Parking path planning method and device, electronic equipment and storage medium

Abstract

The embodiment of the invention discloses a parking path planning method and device, electronic equipment and a storage medium, and the method comprises the steps: collecting first point cloud data of a current parking space through a vehicle-mounted radar in a process that a vehicle is parked in the current parking space; the first point cloud data is processed, a first polygon corresponding to the current parking space is determined, and the first polygon is used for indicating the contour shape of the current parking space; according to a preset polygon corresponding to the preset parking space and a first polygon corresponding to the current parking space, determining a rotation angle interval meeting a preset rotation condition; wherein the preset rotation condition comprises that the preset polygon is located in the first polygon under the condition that the geometric center of the preset polygon is overlapped with the geometric center of the first polygon and the preset polygon is rotated based on any rotation angle in the rotation angle interval; and planning a parking path according to the rotation angle interval and pre-acquired orientation data and attitude data of the vehicle.

Description

Technical Field

[0001] This application belongs to the field of autonomous driving technology, and particularly relates to a parking path planning method, device, electronic device, and storage medium. Background Art

[0002] The ultrasonic radar measures the distance by emitting ultrasonic waves and receiving the reflected waves, but its detection depth is limited, and there are scanning blind spots at far points and occluded locations. For non-regular-shaped parking spaces, such as diagonal parking spaces and L-shaped parking spaces, the shapes of non-regular-shaped parking spaces are complex and the boundaries are blurred, making it difficult to accurately identify them through ultrasonic radars.

[0003] Currently, there are scanning blind spots at far points and occluded locations for ultrasonic radars, resulting in inaccurate initial recognized parking space information and being unable to accurately reflect the true parking space attitude. Due to the scanning blind spots, the initially recognized parking space can only be roughly used as a basis for releasing available parking spaces, and it is difficult to achieve precise parking into non-regular-shaped parking spaces. Ultrasonic radars cannot provide high-precision positioning information, resulting in difficult attitude adjustment during the parking process, and it is easy to stop or fail to park.

[0004] Due to the scanning blind spots and detection depth limitations of ultrasonic radars, it is difficult to achieve precise parking into non-regular-shaped parking spaces. Summary of the Invention

[0005] Embodiments of this application provide a parking path planning method, device, equipment, and storage medium, which can solve the problem that it is currently difficult to achieve precise parking of a vehicle into a non-regular-shaped parking space.

[0006] In a first aspect, embodiments of this application provide a parking path planning method, which includes:

[0007] During the process of a vehicle parking into the current parking space, collect first point cloud data of the current parking space through an in-vehicle radar;

[0008] Process the first point cloud data to determine a first polygon corresponding to the current parking space, where the first polygon is used to indicate the contour shape of the current parking space;

[0009] According to a preset polygon corresponding to a preset parking space and the first polygon corresponding to the current parking space, determine a rotation angle interval that meets the preset rotation condition; where the preset rotation condition includes: when the geometric center of the preset polygon overlaps with the geometric center of the first polygon, and when the preset polygon is rotated based on any rotation angle in the rotation angle interval, the preset polygon is located within the first polygon;

[0010] Plan a parking path according to the rotation angle interval, the previously obtained orientation data and attitude data of the vehicle.

[0011] In a second aspect, an embodiment of the present application provides a parking path planning device, which includes:

[0012] An acquisition module, configured to obtain first point cloud data of the current parking space through an in-vehicle radar during the process of the vehicle parking into the current parking space;

[0013] A processing module, configured to process the first point cloud data to determine a first polygon corresponding to the current parking space, and the first polygon is used to indicate the contour shape of the current parking space;

[0014] A determination module, configured to determine a rotation angle interval that meets a preset rotation condition according to a preset polygon corresponding to a preset parking space and the first polygon corresponding to the current parking space; wherein, the preset rotation condition includes: when the geometric center of the preset polygon overlaps with the geometric center of the first polygon, and the preset polygon is rotated based on any rotation angle in the rotation angle interval, the preset polygon is located within the first polygon;

[0015] A planning module, configured to plan a parking path according to the rotation angle interval, the pre-acquired orientation data and attitude data of the vehicle.

[0016] In a third aspect, an embodiment of the present application provides an electronic device, which includes: a processor and a memory storing computer program instructions; when the processor executes the computer program instructions, the method in the first aspect or any possible implementation manner of the first aspect is implemented.

[0017] In a fourth aspect, an embodiment of the present application provides a readable storage medium, on which computer program instructions are stored, and when the computer program instructions are executed by a processor, the method in the first aspect or any possible implementation manner of the first aspect is implemented.

[0018] In the embodiments of the present application, during the process of the vehicle parking into the current parking space, the on-vehicle radar is used to collect the first point cloud data of the current parking space, and the environmental information of the current parking space is obtained to provide a data basis for subsequent processing. The first point cloud data is processed to determine the first polygon corresponding to the current parking space, and the first polygon is used to indicate the contour shape of the current parking space. By processing the first point cloud data, the contour shape of the current parking space is obtained, which is convenient for subsequent comparison and analysis. According to the preset polygon corresponding to the preset parking space and the first polygon corresponding to the current parking space, the rotation angle interval that meets the preset rotation condition is determined; wherein, the preset rotation condition includes: when the geometric center of the preset polygon overlaps with the geometric center of the first polygon, and the preset polygon is rotated based on any rotation angle in the rotation angle interval, the preset polygon is completely located within the first polygon, which can ensure the angle range within which the vehicle can rotate safely during parking and avoid collisions. According to the rotation angle interval, the previously obtained vehicle orientation data and attitude data, the parking path is planned. Thus, a safe parking path can be generated according to the actual situation of the current parking space to guide the vehicle to complete the parking operation, which is convenient for the vehicle to accurately park into a parking space with an irregular shape. Description of the Drawings

[0019] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings required to be used in the embodiments of the present application. For those of ordinary skill in the art, other drawings can also be obtained based on these drawings without creative efforts.

[0020] Figure 1a It is a schematic diagram of a parking-in scenario provided by an embodiment of the present application;

[0021] Figure 1b It is a schematic diagram of another parking-in scenario provided by an embodiment of the present application;

[0022] Figure 1c It is a schematic diagram of yet another parking-in scenario provided by an embodiment of the present application;

[0023] Figure 1d It is a schematic diagram of still another parking-in scenario provided by an embodiment of the present application;

[0024] Figure 2 It is a flowchart of a parking path planning method provided by an embodiment of the present application;

[0025] Figure 3 It is a schematic structural diagram of a parking path planning device provided by an embodiment of the present application;

[0026] Figure 4 It is a schematic hardware structure diagram of an electronic device provided by an embodiment of the present application. Detailed implementation manners

[0027] The features and exemplary embodiments of various aspects of the present application will be described in detail below. In order to make the objectives, technical solutions and advantages of the present application more clear and understandable, the present application will be further described in detail below in combination with the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only configured to explain the present application and are not configured to limit the present application. For those skilled in the art, the present application can be implemented without some of these specific details. The following description of the embodiments is only intended to provide a better understanding of the present application by showing examples of the present application.

[0028] It should be noted that, in this document, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not expressly listed, or further includes elements inherent to such process, method, article or device. Without further limitation, elements defined by the statement "including..." do not exclude the existence of additional identical elements in the process, method, article or device including the elements.

[0029] The parking path planning method provided by the embodiments of the present application can be applied to at least the following application scenarios, which will be described below.

[0030] As Figure 1a shown, the released parking space accuracy is greatly affected by the vehicle speed when the vehicle itself detects the parking space. When the vehicle speed of the vehicle itself changes greatly or is relatively high, the parking space is not accurate enough, which affects the parking efficiency.

[0031] Specifically, the response time of the ultrasonic sensor is relatively long. When the vehicle speed is relatively high, the sensor may not be able to capture the change of the parking space in time, resulting in inaccurate detection results. Although these sensors have a relatively fast response speed, when driving at a high speed, the change speed of the parking space will also increase, which may cause the detected parking space boundary to be blurred or inaccurate.

[0032] When the vehicle speed is relatively high, the dynamic environment around the vehicle changes relatively fast, such as the movement of other vehicles, the walking of pedestrians, etc. These will all affect the accuracy of parking space detection. When the vehicle speed changes greatly, the movement state of the vehicle is unstable, and the data acquisition of the sensor will also be affected, further reducing the detection accuracy. When the vehicle speed is relatively high, the amount of data that the system needs to process increases, and the calculation and processing delay may cause the parking space detection result to lag, affecting the parking efficiency.

[0033] AsFigure 1b As shown in the figure, the detection depth of ultrasonic sensors for parking spaces is limited. The common detection depth is 3 - 4 meters, while the depth of vertical parking spaces usually reaches more than 5 meters. There is a large detection error in the middle and rear sections of the parking space, and parking spaces with insufficient space for parking may be released.

[0034] Specifically, due to the physical characteristics of ultrasonic sensors, the detection depth of ultrasonic sensors is limited. Ultrasonic sensors measure the distance by emitting ultrasonic waves and receiving reflected waves. Due to the propagation speed and attenuation characteristics of ultrasonic waves, the detection depth of the sensor is usually about 3 - 4 meters. Resolution limitation of ultrasonic sensors: The resolution of ultrasonic sensors decreases at long distances, resulting in reduced detection accuracy.

[0035] The mismatch between the depth of the parking space and the detection depth. The depth of vertical parking spaces usually exceeds 5 meters, far beyond the detection range of ultrasonic sensors. Due to the limited detection depth, ultrasonic sensors cannot detect the middle and rear sections of the parking space, forming a detection blind area. Ultrasonic sensors can accurately detect the front section of the parking space, such as 3 - 4 meters. For the middle and rear sections of the parking space, such as more than 4 meters, ultrasonic sensors cannot provide accurate data, resulting in detection errors. Due to the detection errors in the middle and rear sections, the system may misjudge the size and shape of the parking space and release parking spaces with insufficient space. When a vehicle attempts to park in these incorrectly released parking spaces, it may find that there is insufficient space, resulting in parking failure or the need for readjustment.

[0036] Such as Figure 1c As shown in the figure, due to inaccurate recognition of the side of obstacles when searching for parking spaces, the accuracy of the parking space orientation is not high when the parking space is detected. The parking attitude is restricted and it is easy to stop due to obstacles during the parking process, resulting in parking failure.

[0037] Specifically, sensors such as ultrasonic sensors, cameras, or lidar may have problems with insufficient accuracy in recognizing the side of obstacles, especially in complex environments such as light changes and reflection interference. The dynamic environment around the vehicle, such as other vehicles, pedestrians, and the movement of obstacles, may cause instability in sensor data, affecting the recognition accuracy of the side of obstacles.

[0038] Due to inaccurate recognition of the side of obstacles, the boundary of the parking space may be unclear, resulting in the system being unable to accurately judge the orientation of the parking space. The misjudgment of the shape of the parking space will also affect the accuracy of the parking space orientation. For example, an irregular parking space may be misjudged as a regular parking space.

[0039] Inaccurate parking space orientation can limit the vehicle's attitude during parking, preventing it from entering the parking space at the optimal angle. Due to inaccurate parking space orientation, the vehicle may need to adjust its attitude multiple times to enter the parking space, increasing the parking time and complexity. During the parking process, if the system fails to accurately identify obstacles, it may misjudge the distance when approaching the obstacles, resulting in stopping or colliding with the obstacles. Due to inaccurate obstacle recognition and inaccurate parking space orientation, the vehicle may not be able to smoothly enter the parking space during the parking process, leading to parking failure.

[0040] As Figure 1d shown, it is difficult for irregular space parking spaces to be parked in the correct orientation or for invalid parking.

[0041] Specifically, irregular space parking spaces usually have irregular shapes, such as diagonal parking spaces, L-shaped parking spaces, etc. These shapes are difficult to accurately identify through traditional regular parking space detection algorithms. The boundaries of irregular parking spaces may be unclear, making it difficult for the system to accurately determine the actual size and shape of the parking space. The orientation of irregular parking spaces may be unclear, and it is difficult for the system to determine at what angle the vehicle should enter the parking space. Irregular parking spaces may need to be detected from multiple directions, increasing the complexity and uncertainty of detection.

[0042] The path planning difficulty for irregular parking spaces increases, and the system needs to consider more constraints, such as obstacles and parking space boundaries, to ensure that the vehicle can smoothly enter the parking space. Due to the complex shape of the parking space, the vehicle may need to adjust its attitude multiple times to enter the parking space, increasing the parking time and complexity.

[0043] Irregular parking spaces may cause difficulties in attitude adjustment during the parking process, preventing the vehicle from entering the parking space at the optimal angle. Due to the complex shape of the parking space, the vehicle is more likely to collide with obstacles during the parking process, increasing the risk of stopping. Due to the complexity of irregular parking spaces, the system may not be able to accurately detect the parking space, resulting in invalid parking. Even if the parking space is detected, due to the difficulties in path planning and attitude adjustment, the vehicle may not be able to smoothly enter the parking space, leading to parking failure.

[0044] Figure 2 is a flowchart of a parking path planning method provided by an embodiment of the present application.

[0045] As Figure 2 shown, the parking path planning method may include step 210 - step 240. This method is applied to a parking path planning device and is specifically as follows:

[0046] Step 210, during the process of the vehicle parking into the current parking space, collect the first point cloud data of the current parking space through an in-vehicle radar;

[0047] Point cloud data: It is a set of points in three-dimensional space collected by sensors such as lidar, radar, or cameras. Each point usually contains its coordinates (X, Y, Z) in space. It is used to construct a three-dimensional model of the environment to help the vehicle perceive the surrounding environment.

[0048] The on-vehicle radar, such as lidar or millimeter-wave radar, scans the current parking space and its surrounding environment to generate point cloud data. The point cloud data contains the three-dimensional coordinate information of the parking space and the surrounding objects. Obtain the three-dimensional environment information of the current parking space to provide a data basis for subsequent processing.

[0049] Among them, during the process of the vehicle parking into the current parking space, it is an exploratory parking into the current parking space.

[0050] Exploratory parking: During the parking process, the vehicle continuously moves and adjusts its posture, and at the same time uses ultrasonic detectors to detect the internal space of the parking space in real time. The detected internal space data of the parking space is stored for subsequent analysis and adjustment.

[0051] Due to the scanning blind area and detection depth limitation of the ultrasonic radar, the initially recognized parking space information may be inaccurate and there is uncertainty. During the parking process, the specific internal space layout and obstacle positions in the parking space may be unknown and need to be obtained through dynamic detection.

[0052] During the parking process, the vehicle uses ultrasonic detectors to detect the internal space of the parking space in real time to obtain the specific internal space data of the parking space. The detected internal space data of the parking space is stored for subsequent analysis and adjustment. According to these data, the vehicle posture and path planning can be adjusted in real time to adapt to the specific situation inside the parking space.

[0053] Through dynamic detection and data storage, the vehicle posture and path planning can be adjusted in real time to improve the parking accuracy and success rate. This method is especially suitable for parking spaces with irregular shapes. Through dynamic detection and data analysis, it can better adapt to the specific situation inside the parking space and achieve precise parking.

[0054] During the parking process, dynamically detect and utilize the internal space of the parking space, store and analyze the detection data to improve the parking accuracy and success rate. This method is especially suitable for situations where the initial parking space recognition is inaccurate and the internal space of the parking space is unknown. Through real-time adjustment and data analysis, it can better meet the parking requirements of irregular parking spaces.

[0055] Step 220, process the first point cloud data to determine a first polygon corresponding to the current parking space, and the first polygon is used to indicate the contour shape of the current parking space;

[0056] Polygon: A closed figure formed by connecting multiple line segments. In computer vision and autonomous driving, polygons are often used to represent the contours of objects or the boundaries of regions, and are used to describe the shapes and positions of objects such as parking spaces and obstacles.

[0057] Process the collected point cloud data, extract the contour information of the parking space, and represent it as a polygon. The processing process may include operations such as point cloud filtering, segmentation, and clustering. Obtain the contour shape of the current parking space for subsequent comparison and analysis.

[0058] Step 230: Determine the rotation angle interval that meets the preset rotation conditions according to the preset polygon corresponding to the preset parking space and the first polygon corresponding to the current parking space; wherein, the preset rotation conditions include: when the geometric centers of the preset polygon and the first polygon overlap, and when the preset polygon is rotated based on any rotation angle in the rotation angle interval, the preset polygon is entirely within the first polygon.

[0059] Geometric center: Also known as the centroid or center of gravity, it is the average position of all points within the polygon. For a simple polygon, the geometric center can be determined by calculating the average coordinates of all vertices. It is used to determine the center position of the polygon for alignment and rotation operations.

[0060] Rotation angle interval: Refers to the range of angles within which a polygon is allowed to rotate under specific conditions. It is used to determine the range of angles within which a vehicle can safely rotate during the parking process.

[0061] Calculate the rotation angle interval that meets the preset rotation conditions according to the preset polygon corresponding to the preset parking space and the first polygon corresponding to the current parking space. The preset rotation conditions require that the geometric centers of the preset polygon and the first polygon overlap, and at any angle within the rotation angle interval, the preset polygon can be completely within the first polygon. Determine the range of angles within which a vehicle can safely rotate during the parking process to avoid collisions.

[0062] Step 240: Plan the parking path according to the rotation angle interval, the previously obtained orientation data and attitude data of the vehicle.

[0063] Parking path: Refers to the path planning for a vehicle to move from its current position to the target parking space. It guides how the vehicle can complete the parking operation safely and efficiently.

[0064] Plan a path from the current position to the target parking space according to the rotation angle interval, the orientation data and attitude data of the vehicle. The path planning may include steps such as path optimization and collision detection to ensure the safety and feasibility of the path. Generate a safe parking path to guide the vehicle to complete the parking operation.

[0065] Specifically, geometric methods, hybrid A-star and other parking path planning methods can be used to plan a parking path according to the rotation angle interval, the orientation data and attitude data of the vehicle obtained in advance.

[0066] The geometric method is a path planning method based on geometric constraints, which mainly generates a path by analyzing the geometric characteristics and kinematic constraints of the vehicle. Consider the geometric characteristics of the vehicle, such as length, width, turning radius, etc. According to the geometric characteristics of the vehicle, a smooth path is generated to ensure that the path meets the constraints such as the turning radius and minimum turning radius of the vehicle. Optimize the generated path to ensure that the path is smooth and conforms to the kinematic constraints of the vehicle. The geometric method is simple, intuitive, easy to implement, has high computational efficiency, and is suitable for real-time path planning.

[0067] Hybrid A-star: It is a path planning method that combines the A* algorithm and vehicle kinematic constraints to generate a smooth and feasible path.

[0068] Use the A* algorithm to perform global path search to generate an initial path. Vehicle kinematic constraints: On the basis of the A* algorithm, consider the kinematic constraints of the vehicle, such as turning radius, minimum turning radius, etc., to refine the path. Smooth the refined path to ensure that the path is smooth and conforms to the kinematic constraints of the vehicle.

[0069] Hybrid A-star has good adaptability in complex environments, such as multi-obstacle and narrow spaces. Good path smoothness: The path generated by hybrid A-star has good smoothness and conforms to the kinematic constraints of the vehicle.

[0070] Through accurate point cloud data acquisition and processing, ensure that the vehicle can accurately identify parking spaces and obstacles during parking to avoid collisions. Through the calculation of the rotation angle interval and path planning, generate an optimal parking path to reduce parking time and operation complexity. Using point cloud data and polygon representation, the vehicle can perceive the surrounding environment more comprehensively, improving the robustness and reliability of the autonomous driving system. Thus, the autonomous driving system can achieve efficient and safe parking operations, enhancing the user experience and driving safety.

[0071] In one possible embodiment, step 220 includes:

[0072] Process the first point cloud data to obtain target point cloud data;

[0073] According to the target point cloud data, determine the first polygon corresponding to the current parking space.

[0074] Target point cloud data: It refers to the point cloud data that has been processed and filtered, removing noise and irrelevant points and retaining the key points related to the parking space. It is used to more accurately determine the contour shape of the parking space, reducing errors and interference.

[0075] Preprocess the collected first point cloud data, including operations such as filtering, denoising, and segmentation, to remove irrelevant points such as the ground, walls, or other vehicles, and retain the points related to the parking space. Obtain cleaner and more accurate target point cloud data, which is convenient for subsequent extraction of the parking space contour.

[0076] Utilize the processed target point cloud data, and through algorithms such as clustering and boundary extraction, determine the contour shape of the parking space and represent it as a polygon. The vertices of the polygon are usually composed of the key points in the target point cloud data. Obtain the accurate contour shape of the current parking space, which is convenient for subsequent comparison and analysis.

[0077] Through the preprocessing steps, remove noise and irrelevant points, improve the quality of the target point cloud data, reduce errors and interference. With high-quality target point cloud data, the contour shape of the parking space can be extracted more accurately, ensuring the accuracy of the polygon. Through the accurate polygon representation, the vehicle can more comprehensively perceive the shape and size of the parking space, improving the accuracy and safety of the parking operation.

[0078] Specifically, a filtering algorithm can be used to remove the noise points in the point cloud data. Remove the ground and other irrelevant points through a denoising algorithm. Use a segmentation algorithm to segment the point cloud data into different regions and retain the regions related to the parking space. The point cloud data after filtering, denoising, and segmentation is the target point cloud data.

[0079] Cluster the target point cloud data to identify the key points of the parking space. Use a boundary extraction algorithm to extract the contour of the parking space from the clustered point cloud data. Connect the extracted contour points into a polygon to represent the shape of the parking space. The first polygon corresponding to the current parking space represents the contour shape of the parking space.

[0080] Through filtering, denoising, and segmentation processing, improve the quality of the target point cloud data, reduce errors and interference. With high-quality target point cloud data, the contour shape of the parking space can be extracted more accurately, ensuring the accuracy of the polygon. Through the accurate polygon representation, the vehicle can more comprehensively perceive the shape and size of the parking space, improving the accuracy and safety of the parking operation.

[0081] Thus, the autonomous driving system can achieve efficient and safe parking operations, enhancing the user experience and driving safety.

[0082] Among them, in the steps of processing the first point cloud data to obtain the target point cloud data as mentioned above, the specific steps may include the following:

[0083] Convert the first point cloud data into the global coordinate system to obtain the second point cloud data, and determine the confidence value of each data point of the second point cloud data;

[0084] Filter out the data points with confidence values greater than the preset confidence threshold from the second point cloud data to obtain the third point cloud data;

[0085] Perform filtering processing on the third point cloud data to obtain the target point cloud data.

[0086] Global coordinate system: A coordinate system that is fixed relative to the vehicle or the environment, usually with the initial position of the vehicle or a certain fixed point as the origin. It is used to unify the data collected by different sensors, facilitating subsequent processing and analysis.

[0087] Confidence value: Represents the reliability and accuracy of each data point in the point cloud data. Usually generated by sensors or data processing algorithms, the higher the value, the higher the reliability of the point. It is used to screen high-quality data points and remove noise and unreliable data.

[0088] Preset confidence threshold: A preset value used to screen data points with confidence values higher than this threshold. It ensures that the screened data points have high reliability and accuracy.

[0089] Filtering processing: Refers to removing noise points and irrelevant points in the point cloud data through algorithms, and retaining high-quality data points. It can improve the quality of the point cloud data, facilitating subsequent analysis and processing.

[0090] Convert the first point cloud data from the sensor coordinate system to the global coordinate system to facilitate unifying the data of different sensors. At the same time, calculate the confidence value of each data point, representing the reliability and accuracy of the point. The second point cloud data in the unified coordinate system can be obtained, along with the confidence value of each data point, facilitating subsequent screening.

[0091] According to the preset confidence threshold, filter out the data points with confidence values higher than this threshold and remove unreliable data points. High-quality third point cloud data can be obtained, reducing the influence of noise and unreliable data.

[0092] Perform filtering processing on the screened third point cloud data to further remove noise and irrelevant points and retain high-quality data points. The final target point cloud data can be obtained, improving the data quality and facilitating subsequent extraction of the parking space contour.

[0093] By converting point cloud data to a global coordinate system, the data from different sensors can be unified, making it easier to process and analyze them later. By filtering and removing noise and unreliable data points through confidence value screening, the quality and reliability of the data can be improved. Using high-quality target point cloud data, the contour shape of the parking space can be extracted more accurately, ensuring the accuracy of the polygon.

[0094] Specifically, multiple frames of first point cloud data are collected, and multiple frames of first point cloud data of obstacles are collected at different time points. The first point cloud data collected in each frame is converted into a global coordinate system, and the second point cloud data in a unified coordinate system can be obtained, ensuring that all data are compared in the same coordinate system.

[0095] Compare the second point cloud data between multiple frames and analyze the reliability of each data point. Through multi-frame comparison, obstacle points that appear consistently in multiple frames can be identified, eliminating noise and accidental errors. According to the detection of obstacle point cloud data in multiple frames, a confidence label is added to each data point. A data point with a high confidence level indicates that the data point appears consistently in multiple frames and has high reliability.

[0096] In a certain time window, i.e., in the sliding time domain, the confidence of each data point is analyzed. Data points with low confidence are eliminated, which may be noise or accidental errors. By eliminating noise points with low confidence, high-quality third point cloud data can be obtained, which improves the reliability and accuracy of the data.

[0097] Use a Gaussian filter to smooth the remaining third point cloud data to reduce noise and mutations. Use a mean square error filter to smooth the third point cloud data to further reduce noise. Through filtering, the third point cloud data is smoothed, noise and mutations are reduced, and the continuity and smoothness of the third point cloud data are improved.

[0098] The filtered third point cloud data sets are connected to form continuous contour lines. By connecting the filtered third point cloud data, a complete obstacle outline is formed, which is convenient for subsequent path planning and parking operations.

[0099] The step of determining the first polygon corresponding to the current parking space according to the target point cloud data may specifically include the following steps: obtaining an initial polygon according to the target point cloud data;

[0100] The initial polygon is subjected to polygon approximation processing to determine a first polygon corresponding to the current parking space.

[0101] Initial polygon: It is a polygon initially generated based on the target point cloud data. It may contain more vertices and details, but it is not necessarily the best or most concise representation. It serves as the starting point of polygon approximation processing to facilitate subsequent optimization and simplification.

[0102] Polygon approximation processing: It refers to simplifying the vertices of the initial polygon through an algorithm to make it closer to the actual contour shape, while reducing the number of vertices and improving the calculation efficiency. Optimize the representation of the polygon to make it more concise and more in line with the actual contour, facilitating subsequent comparison and analysis.

[0103] Utilize the key points in the target point cloud data to generate an initial polygon through a boundary extraction algorithm. The initial polygon may contain a relatively large number of vertices and details, but it is not necessarily the optimal or most concise representation. Obtain the preliminary parking space contour shape as the starting point for polygon approximation processing.

[0104] Through a polygon approximation algorithm, such as the Ramer-Douglas-Peucker algorithm, simplify the vertices of the initial polygon to make it closer to the actual contour shape, while reducing the number of vertices and improving the calculation efficiency. Obtain the optimized first polygon, which is more concise and more in line with the actual contour, facilitating subsequent comparison and analysis.

[0105] Generate an initial polygon through a boundary extraction algorithm to ensure the accuracy of the contour shape. Through polygon approximation processing, simplify the vertices of the polygon to make it more concise and more in line with the actual contour, improving the calculation efficiency. Through an accurate polygon representation, the vehicle can more comprehensively perceive the shape and size of the parking space, improving the accuracy and safety of the parking operation.

[0106] In a possible embodiment, when the geometric center of the preset polygon overlaps with the geometric center of the first polygon, and the preset polygon is rotated based on any rotation angle, if the preset polygon is not located within the first polygon, then output a first prompt message, where the first prompt message is used to prompt the user that they cannot park in the current parking space.

[0107] When the geometric center of the preset polygon overlaps with the geometric center of the first polygon, rotate the preset polygon and check whether the preset polygon can be completely located within the first polygon at any rotation angle. If the preset polygon cannot be completely located within the first polygon at all rotation angles, it is considered that the current parking space cannot accommodate the vehicle. By rotating and comparing the polygons, determine whether the current parking space is suitable for the vehicle to park in.

[0108] If the preset polygon cannot be completely located within the first polygon at all rotation angles, the system will output a first prompt message to prompt the user that they cannot park in the current parking space. The prompt message can be conveyed to the driver through the in-vehicle display screen, voice prompt, or other user interface elements. Prompt the driver in a timely manner that the current parking space is not suitable for parking, avoiding unnecessary attempts and potential dangers.

[0109] By means of precise polygon comparison and rotation angle checking, it is ensured that the vehicle does not attempt to park in an unsuitable parking space, thus avoiding collisions and scratches. Through timely prompt information, the driver can quickly understand the availability of the current parking space and avoid wasting time and energy. It can intelligently judge the suitability of the parking space, help the driver select a suitable parking space, and improve the parking efficiency.

[0110] In a possible embodiment, after step 240, the following steps may further be included:

[0111] Judge whether the parking path meets the parking conditions; the parking conditions are used to indicate that the rear axle of the vehicle drives into the current parking space;

[0112] If the parking path does not meet the parking conditions, output a second prompt message, and the second prompt message is used to prompt the user that the vehicle cannot be parked in the current parking space.

[0113] Parking path: It refers to the path planning for the vehicle to move from the current position to the target parking space. It guides how the vehicle can complete the parking operation safely and efficiently.

[0114] Parking conditions: It refers to various conditions and parameters used to judge whether the vehicle can drive into the current parking space safely and accurately. It ensures that the vehicle will not collide or exceed the boundary of the parking space during the parking process.

[0115] Second prompt message: It refers to the prompt message output by the system when the parking path does not meet the parking conditions, and is used to inform the user that the vehicle cannot be parked in the current parking space. It timely notifies the driver that the current parking path is not feasible, avoiding unnecessary attempts and potential dangers.

[0116] After planning the parking path, the system will check this path to judge whether it meets the preset parking conditions. The parking conditions usually include the safety of the path, whether the rear axle of the vehicle can smoothly drive into the parking space, etc. It ensures that the planned parking path is safe and feasible, and avoids the vehicle colliding or exceeding the boundary of the parking space during the parking process.

[0117] If the planned parking path does not meet the parking conditions, the system will output a second prompt message to prompt the user that the vehicle cannot be parked in the current parking space. The prompt message can be conveyed to the driver through the in-vehicle display screen, voice prompt or other user interface elements. It timely notifies the driver that the current parking path is not feasible, avoiding unnecessary attempts and potential dangers.

[0118] By checking whether the parking path meets the parking conditions, it is ensured that the vehicle will not collide or exceed the boundary of the parking space during the parking process. Through timely prompt information, the driver can quickly understand the availability of the current parking path and avoid wasting time and energy. It can intelligently judge the feasibility of the parking path, help the driver select a suitable parking path, and improve the parking efficiency.

[0119] By checking whether the parking path meets the parking conditions, it is ensured that the vehicle will not collide or exceed the boundary of the parking space during parking. Through timely prompt information, the driver can quickly understand the availability of the current parking path and avoid wasting time and energy. The system can intelligently judge the feasibility of the parking path, help the driver select a suitable parking path, and improve the parking efficiency. Thus, the autonomous driving system can achieve efficient and safe parking operations, enhancing the user experience and driving safety.

[0120] In a possible embodiment, after step 240, the following steps may further be included:

[0121] Control the steering and speed of the vehicle, and drive the vehicle into the current parking space along the parking path;

[0122] During parking, when a parking obstacle object is detected, collect the point cloud data of the parking obstacle object;

[0123] According to the point cloud data of the parking obstacle object, determine the second polygon of the current parking space;

[0124] According to the preset polygon corresponding to the preset parking space and the second polygon, re-plan the parking path.

[0125] Parking obstacle object: It refers to an obstacle detected during parking, such as other vehicles, pedestrians, bicycles, roadblocks, etc. It is used to detect and identify possible obstacles during parking to ensure the safety of parking operations.

[0126] Second polygon: It refers to a polygon regenerated according to the point cloud data of the detected parking obstacle object during parking, representing the contour shape of the current parking space. It is used to re-plan the parking path to ensure that the vehicle can safely avoid obstacles and drive into the parking space.

[0127] According to the planned parking path, control the steering and speed of the vehicle, so that the vehicle moves along the path and gradually drives into the current parking space. Realize the automatic parking operation of the vehicle, and improve the parking efficiency and accuracy.

[0128] During the process of the vehicle moving along the parking path, the surrounding environment is detected in real time through in-vehicle sensors, especially possible parking obstacle objects, and their point cloud data is collected. Real-time obtain the position and shape information of the parking obstacle object, providing data support for subsequent obstacle processing.

[0129] Using the collected point cloud data of the obstacle object, a second polygon is generated through a boundary extraction algorithm to represent the contour shape of the current parking space, taking into account the influence of the obstacle object. The updated contour shape of the parking space is obtained, which is convenient for re-planning the parking path. According to the updated second polygon and the preset polygon, the rotation angle interval that meets the preset rotation condition is recalculated, and according to the rotation angle interval, the vehicle's orientation data and attitude data, the parking path is re-planned. A new parking path is generated to avoid the obstacle object and ensure that the vehicle can safely drive into the parking space.

[0130] By collecting the point cloud data of the obstacle object in real time, obstacles during the parking process can be detected and processed in a timely manner, improving parking safety. According to the detected obstacle object in real time, the parking path is dynamically adjusted to ensure that the vehicle can safely avoid the obstacle and drive into the parking space. Through the accurate polygon representation, the vehicle can more comprehensively perceive the shape and size of the parking space, improving the accuracy and safety of parking operations.

[0131] In the embodiment of the present application, during the process of the vehicle parking into the current parking space, the first point cloud data of the current parking space is collected by an on-vehicle radar, and the environmental information of the current parking space is obtained to provide a data basis for subsequent processing. The first point cloud data is processed to determine the first polygon corresponding to the current parking space. The first polygon is used to indicate the contour shape of the current parking space. By processing the first point cloud data, the contour shape of the current parking space is obtained, which is convenient for subsequent comparison and analysis. According to the preset polygon corresponding to the preset parking space and the first polygon corresponding to the current parking space, the rotation angle interval that meets the preset rotation condition is determined; wherein, the preset rotation condition includes: when the geometric center of the preset polygon overlaps with the geometric center of the first polygon, and when the preset polygon is rotated based on any rotation angle in the rotation angle interval, the preset polygon is completely located within the first polygon, which can ensure the angle range within which the vehicle can safely rotate during parking and avoid collisions. According to the rotation angle interval, the previously obtained vehicle orientation data and attitude data, the parking path is planned. Thus, according to the actual situation of the current parking space, a safe parking path can be generated to guide the vehicle to complete the parking operation, which is convenient for the vehicle to accurately park into a parking space with an irregular shape.

[0132] Based on the above parking path planning method shown in FIG. 1, the embodiment of the present application further provides a parking path planning device, as Figure 3 shown, the parking path planning device 300 may include:

[0133] A collection module 310, configured to obtain the first point cloud data of the current parking space by an on-vehicle radar during the process of the vehicle parking into the current parking space;

[0134] A processing module 320, configured to process the first point cloud data to determine a first polygon corresponding to the current parking space, where the first polygon is used to indicate the contour shape of the current parking space;

[0135] A determination module 330, configured to determine a rotation angle range that meets a preset rotation condition according to a preset polygon corresponding to a preset parking space and the first polygon corresponding to the current parking space; where the preset rotation condition includes: when the geometric centers of the preset polygon and the first polygon overlap, and when the preset polygon is rotated based on any rotation angle in the rotation angle range, the preset polygon is located within the first polygon;

[0136] A planning module 340, configured to plan a parking path according to the rotation angle range, the orientation data and the attitude data of the vehicle obtained in advance.

[0137] In a possible embodiment, the parking path planning device 300 may include:

[0138] A first output module, configured to output a first prompt message when the geometric centers of the preset polygon and the first polygon overlap, and when the preset polygon is rotated based on any rotation angle, if the preset polygon is not located within the first polygon, where the first prompt message is used to prompt the user that the vehicle cannot be parked in the current parking space.

[0139] In a possible embodiment, the parking path planning device 300 may include:

[0140] A judgment module, configured to judge whether the parking path meets the parking condition; the parking condition is used to indicate that the rear axle of the vehicle drives into the current parking space;

[0141] A second output module, configured to output a second prompt message if the parking path does not meet the parking condition, where the second prompt message is used to prompt the user that the vehicle cannot be parked in the current parking space.

[0142] In a possible embodiment, the parking path planning device 300 may include:

[0143] A control module, configured to control the steering and speed of the vehicle, and drive the vehicle into the current parking space according to the parking path;

[0144] The acquisition module 310 is further configured to acquire the point cloud data of the obstacle object during parking when an obstacle object is detected.

[0145] The determination module 330 is further configured to determine a second polygon of the current parking space according to the point cloud data of the obstacle stopping object;

[0146] The planning module 340 is further configured to replan a parking path according to the preset polygon corresponding to the preset parking space and the second polygon.

[0147] In a possible embodiment, the processing module 320 is specifically configured to:

[0148] Process the first point cloud data to obtain target point cloud data;

[0149] Determine a first polygon corresponding to the current parking space according to the target point cloud data.

[0150] In a possible embodiment, the processing module 320 is specifically configured to:

[0151] Convert the first point cloud data into a global coordinate system to obtain second point cloud data, and determine a confidence value of each data point of the second point cloud data;

[0152] Filter out data points with a confidence value greater than a preset confidence threshold from the second point cloud data to obtain third point cloud data;

[0153] Perform filtering processing on the third point cloud data to obtain target point cloud data.

[0154] In a possible embodiment, the processing module 320 is specifically configured to:

[0155] Obtain an initial polygon according to the target point cloud data;

[0156] Perform polygon approximation processing on the initial polygon to determine a first polygon corresponding to the current parking space.

[0157] In the embodiments of the present application, during the process of the vehicle parking into the current parking space, the first point cloud data of the current parking space is collected by an in-vehicle radar, and the environmental information of the current parking space is obtained to provide a data basis for subsequent processing. The first point cloud data is processed to determine the first polygon corresponding to the current parking space. The first polygon is used to indicate the contour shape of the current parking space. By processing the first point cloud data, the contour shape of the current parking space is obtained, which is convenient for subsequent comparison and analysis. According to the preset polygon corresponding to the preset parking space and the first polygon corresponding to the current parking space, the rotation angle interval that meets the preset rotation condition is determined; wherein, the preset rotation condition includes: when the geometric center of the preset polygon overlaps with the geometric center of the first polygon, and the preset polygon is rotated based on any rotation angle in the rotation angle interval, the preset polygon is completely located within the first polygon, which can ensure the angle range within which the vehicle can rotate safely during parking and avoid collisions. According to the rotation angle interval, the pre-acquired vehicle orientation data and attitude data, a parking path is planned. Thus, a safe parking path can be generated according to the actual situation of the current parking space to guide the vehicle to complete the parking operation, which is convenient for the vehicle to accurately park into a parking space with an irregular shape.

[0158] Figure 4 FIG. shows a schematic hardware structure diagram of an electronic device provided by an embodiment of the present application.

[0159] The electronic device may include a processor 401 and a memory 402 storing computer program instructions.

[0160] Specifically, the above-mentioned processor 401 may include a central processing unit (CPU), or an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present application.

[0161] The memory 402 may include a mass memory for data or instructions. By way of example and not limitation, the memory 402 may include a hard disk drive (HDD), a floppy disk drive, a flash memory, an optical disk, a magneto-optical disk, a magnetic tape, or a universal serial bus (USB) drive, or a combination of two or more of these. Where appropriate, the memory 402 may include removable or non-removable (or fixed) media. Where appropriate, the memory 402 may be internal or external to the integrated gateway disaster recovery device. In a particular embodiment, the memory 402 is a non-volatile solid-state memory. In a particular embodiment, the memory 402 includes a read-only memory (ROM). Where appropriate, the ROM may be a mask-programmed ROM, a programmable ROM (PROM), an erasable PROM (EPROM), an electrically erasable PROM (EEPROM), an electrically rewritable ROM (EAROM), or a flash memory, or a combination of two or more of these.

[0162] The processor 401 reads and executes the computer program instructions stored in the memory 402 to implement any one of the parking path planning methods in the illustrated embodiments.

[0163] In one example, the electronic device may further include a communication interface 404 and a bus 410. Among them, as Figure 4 shown, the processor 401, the memory 402, and the communication interface 404 are connected through the bus 410 to complete communication with each other.

[0164] The communication interface 404 is mainly used to implement communication between the various modules, devices, units, and / or devices in the embodiments of the present application.

[0165] The bus 410 includes hardware, software, or both, and couples the components of the electronic device to each other. By way of example and not limitation, the bus may include an accelerated graphics port (AGP) or other graphics bus, an enhanced industry standard architecture (EISA) bus, a front-side bus (FSB), a hyperTransport (HT) interconnect, an industry standard architecture (ISA) bus, an InfiniBand interconnect, a low-pin count (LPC) bus, a memory bus, a MicroChannel architecture (MCA) bus, a peripheral component interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a serial advanced technology attachment (SATA) bus, a video electronics standards association local (VLB) bus, or other suitable buses, or a combination of two or more of these. Where appropriate, the bus 410 may include one or more buses. Although the embodiments of the present application describe and illustrate specific buses, the present application contemplates any suitable bus or interconnect.

[0166] The electronic device can execute the parking path planning method in the embodiments of the present application, so as to implement the combination with Figure 2 the described parking path planning method.

[0167] In addition, in combination with the parking path planning method in the above embodiments, the embodiments of the present application can be implemented by providing a computer-readable storage medium. Computer program instructions are stored on the computer-readable storage medium; when the computer program instructions are executed by a processor, the Figure 2 implemented parking path planning method is achieved.

[0168] It should be clear that the present application is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present application is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications, and additions, or change the order between steps after understanding the spirit of the present application.

[0169] The functional blocks shown in the above structural block diagrams can be implemented as hardware, software, firmware, or a combination thereof. When implemented in hardware, it can be, for example, an electronic circuit, an application-specific integrated circuit (ASIC), appropriate firmware, a plug-in, a functional card, and so on. When implemented in software, the elements of the present application are programs or code segments used to perform the required tasks. The program or code segment can be stored in a machine-readable medium or transmitted via a data signal carried in a carrier wave on a transmission medium or a communication link. A "machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical discs, hard disks, fiber optic media, radio frequency (RF) links, and so on. The code segment can be downloaded via a computer network such as the Internet, an intranet, and so on.

[0170] It should also be noted that the exemplary embodiments mentioned in the present application describe some methods or systems based on a series of steps or devices. However, the present application is not limited to the order of the above steps, that is, the steps can be executed in the order mentioned in the embodiments, or different from the order in the embodiments, or several steps can be executed simultaneously.

[0171] As described above, this is only the specific implementation manner of the present application. Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems, modules, and units described above can refer to the corresponding processes in the foregoing method embodiments, and will not be elaborated herein. It should be understood that the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed in the present application can easily think of various equivalent modifications or substitutions, and these modifications or substitutions should all be covered within the protection scope of the present application.

Claims

1. A parking path planning method, characterized in that: The method comprises: When the vehicle is parked in the current parking space, the first point cloud data of the current parking space is collected by the vehicle-mounted radar; Processing the first point cloud data to determine a first polygon corresponding to the current parking space, where the first polygon is used to indicate a contour shape of the current parking space; Determine a rotation angle interval that satisfies a preset rotation condition according to a preset polygon corresponding to the preset parking space and a first polygon corresponding to the current parking space; wherein the preset rotation condition includes: when the geometric center of the preset polygon overlaps with the geometric center of the first polygon, and when the preset polygon is rotated based on any rotation angle in the rotation angle interval, the preset polygon is located within the first polygon; A parking path is planned according to the rotation angle interval and the pre-acquired orientation data and posture data of the vehicle.

2. The method according to claim 1, characterized in that The method further comprises: When the geometric center of the preset polygon overlaps with the geometric center of the first polygon and the preset polygon is rotated based on any rotation angle, if none of the preset polygons are located within the first polygon, a first prompt message is output, and the first prompt message is used to prompt the user that the current parking space cannot be parked.

3. The method according to claim 1, characterized in that After planning a parking path according to the rotation angle interval and the pre-acquired orientation data and posture data of the vehicle, the method further includes: Determining whether the parking path satisfies a parking condition; the parking condition is used to indicate that the rear axle of the vehicle enters the current parking space; If the parking path does not satisfy the parking condition, a second prompt message is output, where the second prompt message is used to prompt the user that the vehicle cannot be parked in the current parking space.

4. The method according to claim 1, characterized in that: After planning a parking path according to the rotation angle interval and the pre-acquired orientation data and posture data of the vehicle, the method further includes: Controlling the steering and speed of the vehicle to drive the vehicle into the current parking space according to the parking path; During the parking process, when an obstacle is detected, point cloud data of the obstacle is collected; Determining a second polygon of the current parking space according to the point cloud data of the obstacle object; The parking path is replanned according to the preset polygon corresponding to the preset parking space and the second polygon.

5. The method according to claim 1, characterized in that The processing of the first point cloud data to determine a first polygon corresponding to the current parking space includes: Processing the first point cloud data to obtain target point cloud data; A first polygon corresponding to the current parking space is determined according to the target point cloud data.

6. The method according to claim 1, characterized in that The step of processing the first point cloud data to obtain target point cloud data includes: Converting the first point cloud data into a global coordinate system to obtain second point cloud data, and determining a confidence value of each data point of the second point cloud data; Filtering out data points having confidence values ​​greater than a preset confidence threshold from the second point cloud data to obtain third point cloud data; The third point cloud data is filtered to obtain target point cloud data.

7. The method according to claim 1, characterized in that Determining the first polygon corresponding to the current parking space according to the target point cloud data includes: Obtaining an initial polygon according to the target point cloud data; The initial polygon is subjected to polygon approximation processing to determine a first polygon corresponding to the current parking space.

8. A parking path planning device, characterized in that: The parking path planning device comprises: The acquisition module is used to acquire the first point cloud data of the current parking space through the vehicle-mounted radar when the vehicle is parked in the current parking space; a processing module, configured to process the first point cloud data to determine a first polygon corresponding to the current parking space, wherein the first polygon is used to indicate a contour shape of the current parking space; A determination module, configured to determine a rotation angle interval that satisfies a preset rotation condition based on a preset polygon corresponding to the preset parking space and a first polygon corresponding to the current parking space; wherein the preset rotation condition includes: when the geometric center of the preset polygon overlaps with the geometric center of the first polygon, and when the preset polygon is rotated based on any rotation angle in the rotation angle interval, the preset polygon is located within the first polygon; A planning module is used to plan a parking path according to the rotation angle interval and the pre-acquired orientation data and posture data of the vehicle.

9. An electronic device, characterized in that: The electronic device comprises: a processor and a memory storing computer program instructions; when the processor executes the computer program instructions, the parking path planning method according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer program instructions, and when the computer program instructions are executed by the processor, the parking path planning method according to any one of claims 1 to 7 is implemented.

Images