a first parking space feature data capacity pool.

[0065] The first parking space feature data set is typically generated using the first parking space feature data in the current time-series frame data, a first parking space feature data capacity pool, and a preset distance, where the preset distance is a configurable distance value. The first parking space feature data is placed in the first parking space feature data capacity pool and compared with the coordinate points of the first parking space feature data in the first parking space feature data capacity pool at the preset distance, thereby generating the first parking space feature data set.

[0066] In some embodiments, a first parking space feature data set is generated based on the first parking space feature data, a preset distance and a first parking space feature data capacity pool, including: determining whether there is first target parking space feature data within a circle with the first parking space feature data as the center and the preset distance as the radius, wherein the first target parking space feature data is any other first parking space feature data existing in the first parking space feature data capacity pool; if the first target parking space feature data exists within the circle, generating the first parking space feature data set based on the first parking space feature data and the first target parking space feature data.

[0067] Optionally, the first parking space feature data capacity pool usually includes multiple data points of the first parking space feature data. The first parking space feature data is placed in the first parking space feature data capacity pool and mapped to the plane coordinate system to compare the distance with other points. Specifically, in the plane coordinate system, the coordinate point corresponding to the first parking space feature data is used as the center and the preset distance is used as the radius of the circle. The coordinate points corresponding to other first parking space feature data in the first parking space feature data capacity pool within this circle are searched. These coordinates are any other first parking space feature data existing in the first parking space feature data capacity pool, which are the first target parking space feature data.

[0068] Optionally, if the first target parking space feature data can be found within a circle with the first parking space feature data as the center and a preset distance as the radius, a first parking space feature data set is generated according to the first parking space feature data and the first target parking space feature data.

[0069] In some embodiments, generating a first parking space feature data set based on the first parking space feature data and the first target parking space feature data includes: obtaining virtual first parking space feature data between the first parking space feature data and the first target parking space feature data, and storing the virtual first parking space feature data into the first parking space feature data set until the first parking space feature data set is filled.

[0070] Specifically, in the plane coordinate system, a first coordinate point of the current first parking space feature data and a second coordinate point corresponding to the first target parking space feature data are obtained. Then, a virtual coordinate point is randomly selected between the first coordinate point and the second coordinate point, and the corresponding virtual first parking space feature data is generated using the virtual coordinate point. Since there is at least one first target parking space feature data, at least one virtual first parking space feature data can be generated. All generated virtual first parking space feature data are then placed into the first parking space feature data set.

[0071] At this time, the first parking space feature data set includes not only the first parking space feature data, but also the virtual first parking space feature data. Finally, all the data in the first parking space feature data set is stored in the first parking space feature data capacity pool.

[0072] In view of the data characteristics of this application scenario, this application improves the SMOTE algorithm to ensure the randomness of the data and avoid one-sided offset. The number of times a neighbor point is used does not exceed 50%. After a part of the virtual coordinate points are generated, the neighbor points need to be replaced. Repeat the above steps until a sufficient number of new virtual coordinate points are generated. Optionally, the preset distance can use commonly used mathematical distances such as Euclidean distance, Manhattan distance, inner product distance and KL distance. Other distances in this article can use these distance units. As long as the distance units remain unified, there is no restriction here.

[0073] In some embodiments, when the first target parking space feature data does not exist within the circle, or when the first target parking space feature data exists within the circle and the amount of virtual first parking space feature data is less than the difference between the preset storage amount and the first storage amount: The first parking space characteristic data is copied and stored in the first parking space characteristic data set until the first parking space characteristic data set is filled.

[0074] Optionally, if the first target parking space feature data does not exist within the circle, it means that there is no other key frame data in the first parking space feature data capacity pool. At this time, the first parking space feature data can be copied multiple times, and then the copied multiple data are stored in the first parking space feature data set.

[0075] Optionally, if the first target parking space feature data exists in the circle, and the number of virtual first parking space feature data is less than the difference between the preset storage capacity and the first storage capacity, it means that there are other key frame data in the first parking space feature data capacity pool, but it is not enough to generate enough virtual first parking space feature data. Therefore, the data missing in the first parking space feature data set is copied to the first parking space feature data to make up the remainder.

[0076] In short, the data in the first parking space feature data set is sufficient to be placed in the first parking space feature data capacity pool to generate a basic data volume that satisfies clustering calculations.

[0077] The above method can fill the capacity of the first parking space feature data capacity pool through the first parking space feature data, and then generate the parking space feature data cluster corresponding to the first parking space feature data through the first parking space feature data capacity.

[0078] In some embodiments, updating the first parking space feature data capacity pool according to the first parking space feature data includes: A first parking space feature data cluster is generated based on the first parking space feature data capacity pool, and the parking space feature data cluster includes a cluster center; a second distance between the first parking space feature data and the cluster center, the first parking space feature data farthest from the cluster center in the first parking space feature data capacity pool, and a third distance between the cluster center and the third parking space feature data are obtained; when the third distance is greater than the second distance, the first parking space feature data is stored in the first parking space feature data capacity pool, and the first parking space feature data farthest from the cluster center is deleted.

[0079] When the first storage capacity is equal to the preset storage capacity, that is, the capacity of the first parking space feature data capacity pool has been filled, it is necessary to update the first parking space feature data capacity pool through the first parking space feature data in the current frame time series parking space data to make the data in the first parking space feature data capacity pool closer to accurate and more real data.

[0080] A first parking space feature data cluster is generated based on the first parking space feature data capacity pool. The parking space feature data cluster includes a cluster center, and there may be multiple cluster centers.

[0081] Specifically, the data in the parking space feature data pool is clustered using the k-means clustering algorithm (Kmeans), resulting in k clusters, each with its own cluster center. Clustering algorithms are unsupervised learning methods that do not require predefined class labels. Instead, they automatically divide data into clusters based on its inherent characteristics. Data points within each cluster have similar characteristics, while data points between different clusters have significant differences. The goal of clustering algorithms is to minimize intra-cluster variation (cohesion) and maximize inter-cluster variation (separation).

[0082] Taking the first parking space characteristic data as the center point of the parking space gear lever as an example, Figure 3 This is a scatter plot of the coordinates of the center point of the wheel stop collected in the first parking space feature data capacity pool. The set capacity of the first parking space feature data capacity pool is 100, which means there are 100 coordinate points in total. Generally, the frame time series parking space data is one point, which includes key frame data (that is, data with high observation values). The data enhancement SMOTE algorithm is used to generate some virtual coordinate points. Figure 2 As shown, the implementation process mainly includes the following steps:

[0083] S201, Initialization: Randomly select K data points as initial cluster centers. The specific value of K needs to be determined according to the distribution characteristics of the actual data to achieve a better clustering result (assuming K=3).

[0084] S202. Assignment: Calculate the distance between each data point and each cluster center, and assign it to the cluster with the closest cluster center.

[0085] S203. Update: For each cluster, calculate the mean of all data points belonging to the cluster as the new cluster center.

[0086] S204, Iteration: Repeat steps 2 and 3 until the position of the cluster center no longer changes significantly.

[0087] At this time, when the first parking space feature data in the current frame time series parking space data is transmitted to the parking space detection device, any cluster center among the initial cluster centers is selected as the first parking space feature data cluster cluster including the cluster center, and the second distance between the first parking space feature data and the cluster center, the first parking space feature data farthest from the cluster center in the first parking space feature data capacity pool, and the third distance between the cluster center and the first parking space feature data farthest from the cluster center are calculated. If the third distance is greater than or equal to the second distance, the first parking space feature data is stored in the first parking space feature data capacity pool, and the first parking space feature data farthest from the cluster center is deleted; if the third distance is less than the second distance, the original data in the first parking space feature data capacity pool is retained.

[0088] For other types of parking space feature data in the current frame time series parking space data, the above method is also used to traverse other types of parking space feature data to obtain parking space feature data, update the parking space feature data capacity pool, and generate parking space feature data clusters corresponding to other types of parking space feature data.

[0089] S103: Determine target parking space feature data according to the parking space data of the current time sequence frame and the parking space feature data cluster.

[0090] The purpose of the parking space detection module generating parking space feature data clusters is to provide accurate parking space data for the parking space posture calculation module of the automatic parking system. Therefore, it is necessary to combine the current time frame parking space data and the parking space feature data clusters to determine the target parking space feature data, and send the target parking space feature data to the parking space posture calculation module for parking space posture calculation.

[0091] In some embodiments, the parking space feature data cluster includes a target cluster center point, and the target cluster center point is the cluster center point with the most parking space feature data in the parking space feature data cluster; Determining target parking space feature data based on the parking space data of the current time sequence frame and the parking space feature data cluster, including: determining whether the parking space data of the current time sequence frame meets the confidence requirement based on the parking space data of the current time sequence frame and the target cluster center point; When the parking space data of the current time series frame meets the confidence requirement, the current time series frame data is set as the target parking space feature data; When the parking space data of the current time series frame does not meet the confidence requirement, the parking space feature data corresponding to the target cluster center point is set as the target parking space feature data.

[0092] According to the method mentioned above, all parking space feature data in the multi-frame time series parking space data are generated into parking space feature data clusters corresponding to different parking space feature data according to their types. When the parking space detection device receives the current frame time series data, it performs cluster calculation on the parking space feature data clusters corresponding to each type of parking space feature data. The parking space feature data clusters are updated and confidence determined based on the parking space feature data existing in the current frame time series data. The target parking space feature data is determined based on the confidence of the current time series frame parking space data and the parking space feature data clusters. Finally, the target parking space feature data is sent to the parking space posture calculation module in the automatic parking system.

[0093] The confidence level indicates whether the various types of parking space feature data in the current time series frame parking space data are credible.

[0094] Optionally, the parking space feature data cluster includes a target cluster center point, and the target cluster center point is a cluster center point with the most parking space feature data in the parking space feature data cluster.

[0095] Specifically, such as Figure 4 As shown in the figure, the data in the parking space feature data capacity pool is clustered by using the k-means clustering algorithm (Kmeans) clustering algorithm to obtain k clusters. Each cluster has its own cluster center (K1, K2, K3 in the figure). The cluster center point with the most parking space feature data among the k clusters is determined as the target cluster center point (such as K3), and the confidence of the parking space feature data is judged by the target cluster center point.

[0096] When the parking space data of the current time series frame meets the confidence requirement, the current time series frame data is set as the target parking space feature data; when the parking space data of the current time series frame does not meet the confidence requirement, the parking space feature data corresponding to the target cluster center point is set as the target parking space feature data.

[0097] The determined target parking space feature data is then sent to the parking space posture calculation module for parking space posture calculation.

[0098] In some embodiments, whether the parking space data of the current time frame meets the confidence requirement is determined based on the parking space data of the current time frame and the target cluster center point, including: obtaining the fourth distance between the parking space data of the current time frame and the target cluster center point; when the fourth distance is less than or equal to the set distance threshold, the parking space data of the current time frame meets the confidence requirement; when the fourth distance is greater than the set distance threshold, the parking space data of the current time frame does not meet the confidence requirement.

[0099] Specifically, after obtaining each type of parking space feature data in the current frame time series parking space data, the fourth distance between the parking space feature data and the target cluster center in the corresponding parking space feature data cluster cluster is calculated. When the fourth distance is less than or equal to the set distance threshold, it is determined that the parking space feature data meets the confidence requirement; when the fourth distance is greater than the set distance threshold, the parking space data of the current time series frame does not meet the confidence requirement.

[0100] Taking the first parking space feature data in the parking space data of the current time series frame as an example, the data in the first parking space feature data capacity pool is clustered by using the k-means clustering algorithm (Kmeans) clustering algorithm to obtain three clusters, each cluster having its own cluster center. The cluster center point with the most first parking space feature data in the three clusters is determined as the target cluster center point, and the confidence of the first parking space feature data is calculated using the target cluster center point. When the first parking space feature data in the current frame image is received, the distance between the coordinate point of the first parking space feature data and the target cluster center point is calculated as a fourth distance. When the fourth distance is less than a set distance threshold, it is determined that the first parking space feature data meets the confidence requirement, and the first feature data is set as the target parking space feature data. When the fourth distance is greater than or equal to the set distance threshold, the first parking space feature data does not meet the confidence requirement, and the parking space feature data corresponding to the target cluster center point is set as the target parking space feature data.

[0101] Accordingly, the embodiment of the present application also provides a parking space detection device, such as Figure 6 As shown, the parking space detection device includes: The data receiving module 601 is used to receive multiple frames of time-series parking space data, wherein the multiple frames of time-series parking space data include the current frame of time-series parking space data; A data processing module 602, connected to the data receiving module 601, is configured to generate parking space feature data clusters based on the multiple frames of time-series parking space data; The data confirmation module 603 is connected to the data processing module 602 and is used to determine the target parking space feature data according to the parking space data of the current time sequence frame and the parking space feature data cluster.

[0102] Specifically, the parking space detection device may also perform other methods in the parking space detection method, such as: Receive multiple frames of time-series parking space data, wherein the multiple frames of time-series parking space data include current frame time-series parking space data; generate parking space feature data clusters according to the multiple frames of time-series parking space data; and determine target parking space feature data according to the current frame time-series parking space data and the parking space feature data clusters.

[0103] In some embodiments, the parking space data includes parking space feature data, and the parking space feature data includes at least one of parking space corner point coordinate data, parking space direction data, and object type coordinate point data in the parking space.

[0104] In some embodiments, generating a cluster of parking space feature data based on the multiple frames of time-series parking space data includes: Acquire first parking space feature data, where the first parking space feature data is the parking space feature data in the current time series frame data; update the first parking space feature data capacity pool according to the first parking space feature data; and generate a first parking space feature data cluster based on the first parking space feature data capacity pool, where the parking space feature data cluster includes the first parking space feature data cluster.

[0105] In some embodiments, the first parking space feature data capacity pool includes a first storage capacity and a preset storage capacity, the first storage capacity being the storage quantity of the first parking space feature data currently stored in the first parking space feature data capacity pool, and the preset storage capacity being the maximum storage capacity of the first parking space feature data capacity pool; Updating the first parking space feature data capacity pool according to the first parking space feature data includes: When the first storage capacity is less than the preset storage capacity, the first parking space feature data is stored in the first parking space feature data capacity pool; when the first storage capacity is greater than or equal to the preset storage capacity, the first parking space feature data capacity pool is updated according to the first parking space feature data.

[0106] In some embodiments, storing the first parking space feature data into the first parking space feature data capacity pool includes: determining whether the first parking space feature data is key frame data; In a case where the first parking space feature data is not the key frame data, storing the first parking space feature data into the first parking space feature data capacity pool; In the case that the first parking space feature data is the key frame data, the first parking space feature data is enhanced to generate a first parking space feature data set, and the first parking space feature data set is stored in the first parking space feature data capacity pool, wherein the first parking space feature data set includes the first parking space feature data, and the amount of data in the first parking space feature data set is equal to the difference between the preset storage capacity and the first storage capacity.

[0107] In some embodiments, the enhancing the first parking space feature data to generate a first parking space feature data set includes: generating the first parking space feature data set according to the first parking space feature data, a preset distance, and a first parking space feature data capacity pool.

[0108] In some embodiments, generating the first parking space feature data set according to the first parking space feature data, a preset distance, and a first parking space feature data capacity pool includes: Determining whether first target parking space characteristic data exists within a circle with the first parking space characteristic data as the center and the preset distance as the radius, wherein the first target parking space characteristic data is any other first parking space characteristic data existing in the first parking space characteristic data capacity pool; In a case where the first target parking space feature data exists within the circle, the first parking space feature data set is generated according to the first parking space feature data and the first target parking space feature data.

[0109] In some embodiments, generating the first parking space feature data set according to the first parking space feature data and the first target parking space feature data includes: A virtual first parking space feature data is obtained between the first parking space feature data and the first target parking space feature data, and the virtual first parking space feature data is stored in the first parking space feature data set until the first parking space feature data set is filled.

[0110] In some embodiments, the method further comprises: When the first target parking space feature data does not exist within the circle, or when the first target parking space feature data exists within the circle and the amount of the virtual first parking space feature data is less than the difference between the preset storage capacity and the first storage capacity: copy the first parking space feature data and store it in the first parking space feature data set until the first parking space feature data set is filled.

[0111] In some embodiments, updating the first parking space feature data capacity pool according to the first parking space feature data includes: Generate the first parking space feature data cluster based on the first parking space feature data capacity pool, and the parking space feature data cluster includes a cluster center; obtain the second distance between the first parking space feature data and the cluster center, the first parking space feature data farthest from the cluster center in the first parking space feature data capacity pool, and the third distance between the cluster center and the third parking space feature data; when the third distance is greater than the second distance, store the first parking space feature data in the first parking space feature data capacity pool, and delete the first parking space feature data farthest from the cluster center.

[0112] In some embodiments, the parking space feature data cluster includes a target cluster center point, and the target cluster center point is the cluster center point with the most parking space feature data in the parking space feature data cluster. Determining the target parking space feature data based on the current time series frame parking space data and the parking space feature data cluster includes: Determine whether the parking space data of the current time frame meets the confidence requirement based on the parking space data of the current time frame and the target cluster center point; if the parking space data of the current time frame meets the confidence requirement, set the current time frame data as the target parking space feature data; if the parking space data of the current time frame does not meet the confidence requirement, set the parking space feature data corresponding to the target cluster center point as the target parking space feature data.

[0113] In some embodiments, determining whether the parking space data of the current time sequence frame meets the confidence requirement based on the parking space data of the current time sequence frame and the target cluster center point includes: Obtain a fourth distance between the parking space data of the current time sequence frame and the target cluster center point; when the fourth distance is less than or equal to the set distance threshold, the parking space data of the current time sequence frame meets the confidence requirement; when the fourth distance is greater than the set distance threshold, the parking space data of the current time sequence frame does not meet the confidence requirement.

[0114] The specific execution steps and detailed logical relationships are like the steps and logic in the parking space detection method.

[0115] The specific implementation of the above operations can be found in the previous embodiments and will not be described in detail here.

[0116] As can be seen from the above, this application determines the target parking space feature data based on the current time frame parking space data and parking space feature data clustering cluster through the data confirmation module 603, thereby improving the data accuracy of parking space detection during automatic parking and solving problems such as unnecessary parking and off-center parking caused by parking space data errors.

[0117] The present application also provides an automatic parking system. Figure 7 As shown, the automatic parking system includes: The parking control module 704 and the parking space detection device 702 are connected in communication with each other. The parking space detection device 702 is used to detect parking space data to determine target parking space feature data; The parking control module 704 is used to control the vehicle to achieve automatic parking according to the target parking space characteristic data.

[0118] Specifically, the parking space detection device 702 determines the target parking space characteristic data of the parking space according to the parking space data by executing the above-mentioned parking space detection method, and sends the target parking space characteristic data to the parking control module 704; the parking control module 704 controls the vehicle to realize automatic parking based on the target parking space characteristic data.

[0119] In some embodiments, the automatic parking system includes a parking space data acquisition module 701 , which is connected to the parking space detection device 702 and is configured to acquire the parking space data through a visual sensor.

[0120] Optionally, the automatic parking system includes a parking space data acquisition module 701, which includes a visual sensor and a parking space data processing unit. The visual sensor is used to capture or record images or videos of the parking space and surrounding environment during the parking process. The captured or recorded images or videos are sent to the parking space data processing unit, which extracts parking space features from the data. The sent parking space feature data includes various types of parking space-related feature information, such as parking space corner point data, parking space orientation data (e.g., angle values), and coordinate data of relevant objects within the parking space (e.g., limit blocks, gear levers, obstacles within the parking space, ground locks, etc.). The extracted parking space feature data is processed for missing values and outliers, and then a determination is made as to whether the frame data is a key frame. The various types of parking space feature data and information such as whether the frame data is a key frame are sent to the parking space detection device 702.

[0121] The visual sensor may be one or more of a camera, a streaming media, or other visual sensors, and is not limited here.

[0122] In some embodiments, the automatic parking system further includes a target position pose calculation module 703, which is connected to the parking space detection device 702 and is configured to receive the target position feature data sent by the parking space detection device 702 and calculate the parking space position pose according to the target position feature data; Specifically, to achieve automatic parking control of the vehicle, it is also necessary to calculate and analyze the parking position of the target feature data determined by the parking detection device 702. Therefore, after the parking detection device 702 determines the target feature data, it sends the target feature data to the target position calculation module 703. The target position calculation module 703 analyzes and calculates the parking position through the target feature data and historical data, thereby constantly updating more accurate parking features, improving the accuracy of the parking position during the automatic parking process, and solving problems such as unnecessary parking and parking deviation caused by parking data errors. In some embodiments, the target position calculation module 703 is connected to the parking control module 704, and the parking control module is also used to control the vehicle to automatically park according to the parking position.

[0123] The parking space posture data calculated by the target posture calculation module 70 is sent to the parking control module 704 for vehicle parking guidance data path planning, thereby realizing automatic parking of the vehicle.

[0124] Depending on the architecture of different automatic parking systems, the target feature data determined by the parking space detection device 702 of this method can be adapted to different parking control modules to perform parking control in various scenarios, such as: The target feature data is the high-confidence parking space boundary coordinates and dimensions, which can reduce the posture calculation deviation caused by data errors, improve the accuracy of the parking path, filter interference through the credibility data, and ensure the accuracy of the safety boundary verification (such as avoiding scratches caused by misjudgment of parking space size); if the target feature data is obstacle data, it can eliminate false parking space occupancy information caused by environmental interference (such as shadows misjudged as obstacles), avoid false triggering of emergency braking, and verify whether the posture path is within the safety boundary (such as the distance to the obstacle is ≥5cm, and the steering wheel angle does not exceed the mechanical limit); the high-confidence target feature data can also be synchronized to the global map (such as updating the parking space occupancy status, correcting the parking space size deviation), and marking low-confidence areas for manual review, thereby outputting a high-precision parking lot map database.

[0125] The specific implementation of the above operations can be found in the previous embodiments and will not be described in detail here.

[0126] From the above, it can be seen that this application determines the target parking space feature data based on the current time frame parking space data and parking space feature data clustering cluster through the automatic parking system, thereby improving the data accuracy of parking space detection during the automatic parking process, and solving problems such as unnecessary parking and off-center parking caused by parking space data errors, thereby improving the accuracy of the parking path and outputting a high-precision parking lot map database.

[0127] The present application also provides a controller, such as Figure 8 , which shows a schematic diagram of the structure of the controller involved in the embodiment of the present application, specifically:

[0128] The controller may include one or more processors 1001 of processing cores, one or more computer-readable storage media memories 1002, a power supply 1003, an input unit 1004 and other components. Those skilled in the art will appreciate that Figure 6The controller structure shown in does not constitute a limitation of the controller, and may include more or fewer components than shown in the figure, or a combination of certain components, or a different arrangement of components. Among them: the processor 1001 is the control center of the controller, which uses various interfaces and lines to connect the various parts of the entire controller, and executes various functions of the controller and processes data by running or executing software programs and / or modules stored in the memory 1002, and calling data stored in the memory 1002, thereby monitoring the controller as a whole. Optionally, the processor 1001 may include one or more processing cores; preferably, the processor 1001 may integrate an application processor and a modem processor, wherein the application processor mainly processes the operating system, user interface and computer programs, etc., and the modem processor mainly processes wireless communications. It can be understood that the above-mentioned modem processor may not be integrated into the processor 1001.

[0129] Memory 1002 can be used to store software programs and modules. Processor 1001 executes various functional applications and data processing by running the software programs and modules stored in memory 1002. Memory 1002 may primarily include a program storage area and a data storage area. The program storage area may store an operating system and computer programs required for at least one function (such as sound playback or image playback); the data storage area may store data generated based on the use of the controller. Furthermore, memory 1002 may include high-speed random access memory and non-volatile memory, such as at least one disk storage device, flash memory device, or other volatile solid-state storage device. Accordingly, memory 1002 may also include a memory controller to provide processor 1001 with access to memory 1002.

[0130] The controller also includes a power supply 1003 for supplying power to various components. Preferably, the power supply 1003 can be logically connected to the processor 1001 via a power management system, thereby enabling the power management system to manage charging, discharging, and power consumption. The power supply 1003 can also include one or more DC or AC power supplies, a recharging system, a power failure detection circuit, a power converter or inverter, a power status indicator, and other arbitrary components.

[0131] The controller may further include an input unit 1004, which may be configured to receive input digital or character information and generate keyboard, mouse, joystick, optical or trackball signal inputs related to user settings and function control.

[0132] Although not shown, the controller may further include a display unit, etc., which will not be described in detail here. Specifically, in this embodiment, the processor 1001 in the controller loads the executable files corresponding to one or more computer program processes into the memory 1002 according to the following instructions, and the processor 1001 runs the computer program stored in the memory 1002 to implement various functions, including: receiving multiple frames of time-series parking space data, the multiple frames of time-series parking space data including the current frame of time-series parking space data; generating a cluster of parking space feature data based on the multiple frames of time-series parking space data; and determining target parking space feature data based on the current frame of time-series parking space data and the cluster of parking space feature data.

[0133] This application generates parking space feature data clusters from multiple frames of time-series parking space data, and combines the current time-series frame parking space data with the parking space feature data clusters to determine target parking space feature data. This improves the data accuracy of parking space detection during automatic parking, and resolves issues such as unnecessary parking and misaligned parking caused by parking space data errors. The specific implementation of each of the above operations can be found in the previous examples and will not be repeated here.

[0134] The present application also provides a vehicle including a controller or an automatic parking system. The vehicle can execute the parking methods provided in the various optional implementations of the above embodiments via the controller or the automatic parking system. The vehicle can be a gasoline-powered vehicle, a plug-in hybrid vehicle, or a new energy vehicle, and this application does not specifically limit this.

[0135] The present application also provides a computer program product or computer program, which includes computer instructions stored in a computer-readable storage medium. A vehicle processor reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the vehicle to perform the parking space detection method provided in various optional implementations of the above embodiments.

[0136] Those skilled in the art will appreciate that all or part of the steps in the various methods of the above embodiments may be accomplished by a computer program, or by controlling related hardware through a computer program. The computer program may be stored in a computer-readable storage medium and loaded and executed by a processor.

[0137] According to one aspect of the present application, a computer-readable storage medium is provided, on which instructions are stored. When the instructions are executed by a processor, the processor is configured to execute the above-mentioned parking space detection method.

[0138] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, compact disc read-only memory (CD-ROM), optical storage, etc.) containing computer-usable program code.

[0139] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems) and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0140] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0141] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 The steps for the function specified in one or more boxes.

[0142] In a typical configuration, a computing device includes one or more processors (Central Processing Unit, CPU), input / output interfaces, network interfaces, and memory.

[0143] Memory may include non-permanent memory in a computer-readable medium, random access memory (RAM), and / or non-volatile memory, such as read-only memory (ROM) or flash random access memory (flash RAM). Memory is an example of a computer-readable medium.

[0144] Computer-readable media include permanent and non-permanent, removable and non-removable media, and can be implemented by any method or technology for information storage. Information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random-access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassettes, magnetic tape, disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable media does not include transitory media such as modulated communication signals and carrier waves.

[0145] In the description of this application, the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more features. In the description of this application, "plurality" means two or more, unless otherwise specifically defined.

[0146] In the above embodiments, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0147] The embodiments, implementation methods and related technical features of the present application can be combined and replaced with each other without conflict.

[0148] The above are merely preferred embodiments of the present application and do not constitute any form of limitation to the present application. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present application without departing from the content of the technical solution of the present application are still within the scope of the technical solution of the present application.

Claims

1. A parking space detection method, characterized in that: include: receiving a plurality of frames of time-series parking space data, wherein the plurality of frames of time-series parking space data includes a current frame of time-series parking space data; Generate a parking space feature data cluster according to the multiple frames of time-series parking space data; Target parking space feature data is determined according to the current frame time series parking space data and the parking space feature data cluster.

2. The method according to claim 1, characterized in that The parking space data includes parking space feature data, and the parking space feature data includes at least one of parking space corner point coordinate data, parking space direction data, and parking space object type coordinate point data.

3. The method according to claim 2, characterized in that Generating a parking space feature data cluster according to the multiple frames of time-series parking space data includes: Acquire first parking space feature data, where the first parking space feature data is the parking space feature data in the current time series frame data; Updating the first parking space feature data capacity pool according to the first parking space feature data; A first parking space feature data cluster is generated based on the first parking space feature data capacity pool, and the first parking space feature data cluster includes the first parking space feature data cluster.

4. The method according to claim 3, characterized in that The first parking space feature data capacity pool includes a first storage capacity and a preset storage capacity, the first storage capacity is the storage quantity of the first parking space feature data currently stored in the first parking space feature data capacity pool, and the preset storage capacity is the maximum storage capacity of the first parking space feature data capacity pool; Updating the first parking space feature data capacity pool according to the first parking space feature data includes: When the first storage capacity is less than the preset storage capacity, storing the first parking space characteristic data into the first parking space characteristic data capacity pool; In a case where the first storage capacity is greater than or equal to the preset storage capacity, the first parking space feature data capacity pool is updated according to the first parking space feature data.

5. The method according to claim 4, characterized in that The storing the first parking space feature data into the first parking space feature data capacity pool includes: Determining whether the first parking space feature data is key frame data; In a case where the first parking space feature data is not the key frame data, storing the first parking space feature data into the first parking space feature data capacity pool; In the case that the first parking space feature data is the key frame data, the first parking space feature data is enhanced to generate a first parking space feature data set, and the first parking space feature data set is stored in the first parking space feature data capacity pool, wherein the first parking space feature data set includes the first parking space feature data, and the amount of data in the first parking space feature data set is equal to the difference between the preset storage capacity and the first storage capacity.

6. The method according to claim 5, characterized in that The performing enhancement processing on the first parking space feature data to generate a first parking space feature dataset includes: The first parking space feature data set is generated according to the first parking space feature data, a preset distance, and a first parking space feature data capacity pool.

7. The method according to claim 6, characterized in that The generating the first parking space feature data set according to the first parking space feature data, the preset distance, and the first parking space feature data capacity pool includes: Determining whether first target parking space characteristic data exists within a circle with the first parking space characteristic data as the center and the preset distance as the radius, wherein the first target parking space characteristic data is any other first parking space characteristic data existing in the first parking space characteristic data capacity pool; In a case where the first target parking space feature data exists within the circle, the first parking space feature data set is generated according to the first parking space feature data and the first target parking space feature data.

8. The method according to claim 7, characterized in that The generating the first parking space feature data set according to the first parking space feature data and the first target parking space feature data includes: A virtual first parking space feature data is obtained between the first parking space feature data and the first target parking space feature data, and the virtual first parking space feature data is stored in the first parking space feature data set until the first parking space feature data set is filled.

9. The method according to claim 8, characterized in that The method further comprises: In a case where the first target parking space characteristic data does not exist within the circle, or in a case where the first target parking space characteristic data exists within the circle and the amount of the virtual first parking space characteristic data is less than the difference between the preset storage amount and the first storage amount: The first parking space feature data is copied and stored in the first parking space feature data set until the first parking space feature data set is filled.

10. The method according to claim 4, characterized in that Updating the first parking space feature data capacity pool according to the first parking space feature data includes: generating the first parking space feature data cluster based on the first parking space feature data capacity pool, wherein the parking space feature data cluster includes a cluster center; Obtaining a second distance between the first parking space feature data and the cluster center, the first parking space feature data in the first parking space feature data capacity pool that is farthest from the cluster center, and a third distance between the cluster center and the third parking space feature data; When the third distance is greater than the second distance, the first parking space feature data is stored in the first parking space feature data capacity pool, and the first parking space feature data farthest from the cluster center is deleted.

11. The method according to any one of claims 2 to 10, characterized in that The parking space feature data cluster includes a target cluster center point, which is a cluster center point with the most parking space feature data in the parking space feature data cluster. The determining of the target parking space feature data based on the current time series frame parking space data and the parking space feature data cluster includes: Determining whether the parking space data in the current time sequence frame meets the confidence requirement according to the parking space data in the current time sequence frame and the target cluster center point; If the parking space data of the current time sequence frame meets the confidence requirement, setting the current time sequence frame data as the target parking space feature data; When the parking space data of the current time sequence frame does not meet the confidence requirement, the parking space feature data corresponding to the target cluster center point is set as the target parking space feature data.

12. The method according to claim 11, characterized in that The determining, based on the parking space data of the current time sequence frame and the target cluster center point, whether the parking space data of the current time sequence frame meets the confidence requirement includes: Obtaining a fourth distance between the parking space data of the current time series frame and the target cluster center point; When the fourth distance is less than or equal to the set distance threshold, the parking space data of the current time sequence frame meets the confidence requirement; When the fourth distance is greater than the set distance threshold, the parking space data of the current time sequence frame does not meet the confidence requirement.

13. A parking space detection device, characterized in that: include: A data receiving module, the data receiving module is used to receive multiple frames of time-series parking space data, the multiple frames of time-series parking space data including the current frame of time-series parking space data; A data processing module, connected to the data receiving module, configured to generate a parking space feature data cluster based on the multiple frames of time-series parking space data; A data confirmation module is connected to the data processing module and is used to determine target parking space feature data based on the current time sequence frame parking space data and the parking space feature data cluster.

14. An automatic parking system, characterized in that: include: a parking control module and the parking space detection device according to claim 13, wherein the parking control module and the parking space detection device are communicatively connected; The parking space detection device is used to detect parking space data to determine target parking space characteristic data; The parking control module is used to control the vehicle to achieve automatic parking according to the target parking space characteristic data.

15. The automatic parking system according to claim 14, characterized in that: The automatic parking system further includes a parking space data acquisition module, which is connected to the parking space detection device and is used to obtain the parking space data through a visual sensor.

16. The automatic parking system according to claim 14, characterized in that: The system further comprises a target position and posture calculation module, the target position and posture calculation module being connected to the parking space detection device and configured to receive the target position feature data sent by the parking space detection device and perform parking space position and posture calculation based on the target position feature data; The target posture calculation module is connected to the parking control module, and the parking control module is further used to control the vehicle to achieve automatic parking according to the parking posture calculation result.

17. A controller, characterized in that: It comprises a memory and a processor; the memory stores a computer program, and the processor is used to run the computer program in the memory to execute the parking space detection method according to any one of claims 1 to 12.

18. A vehicle, characterized in that: The controller according to claim 17 or the automatic parking system according to any one of claims 14 to 16 may be included.

19. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the parking space detection method according to any one of claims 1 to 12 is implemented.

20. A computer program product, characterized in that The method comprises a computer program or an instruction, which, when executed by a processor, implements the parking space detection method according to any one of claims 1 to 12.

Images