Parking space berthability judgment method and vehicle

Abstract

The invention provides a parking space berthability judgment method and a vehicle, and the method comprises the steps: determining the parking space information and obstacle information of a target parking space based on the detection information of a sensor of the vehicle; determining a to-be-observed parking space according to a comparison result of the parking space information of the target parking space and the parking space information of historical parking spaces; if it is determined that the to-be-observed parking space is located in an observation area according to the detection information, grid observation attributes of the to-be-observed parking space are determined based on a preset grid map; and according to the grid observation attribute and the obstacle information of the to-be-observed parking space, determining the available parking state information of the to-be-observed parking space. According to the method, accurate parking space state identification can be realized, and accurate parking space information can be provided for a user.

Description

Technical Field

[0001] This application relates to the field of vehicles, and more particularly to a method for determining parking space availability and a vehicle. Background Technology

[0002] In the field of automated parking technology, parking space status recognition is crucial. Related visual recognition, ultrasonic detection, and multi-sensor fusion solutions all suffer from problems caused by sensor inaccuracy, such as inaccurate obstacle recognition and parking space outline judgment. This makes it difficult to accurately determine the availability of parking spaces and consequently, to provide users with accurate parking space information. Summary of the Invention

[0003] This application discloses a method and vehicle for determining parking space availability, which can solve the problem that inaccurate parking space status identification methods make it difficult to provide users with accurate parking space information.

[0004] This application provides a method for determining the parking availability of a parking space. The method includes: determining the parking space information and obstacle information of a target parking space based on the detection information of a vehicle's sensors; determining a parking space to be observed based on a comparison between the parking space information of the target parking space and the parking space information of historical parking spaces; if the detection information determines that the parking space to be observed is located within an observation area, determining the grid observation attributes of the parking space to be observed based on a preset grid map; and determining the parking availability status information of the parking space to be observed based on the grid observation attributes and the obstacle information of the parking space to be observed.

[0005] In some embodiments of this application, the target parking space represents the parking space currently detected by the sensor; the parking space information of the target parking space includes the parking space number of the target parking space; the historical parking space represents the parking space previously detected by the sensor; the parking space information of the historical parking space includes the parking space number of the historical parking space; determining the parking space to be observed based on the comparison result of the parking space information of the target parking space and the parking space information of the historical parking space includes: comparing the parking space number of the target parking space with the parking space number of the historical parking space; if the parking space number of the target parking space does not match the parking space number of the historical parking space, determining the corresponding target parking space as the parking space to be observed.

[0006] In some embodiments of this application, the parking space information of the historical parking space includes the detection status information and parking availability information of the historical parking space; the method further includes: if the detection status information of the historical parking space indicates that the historical parking space is in an undetectable state, or the position of the wheel chock in the historical parking space does not meet the preset requirements, determining that the historical parking space is in an unparkable state; if the parking availability information of the historical parking space indicates that the historical parking space is not in an unparkable state, using the historical parking space as the target parking space.

[0007] In some embodiments of this application, the sensor includes an ultrasonic sensor and / or a laser sensor, and the method further includes determining whether the parking space to be observed is located within the observation area based on the detection information, including: determining the coordinates of the center point of the entrance of the parking space to be observed in the vehicle coordinate system based on the coordinates of the corner point of the parking space to be observed in the vehicle coordinate system; if the detection information of the ultrasonic sensor indicates that the center point coordinates are within the detection range of the ultrasonic sensor, determining that the parking space to be observed is located within the observation area; or, if the detection information of the laser sensor indicates that the center point coordinates are within the detection range of the laser sensor, determining that the parking space to be observed is located within the observation area.

[0008] In some embodiments of this application, the sensor further includes multiple image sensors. Determining whether the parking space to be observed is located within the observation area based on the detection information further includes: if the detection information from the multiple image sensors indicates that the center point coordinates are within the detection range of the multiple image sensors, determining that the parking space to be observed is located within the observation area; or, if the parking space type of the parking space to be observed is a line parking space, selecting an image sensor combination from the multiple image sensors based on the center point coordinates; determining a fan-shaped observation range based on the image sensor combination; and if the center point coordinates are within the fan-shaped observation range, determining whether the parking space to be observed is located within the observation area.

[0009] In some embodiments of this application, the parking space information of the target parking space includes the coordinates of the corner point of the target parking space in the vehicle coordinate system corresponding to the vehicle; the obstacle information includes at least one of the following: the coordinates of the obstacle in the vehicle coordinate system, and the obstacle type; the grid observation attributes include poor observation, completely empty, completely not empty, and uncertain; the determination of the grid observation attributes of the parking space to be observed based on the preset grid map includes at least one of the following methods: determining the vehicle position of the vehicle in the preset grid map; determining the parking space area of ​​the parking space to be observed in the preset grid map based on the vehicle position and the coordinates of the corner point of the parking space to be observed in the vehicle coordinate system; if the parking space area indicates that any corner point of the parking space to be observed is located outside the preset grid map, determining the grid observation attribute of the parking space to be observed as poor observation; counting the number of grids of each category according to the category label of each grid in the preset grid map, wherein the category label includes empty, obstacle, and unknown; determining the grid observation attribute based on the number of grids of each category and the total number of grids corresponding to the parking space area.

[0010] In some embodiments of this application, determining the grid observation attribute based on the number of grids of each category and the total number of grids corresponding to the parking space area includes: if all grids in the parking space area are of unknown category, determining the grid observation attribute as poor observation; if there are no grids of the obstacle category in the parking space area, determining the grid observation attribute as completely empty; if there are grids of the obstacle category in the parking space area, and the sum of the number of grids of the obstacle category and the unknown category exceeds a preset first proportion threshold, determining the grid observation attribute as completely not empty; if there are grids of the obstacle category in the parking space area, but the sum of the number of grids of the obstacle category and the unknown category exceeds a preset first proportion threshold, determining the grid observation attribute as completely not empty; if there are grids of the obstacle category in the parking space area, but the sum of the number of grids of the obstacle category and the unknown category exceeds a preset first proportion threshold, determining the grid observation attribute as completely not empty. If the sum of the quantities does not exceed the first proportion threshold, the grid observation attribute is determined to be uncertain; if there are no obstacle category grids in the parking space area, and the number of unknown category grids exceeds the preset second proportion threshold, the grid observation attribute is determined to be poor observation; if there are no obstacle category grids in the parking space area, and the number of idle category grids exceeds the first proportion threshold, the grid observation attribute is determined to be completely idle; if there are no obstacle category grids in the parking space area, but the number of idle category grids does not exceed the first proportion threshold, the grid observation attribute is determined to be uncertain.

[0011] In some embodiments of this application, determining the parking availability information of the parking space to be observed based on the grid observation attributes and the obstacle information of the parking space to be observed includes: determining the grid observation attributes of the parking space to be observed multiple times based on the sampling rate of the sensor during the vehicle's operation; determining the minimum distance between the obstacles on the left and right sides of the parking space to the center point of the parking space based on the obstacle information; updating the parking availability count value corresponding to a preset parking availability counter based on the multiple determined grid observation attributes and the minimum distance; determining the parking availability status information of the parking space to be observed and the parking space aisle space based on the parking availability count value and the minimum distance; and determining the parking availability status information of the parking space to be observed based on the parking availability status information of the parking space to be observed and the parking space aisle space.

[0012] In some embodiments of this application, the grid observation attributes include completely idle, completely non-idle, and uncertain. Updating the preset parking counter value based on the repeatedly determined grid observation attributes and the minimum distance includes: if the grid observation attribute is completely idle, increasing the parking counter value by a preset value; if the grid observation attribute is completely non-idle, decreasing the parking counter value by the preset value; if the grid observation attribute is uncertain and the minimum distance is less than a preset distance threshold, decreasing the parking counter value by the preset value; if the grid observation attribute is uncertain and the minimum distance is greater than or equal to the distance threshold, increasing the parking counter value by the preset value.

[0013] In some embodiments of this application, determining the release status information of the parking space and the release status information of the parking aisle space based on the available parking count value and the minimum distance includes: if the available parking count value is greater than a preset counting threshold, and at least one of the minimum distances is greater than a preset distance threshold, determining that the parking space release status information is releaseable; determining the position of the entrance area of ​​the parking space and the centerline position of the parking space according to the coordinates of the corner point of the parking space in the vehicle coordinate system; if the distance between the position of an obstacle in the entrance area and the centerline position is greater than or equal to a preset aisle safety distance threshold, determining that the release status information of the parking aisle space is releaseable.

[0014] In some embodiments of this application, determining the parking availability status of the parking space to be observed based on the releaseable status information of the parking space to be observed and the releaseable status information of the parking space aisle space includes: if the releaseable status information of both the parking space and the parking space aisle space is releaseable, determining that the parking availability status information of the parking space to be observed is parking available.

[0015] In some embodiments of this application, the method further includes: outputting available historical parking space information, responding to a user's selection operation for any available historical parking space, and controlling the vehicle to park in any historical parking space.

[0016] This application also provides a vehicle, the vehicle including a processor and a memory, the processor being used to implement the parking space availability determination method when executing a computer program stored in the memory.

[0017] This application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the parking space availability determination method.

[0018] In the parking space availability determination method provided in this application, the vehicle's sensors acquire details of the currently detected target parking space and obstacle distribution information. By comparing this information with historical parking space information, the method efficiently locates the parking space to be observed, avoiding invalid processing. If the parking space to be observed is within a preset observation area, its grid observation attributes are precisely defined using a grid map, discretizing complex information for easier analysis. Finally, by comprehensively considering grid attributes and obstacle information, the parking space availability status is accurately and reliably determined, effectively overcoming the shortcomings of existing methods such as unreasonable observation areas and insufficient judgment accuracy. This significantly improves the reliability of parking space release determination and its adaptability to different environments. Attached Figure Description

[0019] Figure 1 This is a flowchart of a parking space availability determination method provided in an embodiment of this application.

[0020] Figure 2 This is an example diagram of a vehicle coordinate system provided in an embodiment of this application.

[0021] Figure 3 This is an example diagram of an ultrasonic sensor used to detect parking spaces according to an embodiment of this application.

[0022] Figure 4 This is an example diagram of a laser sensor used to detect parking spaces according to an embodiment of this application.

[0023] Figure 5 This is an example diagram of an image sensor used to detect parking spaces according to an embodiment of this application.

[0024] Figure 6 This is an example diagram illustrating a detailed process of step S104 provided in an embodiment of this application.

[0025] Figure 7 This is a flowchart of a parking space availability determination method provided in another embodiment of this application.

[0026] Figure 8 This is a schematic diagram of the structure of a vehicle provided in one embodiment of this application. Detailed Implementation

[0027] For ease of understanding, some concepts related to the embodiments of this application are illustrated and explained by way of example for reference.

[0028] It should be noted that in this application, "at least one" means one or more, and "more than one" means two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone, where A and B can be singular or plural. The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and drawings of this application are used to distinguish similar objects, not to describe a specific order or sequence.

[0029] In the field of automatic parking technology, parking space status recognition is a key link in improving the efficiency and safety of automatic parking. Its core lies in accurately judging the availability status of vacant parking spaces, thereby providing users with reliable parking space information.

[0030] The relevant technologies mainly include the following parking space status recognition schemes: Visual recognition schemes utilize surround-view cameras to acquire 360° environmental images around the vehicle. After distortion correction and dynamic stitching to form a top-down BEV image, parking space searches are performed. Visual algorithms identify obstacles and distinguish parking space types. However, surround-view cameras are affected by the fisheye ultra-wide-angle effect, resulting in significant edge distortion at the observation angle and distortion of the 3D image, affecting the accuracy of obstacle recognition. Ultrasonic detection schemes target unmarked parking spaces, using ultrasonic sensors to scan and cluster echo feature points from multiple probes to form the parking space outline and detect obstacles. However, ultrasonic sensors are prone to detection errors when encountering irregularly shaped obstacles, complex environments, or unreasonable ultrasonic scanning angles. This leads to inaccurate judgment of the parking space outline and the detection of ultrasonic obstacle points extending outwards, encroaching on available parking space space without being released. Multi-sensor fusion schemes, while integrating visual and ultrasonic sensor data and combining real-time vehicle detection to improve recognition flexibility and update parking space status, are still limited by sensor accuracy issues and cannot completely avoid the aforementioned shortcomings.

[0031] Therefore, related visual recognition, ultrasonic detection, and multi-sensor fusion solutions all suffer from problems such as inaccurate obstacle recognition and parking space contour judgment caused by sensor accuracy, making it difficult to accurately determine the availability of parking spaces and thus difficult to provide users with accurate parking space information.

[0032] To address the aforementioned issues, this application provides a method and vehicle for determining parking space availability. This application utilizes vehicle sensors to acquire details of the currently detected target parking space and obstacle distribution information. By comparing this information with historical parking space information, it efficiently identifies the parking space to be observed, avoiding invalid processing. If the parking space to be observed is within a preset observation area, its grid observation attributes are precisely defined using a grid map, discretizing complex information for easier analysis. Finally, by comprehensively considering grid attributes and obstacle information, the parking space availability status is accurately and reliably determined, effectively overcoming the shortcomings of existing methods such as unreasonable observation areas and insufficient judgment accuracy. This significantly improves the reliability of parking space release judgment and its adaptability to different environments.

[0033] Figure 1 This is a flowchart of a parking space availability determination method provided in an embodiment of this application, applied to vehicles (e.g., Figure 8 In vehicle 120). Depending on different needs, the order of steps in this flowchart can be changed, and some steps can be omitted.

[0034] Step S101: Based on the detection information from the vehicle's sensors, determine the parking space information and obstacle information of the target parking space.

[0035] In some embodiments of this application, the sensor includes at least one combination of the following sensors: multiple image sensors, an image sensor combined with an ultrasonic sensor, or an image sensor combined with a laser sensor. For example, multiple image sensors can acquire image information of the parking space and its surrounding environment from different positions of the vehicle (e.g., front, rear, left, right, etc.), and accurately obtain features such as parking space markings and shape through image stitching and recognition technology. In a combination of an image sensor and an ultrasonic sensor, the image sensor is used to identify the type and general outline of the parking space, while the ultrasonic sensor accurately measures the distance to obstacles by emitting and receiving ultrasonic waves, compensating for the image sensor's deficiency in acquiring depth information. In a combination of an image sensor and a laser sensor, the laser sensor provides high-precision three-dimensional point cloud data, which can more meticulously depict the shape and position of obstacles within the parking space, corroborating the information from the image sensor and improving information accuracy.

[0036] Based on the above embodiments, different combinations of sensors installed in the vehicle can leverage their respective advantages to achieve more comprehensive and accurate information acquisition.

[0037] In some embodiments of this application, the target parking space refers to the parking space currently detected by the sensor. While the vehicle is in motion, the sensor continuously detects the surrounding environment and captures potential parking space information in real time. The parking space information of the target parking space includes at least one of the following: parking space number, coordinates of the parking space corner point in the vehicle's corresponding vehicle coordinate system, and parking space type.

[0038] In one example, the parking space number is used to uniquely identify each parking space, facilitating subsequent management and tracking. The coordinates of the parking space's corner points in the vehicle coordinate system can accurately determine the specific location and spatial range of the parking space around the vehicle. For example, a vehicle coordinate system oxy can be established with the center point of the vehicle's rear axle as the origin o, where the vehicle's forward direction is the positive x-axis, and the leftward direction is the positive y-axis, for example... Figure 2 As shown. A parking space corner point can include multiple corner points. For example, if the parking space is a parallelogram area, the parking space corner point can be the four corner points of the parallelogram area.

[0039] In one example, parking space types include perpendicular, horizontal, and angled, each with different requirements for parking angle and space. For instance, perpendicular parking spaces typically require vehicles to back into them perpendicularly, while horizontal parking spaces are suitable for vehicles to drive into directly. Parking space type helps vehicles determine whether a space is suitable for them based on their size and parking capacity.

[0040] In some embodiments of this application, obstacle information includes at least one of the following: the coordinates of the obstacle in the vehicle coordinate system, and the obstacle type. In one example, during the process of detecting parking spaces, the sensor detects obstacles within and around the parking space that may affect parking. For obstacle points output by upstream modules (such as a surround-view obstacle point detection module), the attributes of each obstacle point include: the distance value to the center point of the rear axle of the vehicle (vehicle coordinate system), and the obstacle type at the current point (such as wheel chocks, vehicles, pedestrians, fixed obstacles, etc.).

[0041] By measuring the distance from an obstacle to the center of the vehicle's rear axle, the relative position of the obstacle to the vehicle can be determined, allowing for the assessment of whether the obstacle is within a parking space and its distance from the vehicle. Obstacle type information helps the vehicle assess the obstacle's impact on the parking process. For example, pedestrians are mobile, requiring more caution when parking, while fixed obstacles are relatively stable, allowing the vehicle to plan a reasonable parking path based on their position and size. Combining this obstacle point information provides a comprehensive and accurate understanding of the obstacle distribution within the target parking space, offering crucial information for subsequently determining the parking space's availability.

[0042] In some embodiments of this application, the vehicle can receive grid map information of a preset grid map output by the upstream module. This information includes map size, resolution (e.g., 10cm), category label of each grid (e.g., free, obstacle, unknown, etc.), and the vehicle's position information in the map.

[0043] In one example, the map size represents the actual area covered by the raster map, which can be set based on a preset area where the vehicle is located. For example, the area formed by extending a certain distance outward from the vehicle (e.g., the origin of the vehicle's coordinate system) ensures that the surrounding environment where the vehicle may drive and park is covered, providing a comprehensive spatial reference for subsequent vehicle route planning and decision-making.

[0044] In one example, resolution refers to the actual physical size represented by each grid cell. A resolution of 10cm means that each grid cell corresponds to a 10cm x 10cm area in the actual environment. Higher resolution allows for a more detailed description of environmental information, much like dividing space with a finer grid, capturing smaller objects and more complex environmental features. For example, smaller obstacles, such as roadside bollards or low stakes, might be ignored or have their location and shape not accurately represented in a low-resolution map. However, in a 10cm resolution grid map, these small obstacles can be clearly divided into the corresponding grid cells, providing a more precise basis for vehicle obstacle avoidance and parking operations.

[0045] In one example, the category label for each grid cell visually reflects its status. The obstacle label explicitly indicates the presence of impassable objects within the grid area. These objects could be fixed buildings, other parked vehicles, or temporary obstacles such as construction barriers. Vehicles must avoid areas represented by these obstacle grids during driving and parking to ensure safety. The free area label indicates that the grid area is freely passable, allowing vehicles to drive or park within the planned route. The unknown area label indicates that the grid area has not yet been detected or the information is unclear; it may be in a blind spot for the vehicle's sensors or a newly appeared area that has not yet been identified. For unknown grids, vehicles need to remain vigilant and conduct further detection upon approach to determine their actual status.

[0046] In one example, the vehicle's location information on the map provides precise positioning within the coordinate system corresponding to a pre-defined grid map. For instance, the coordinate system of the pre-defined grid map can be identical to the vehicle's coordinate system. This information allows the vehicle to clearly understand its specific location within the grid map. Combined with the map size and resolution, the vehicle can accurately determine its relative position to surrounding obstacles. For example, the vehicle can determine how many grid squares away it is from an obstacle ahead, allowing it to adjust its distance based on its speed and braking performance to avoid collisions. Simultaneously, the vehicle's location information is also the foundation for path planning and decision-making. Based on its own position and the target location, the vehicle can search for a safe and feasible path on the grid map.

[0047] In some embodiments of this application, a preset grid map can be updated based on obstacle information during vehicle movement. As the vehicle moves, its various sensors (such as lidar, cameras, and ultrasonic sensors) continuously detect the surrounding environment and acquire obstacle information in real time. This newly acquired obstacle information may differ from the information already present in the preset grid map. For example, areas originally marked as empty squares on the map may suddenly have new obstacles, such as temporarily parked vehicles or fallen objects; or areas originally marked as obstacle squares may have obstacles removed, becoming passable spaces.

[0048] When these situations occur, the vehicle fuses the newly detected obstacle information with a pre-defined grid map. For example, based on sensor data, the vehicle determines the location of the new obstacle in the vehicle coordinate system, then transforms it to the grid map coordinate system to find the corresponding grid cell. For newly appearing obstacles, the grid cell type is updated from empty or unknown cell to obstacle cell; for removed obstacles, the grid cell type is updated from obstacle cell to empty cell. For some boundary areas or areas with unclear information, the vehicle also combines data from multiple sensors and historical information to make comprehensive judgments and reasonable inferences to update the grid map more accurately.

[0049] By dynamically updating the preset grid map based on obstacle information, this method ensures that the grid map remains highly consistent with the actual environment. This provides a more accurate and reliable environmental perception basis for advanced driver assistance functions such as automatic parking, adaptive cruise control, and automatic emergency braking, effectively improving vehicle safety and intelligence in complex environments. For example, during automatic parking, the vehicle can adjust its parking path in real time based on the updated grid map, avoiding newly appearing obstacles and successfully completing the parking operation. During adaptive cruise control, the vehicle can anticipate changes in the road ahead based on updated map information, adjusting its speed and distance accordingly to ensure smooth and safe driving.

[0050] Step S102: Based on the comparison results between the parking space information of the target parking space and the parking space information of historical parking spaces, determine the parking space to be observed.

[0051] In some embodiments of this application, "historical parking spaces" refer to parking spaces that have been previously detected by sensors. For example, historical parking spaces are obtained by the vehicle during its previous driving process, through various onboard sensors detecting the surrounding environment and processing and judging the detected parking space information. For example, parking spaces judged by the parking availability judgment method provided in the embodiments of this application.

[0052] In one example, the parking space information for a historical parking space includes at least one of the following: parking space number, coordinates of the parking space corner point in the vehicle's coordinate system, parking space type, detection status information, and parking availability information. The parking space number uniquely identifies each historical parking space for easy management and tracking; when a vehicle passes by again, historical records can be retrieved for comparison with the current detection status.

[0053] The coordinates of the parking space corner points within the vehicle's corresponding vehicle coordinate system can accurately determine the specific location and spatial range of a historical parking space around the vehicle. These corner point coordinates are key parameters describing the shape and location of the parking space. Based on this coordinate information, the vehicle can intuitively understand its relative position to the parking space, providing fundamental data for subsequent decisions on whether to reconsider the parking space. For example, when a vehicle approaches a historical parking space, the distance and angle between the vehicle and the parking space can be calculated based on the corner point coordinates to assess whether parking is suitable.

[0054] Parking space types include perpendicular, horizontal, and angled. Parking space type helps vehicles determine whether a space is suitable for them based on their size and parking capacity. Vehicles can plan their parking strategies in advance based on historical parking space types, combined with their own parking algorithms and performance.

[0055] The detection status information (det_status) reflects the status of parking spaces during the historical detection process, including detectable (dettrue) and undetectable (det false) states. Detectable indicates that the parking space has been clearly and accurately detected, while undetectable indicates that the parking space is partially obscured, blurry, or otherwise obscured.

[0056] Parking availability status (valid_status) is a comprehensive assessment of whether a historical parking space is suitable for parking. Parking availability status includes "parkable" (e.g., valid_status = 1) and "unparkable" (e.g., valid_status = -1). The parking availability status takes into account factors such as the size and location of the parking space, surrounding obstacles, and the vehicle's own performance. For example, even if a parking space appears large enough to accommodate a vehicle, if there are numerous fixed obstacles around it preventing the vehicle from entering or exiting smoothly, then the parking availability status of that space will be determined as "unparkable." Parking availability status provides crucial reference information for deciding whether to choose a parking space in the current scenario.

[0057] In some embodiments of this application, determining the parking space to be observed based on the comparison results between the parking space information of the target parking space and the parking space information of historical parking spaces includes: comparing the parking space number of the target parking space with the parking space number of historical parking spaces; if the parking space number of the target parking space does not match the parking space number of historical parking spaces, the corresponding target parking space is determined to be the parking space to be observed. For example, if the parking space number of the target parking space is different from the parking space number of historical parking spaces, it indicates that it may be a parking space that has not been detected before, and the corresponding target parking space can be determined to be the parking space to be observed.

[0058] In one example, if the parking space number of the target parking space is the same as the parking space number of a historical parking space, it may be the same parking space that was previously detected. Further information can be combined to determine whether it is the parking space to be observed.

[0059] For example, by checking the detection status and parking availability information of a historical parking space, if the historical detection status is "detectable" and the parking availability information is "parkable," but the corner coordinates of the current target parking space have changed, it may indicate that the parking space situation has changed. Mark the target parking space as a parking space to be observed for detailed detection and confirmation.

[0060] Alternatively, calculate the distance and angle differences between the target parking space corner coordinates and the historical parking space corner coordinates. If the difference is within a preset threshold range, it may be different detection results for the same parking space; if it exceeds the threshold, it may be a newly appeared parking space or a significant change in the original parking space. List the target parking spaces with large coordinate differences as parking spaces to be observed, analyze the reasons for the changes, and determine whether it is due to environmental changes or sensor errors.

[0061] For example, when the target parking space type is different from the historical parking spaces, such as previously being a perpendicular parking space and now being a horizontal parking space, it may be due to sensor misidentification or parking space modification. Target parking spaces with such obvious changes in parking space type are identified as parking spaces to be observed, and their true type is verified through re-detection and manual confirmation.

[0062] Based on the above embodiments, and by combining the comparison results of various information sources, the vehicle can accurately select the parking spaces to be observed. For each parking space, the vehicle can allocate more sensor resources for more precise detection, or reduce its driving speed to observe the surrounding conditions of the parking space more carefully, thereby obtaining more accurate parking space information and providing a reliable basis for subsequent parking decisions.

[0063] In some embodiments of this application, if the detection status information of a historical parking space indicates that the historical parking space is in an undetectable state, or if the position of the wheel chock in the historical parking space does not meet preset requirements, the historical parking space is determined to be in an unparkable state. If the parking availability information of a historical parking space indicates that the historical parking space is not in an unparkable state, the historical parking space is used as the target parking space.

[0064] In one example, a wheel chock position not meeting preset requirements means that when the detection status information is detectable, the wheel chock position makes it inconvenient or impossible for the vehicle to park. For example, a perpendicular parking space typically consists of several key parts, including the parking space entrance, the wheel chock, and the bottom of the parking space. The parking space entrance is the passage for vehicles to enter the parking space; the wheel chock is generally located near the entrance of the parking space to restrict the parking position of vehicles and prevent vehicles from over-driving in; the bottom of the parking space is the area where the vehicle is finally parked and the rear of the vehicle reaches.

[0065] When viewed from the side of the vehicle's position, corresponding to the parking space entrance area, the wheel chock position in this layout meets the preset requirements. In this case, the vehicle can smoothly enter the parking space through the entrance and, following normal driving operations and parking procedures, can be parked in a suitable position without being obstructed by an unreasonably positioned wheel chock.

[0066] If the wheel chock is positioned at the bottom of the parking space when viewed from the side of the vehicle, it means that when the vehicle tries to enter the space, it will be blocked in front of the vehicle's path. Because of this obstruction, the vehicle cannot drive into the parking space normally, making parking impossible. Therefore, the wheel chock's position in this layout does not meet the intended requirements.

[0067] In one example, if a historical parking space is determined to be in an unparkable state (valid_status = -1), then when filtering parking spaces within the observation area to be calculated, these unparkable parking spaces will be skipped and not included in the list of parking spaces in the observation area to be calculated.

[0068] In one example, when a historical parking space is not currently unavailable, it can be selected as the target parking space. During actual vehicle operation, due to the vehicle's motion characteristics, the sensing angle and area of ​​its onboard sensors dynamically change as the vehicle moves. This dynamic change may limit the sensors' perception of the parking space status; for example, at certain times, the sensor may fail to accurately determine the parking space status due to limitations in viewing angle or area. By re-detecting a historical parking space that is not currently unavailable as the target parking space, this deficiency in sensor perception can be effectively compensated for, further verifying the actual parking space availability and ensuring the accuracy and reliability of the parking space availability assessment.

[0069] Step S103: If the parking space to be observed is determined to be within the observation area based on the detection information, the grid observation attributes of the parking space to be observed are determined based on the preset grid map.

[0070] In some embodiments of this application, the detection information from sensors can be used to determine whether the parking space to be observed is located within the observation area. For example, based on the coordinates of the corner points of the parking space to be observed in the vehicle coordinate system, the coordinates of the center point of the entrance to the parking space to be observed in the vehicle coordinate system can be determined. If the detection information from the ultrasonic sensor indicates that the center point coordinates are within the detection range of the ultrasonic sensor, it is determined that the parking space to be observed is located within the observation area. Alternatively, if the detection information from the laser sensor indicates that the center point coordinates are within the detection range of the laser sensor, it is determined that the parking space to be observed is located within the observation area.

[0071] In one example, to determine the coordinates of the center point of the entrance to the parking space to be observed in the vehicle coordinate system, the coordinates of the two corner points of the parking space to be observed in the vehicle coordinate system can be obtained based on real-time vehicle image detection. The coordinates of the center point of the parking space (slot_entrance_center) can then be calculated using the midpoint formula. For example, if the coordinates of corner point A are (x1, y1) and the coordinates of corner point B are (x2, y2), then the coordinates of the center point are (x, y) = ((x1+x2) / 2, (y1+y2) / 2).

[0072] In one example, when the ultrasonic sensor detects that the front of the vehicle has swept past an obstacle in front of the parking space, it verifies whether the coordinates of the center point meet the ultrasonic sensor's detection threshold condition. For example, |x|≤15 and |y|≤5 (in meters). The ultrasonic sensor detection threshold is calibrated based on the effective detection range of the ultrasonic sensor, ensuring that the parking space is within the coverage area of ​​the ultrasonic sensor (e.g., ...). Figure 3 (As shown). If the conditions are met, it is determined to be within the observation area, triggering subsequent raster observation attribute analysis.

[0073] In one example, for a 3D parking space constructed from laser sensor detection information, the sensor's 3D target detection information (such as point cloud data) can be combined to determine whether the center point falls within a threshold range of ±3 meters laterally and ±5 meters longitudinally for the vehicle. This range is comprehensively calibrated using the sensor's field of view, installation position, and actual detection distance to ensure that the parking space is within the sensor's effective detection area. Figure 4 As shown.

[0074] In another example, if the detection information from multiple image sensors indicates that the center point coordinates are within the detection range of the multiple image sensors, it is determined that the parking space to be observed is located within the observation area. For a three-dimensional parking space constructed by multiple image sensors, the sensor's three-dimensional target detection information (such as depth images) can be combined to determine whether the center point falls within the threshold range of ±3 meters laterally and ±5 meters longitudinally for the vehicle.

[0075] In some embodiments of this application, determining whether the parking space to be observed is located within the observation area based on the detection information further includes: if the parking space type is a line parking space, selecting an image sensor combination from multiple image sensors based on the center point coordinates; determining a fan-shaped observation range based on the image sensor combination; and determining whether the parking space to be observed is located within the observation area if the center point coordinates are within the fan-shaped observation range.

[0076] In one example, in an online parking space scenario, the combination of image sensors used for observation can be determined based on the sign of the coordinates (x, y) of the parking space's center point. If x > 0 (the parking space's center point is located in the area in front of the vehicle), an image sensor mounted in front of the vehicle (e.g., a front fisheye camera) is selected. If x < 0 (the parking space's center point is located in the area behind the vehicle), an image sensor mounted behind the vehicle (e.g., a rear fisheye camera) is selected. If y > 0 (the parking space's center point is located in the area to the left of the vehicle), an image sensor mounted on the left side of the vehicle (e.g., a left fisheye camera) is selected. If y < 0 (the parking space's center point is located in the area to the right of the vehicle), an image sensor mounted on the right side of the vehicle (e.g., a right fisheye camera) is selected. Thus, by combining the signs of (x, y), four combinations of image sensors can be selected: front left, front right, rear left, and rear right.

[0077] In one example, the coordinates of the center point of the storage area in the vehicle coordinate system are transformed to the camera coordinate system of any image sensor in the image sensor assembly (e.g., the front fisheye camera), resulting in (pos_in_cam1, pos_in_cam2). During the transformation, parameters such as the installation position, optical center coordinates, and rotation matrix of this image sensor can be considered. Based on the camera intrinsic parameters (focal length, field of view, etc.) and installation height of this image sensor, the boundary angle and distance threshold of the fan-shaped observation area of ​​this image sensor are calculated. For example... Figure 5 As shown, if the center point of the parking space is within the sector of any image sensor in the image sensor combination, the parking space is determined to be within the observation area.

[0078] For example, in the image sensor combination, the front fisheye camera has a horizontal field of view of 180° and a longitudinal detection distance of 0.5 to 10 meters. If the center point of the parking space is within the fan-shaped observation range of the front fisheye camera (angle deviation ≤ ±15°, distance within the effective detection range), the parking space is determined to be within the observation area.

[0079] In one example, the parking space to be observed must meet the observation area determination criteria for at least one of the following scenarios: ultrasonic, visual / laser, or linear parking. For example, an ultrasonic parking space must simultaneously meet the center point coordinate threshold and the effective range of the sensor; a linear parking space must have a fan-shaped area covered by at least one camera combination covering the center point.

[0080] In one example, if the detection information determines that the parking space to be observed is outside the observation area, it is determined not to determine the parking availability status of the parking space. For example, for parking spaces outside the observation area, the process can jump to step S105 to mark them as unparkable, avoiding unnecessary computational resource consumption.

[0081] Based on the above embodiments, by using a multi-sensor fusion strategy and combining different sensor characteristics to customize the observation area determination rules, the accuracy of parking space detection is ensured, and the system's computational efficiency is optimized, providing reliable prerequisites for subsequent grid observation attribute analysis and parking availability determination.

[0082] In some embodiments of this application, the parking space status can be dynamically assessed based on vehicle sensor data and grid map information. The vacancy, occupancy, and unknown status of parking space areas can be quantified using a gridding method, assigning an observation attribute label to each parking space, such as a grid observation attribute. Grid observation attributes include poor observation (observe_bad), completely free (observe_perfect_free), completely not free (observe_perfect_no_free), and uncertain (observe_uncertain). Poor observation indicates that the grid detection information within the parking space area in the grid map is unreliable, such as abnormal detection data, including missing or conflicting data. Completely free indicates that the grids within the parking space area in the grid map are free of obstacles and have no unknown areas. Completely not free indicates that the grids within the parking space area in the grid map are occupied by obstacles or a high proportion of unknown areas. Uncertain indicates that some grids within the parking space area in the grid map are observable, but it is impossible to definitively determine whether they are vacant or occupied.

[0083] In some embodiments of this application, determining the grid observation attributes of the parking space to be observed based on a preset grid map includes: (a) Determine the vehicle position in the preset grid map. Based on the vehicle position and the coordinates of the corner point of the parking space to be observed in the vehicle coordinate system, determine the parking space area of ​​the parking space to be observed in the preset grid map.

[0084] In one example, the coordinates of multiple corner points of a parking space in the vehicle's coordinate system can be projected to the map coordinate system corresponding to a preset raster map through coordinate transformation. Based on the corner point coordinates, it can be determined whether all corner points of the parking space are within the effective range of the raster map (such as within the map boundary).

[0085] (b) If the parking space area indicates that any corner point of the parking space to be observed is outside the preset grid map, the grid observation attribute of the parking space to be observed is determined to be poor observation.

[0086] In one example, if any corner point extends beyond the raster map's boundaries, the raster observation attribute can be determined as "observe_bad". This is because areas outside the map boundaries lack raster data support, making it impossible to accurately determine the parking space status (for example, partial truncation of parking spaces may lead to misjudgments).

[0087] (c) Count the number of grids of each category based on the category label of each grid in the preset grid map.

[0088] In one example, based on the category label of each grid in the preset grid map (e.g., free, obstacle, unknown), the number of grids in each of these three categories is counted. For example, iterating through all grids corresponding to the entire parking area, grids marked "free" are grouped into one category and counted, resulting in the number of grids with the attribute "free" as free_num; grids marked "obstacle" are grouped into another category and counted, resulting in the number of grids with the attribute "obstacle" as obs_num; grids marked "unknown" are grouped into a third category and counted, resulting in the number of grids with the attribute "unknown" as unknown_num, thus obtaining the specific number of grids in each category.

[0089] (d) Determine the grid observation attributes based on the number of grids in each category and the total number of grids corresponding to the parking space area.

[0090] After counting the number of grid cells for each category, the grid observation attributes for that parking space area are determined by combining this with the total number of grid cells corresponding to that parking space area (slot_area_num). For example: If all grids within the parking space area are of unknown category, meaning the number of grids of unknown category equals the total number of grids corresponding to the parking space area, then it is considered that there is no valid observation data for that parking space area, and its status is unreliable. Therefore, the grid observation attribute is set to "poor observation".

[0091] If there are no obstacle category grids within the parking space area (obs_num = 0), meaning the number of obstacle category grids is zero, and the number of free category grids is equal to the total number of grids corresponding to the parking space area, this indicates that the parking space area is completely observable and has no obstacles. In this case, the grid observation attribute is set to "completely free".

[0092] If there are grids of the obstacle category within the parking space area (obs_num != 0), meaning the number of grids of the obstacle category is not zero, and the sum of the number of grids of the obstacle category and the unknown category accounts for more than the total number of grids corresponding to the parking space area, this means that obstacles or unknown areas occupy most of the space in the parking space area, and the parking space is definitely unusable. Therefore, the grid observation attribute is set to "completely not idle".

[0093] If there are grids of the obstacle category in the parking space area, that is, the number of grids of the obstacle category is not zero, but the sum of the number of grids of the obstacle category and the unknown category does not exceed the first proportion threshold, it means that the proportion of obstacles and unknown areas has not reached the level that makes the parking space unusable, and the parking space status is in an ambiguous state. At this time, the grid observation attribute is set to "uncertain (with obstacles)".

[0094] If there are no grids of the obstacle category in the parking space area, that is, the number of grids of the obstacle category is zero, and the proportion of the number of grids of unknown category to the total number of grids corresponding to the parking space area exceeds the preset second proportion threshold (e.g., 80%), the reliability of the parking space status is low due to the high proportion of unknown areas. Therefore, the grid observation attribute is set to "poor observation (no obstacles but high proportion of unknowns)".

[0095] If there are no obstacle category grids in the parking space area, that is, the number of obstacle category grids is zero, and the number of vacant category grids accounts for more than the first proportion threshold of the total number of grids corresponding to the parking space area, this indicates that the vacant area dominates the parking space area, so the grid observation attribute is set to "completely vacant (no obstacles and high vacancy rate)".

[0096] If there are no obstacle category grids in the parking space area, that is, the number of obstacle category grids is zero, but the proportion of the number of vacant category grids to the total number of grids corresponding to the parking space area does not exceed the first proportion threshold, such as 50% or less, that is, the vacant area proportion is insufficient and the parking space status is not clear enough. In this case, the grid observation attribute is set to "uncertain (no obstacles but low vacant proportion)".

[0097] Based on the above embodiments, by combining vehicle sensor data with preset grid map information, the parking space status can be dynamically assessed. A gridding method is used to quantify the vacancy, occupancy, and unknown status of parking space areas, thereby accurately assigning observation attribute labels to each parking space. For example, first, the vehicle's position in the preset grid map is determined, and the area of ​​the parking space to be observed is determined accordingly. If any corner of the parking space is outside the map, it is considered a poor observation. Next, the number of vacant, obstacle, and unknown grids within the parking space area is counted. Based on these numbers and the total number of grids in the parking space area, grid observation attributes are determined according to different rules. In this way, the parking space status can be comprehensively and accurately reflected, providing a reliable basis for autonomous driving parking decisions and effectively improving parking safety and success rate.

[0098] Step S104: Determine the parking availability information of the parking space to be observed based on the grid observation attributes and the obstacle information of the parking space to be observed.

[0099] In some embodiments of this application, a multi-dimensional dynamic evaluation mechanism can be used to combine historical state information with real-time observation data to achieve accurate judgment of parking space availability. In one example, the detailed process of step S104 is as follows: Figure 6 As shown, the steps include the following.

[0100] Step S601: During the vehicle's operation, the grid observation attributes of the parking space to be observed are determined multiple times based on the sensor's sampling rate.

[0101] In one example, the grid observation attributes of the parking space to be observed are determined multiple times based on the sensor's sampling rate. For example, the sampling rate of a laser sensor is 10-20Hz, and the sampling rate of an image sensor is 30fps, allowing multiple independent samplings of the same parking space area during vehicle movement. The time-series data obtained from multiple samplings are aligned, for example, by synchronizing the timestamps of the multi-sensor data using a vehicle kinematics model (such as wheel speedometers) to eliminate measurement bias caused by motion. Specifically, for the sensor data at each sampling time, a grid matching algorithm can be used to map the point cloud or image to a grid map coordinate system, updating the instantaneous category label of each grid. Based on the instantaneous category label of each grid, the grid observation attributes of the parking space to be observed are determined.

[0102] In this way, parking space status information can be obtained more comprehensively and accurately, avoiding misjudgments caused by single sampling errors.

[0103] Step S602: Based on the obstacle information, determine the minimum distance between the obstacles on the left and right sides of the parking space to be observed and the center point of the parking space.

[0104] In one example, based on obstacle information, the minimum distances from obstacles on both sides of the parking space to the center point of the space are determined, such as the minimum distance on the left (left_dis) and the minimum distance on the right (right_dis). These distances are calculated using a grid map or obstacle point set and are used to quantify the available space width on both sides of the parking space. The parking space threshold can be determined based on vehicle dimensions (such as the sum of vehicle width and vehicle safety margin) or parking space design specifications to ensure vehicles can park smoothly.

[0105] Step S603: Update the parking count value corresponding to the preset parking counter based on the grid observation attributes and minimum distance determined multiple times.

[0106] In one example, the available parking counter (map_valid_counter) is updated based on the raster observation attributes and minimum distance, with the available parking counter value ranging from [-release_counter, release_counter].

[0107] In one example, if the grid observation attribute is completely free, the parking availability counter (map_valid_counter) is incremented by a preset value. For example, incrementing the parking availability counter by 1 indicates that the parking space is free of obstacles and has no unknown areas, representing an ideal parking availability state.

[0108] In one example, if the grid observation attribute is "not available at all," the available parking count is reduced by a preset value. For example, reducing the available parking count by 1 indicates that the parking space is occupied by obstacles or a high proportion of unknown areas and is unavailable.

[0109] In one example, if the grid observation attribute is uncertain and the minimum distance is less than a preset distance threshold, the available parking count is decreased by a preset value. If the grid observation attribute is uncertain and the minimum distance is greater than or equal to the distance threshold, the available parking count is increased by a preset value. For example, if left_dis + right_dis < parking space threshold, the available parking count is decreased by 1 (risk of insufficient space); conversely, if left_dis + right_dis ≥ parking space threshold, the available parking count is increased by 1 (sufficient space). In this way, the grid state and obstacle distance can be dynamically combined to improve the accuracy of the judgment.

[0110] Step S604: Based on the parking count and the minimum distance, determine the release state information of the parking space to be observed and the release state information of the parking space aisle space.

[0111] In one example, if the available parking count is greater than a preset counting threshold (e.g., release_counter), and at least one side of the minimum distance is greater than a preset distance threshold, the parking space is determined to be releaseable. For example, when the middle area of ​​the parking space is vacant and there is sufficient space on both sides, the vehicle parking conditions are met.

[0112] In one example, the positions of the parking space entrance area and the centerline of the parking space are determined based on the coordinates of the corner point of the parking space in the vehicle coordinate system. If the distance between the obstacle position in the entrance area and the centerline position is greater than or equal to a preset aisle safety distance threshold, the releasable state information of the parking space aisle space is determined to be releasable.

[0113] For example, the parking space entrance area (slot_entrance_roi) and the parking space centerline are calculated based on the corner point of the parking space. If the distance from an obstacle within the entrance area to the parking space centerline is greater than or equal to the aisle safety distance threshold, it is considered releaseable. The aisle safety distance threshold can be determined based on the vehicle's turning radius and safety margin. This ensures that vehicles can smoothly enter the parking space, avoiding collisions caused by narrow aisles.

[0114] Step S605: Based on the releaseable state information of the parking space to be observed and the releaseable state information of the parking space aisle space, determine the parking availability state information of the parking space to be observed.

[0115] In one example, if the resolvable state information of both the parking space and the aisle space is resolvable, the parking space's drivability is determined to be drivable. Thus, the parking space is ultimately determined to be drivable only when both the parking space and the aisle space are resolvable. For example, if the parking space has sufficient internal space but the aisle is blocked by an obstacle, it is still considered unparkable, ensuring the safety of the parking process.

[0116] In another example, if the grid observation attribute is poor observation, the original parking availability information is used for historical parking spaces, while the parking availability data of the parking spaces to be observed is discarded directly to avoid introducing unreliable data.

[0117] Based on the above embodiments, a multi-dimensional dynamic evaluation mechanism, combining grid observation attributes and obstacle information obtained from multiple sensor samplings, can accurately update the parking availability counter. Furthermore, by considering the release conditions of both the parking space and the aisle space, a reliable judgment of parking availability can be achieved. This not only effectively avoids the impact of single sampling errors and poorly observed data, but also significantly improves the accuracy and robustness of the judgment by dynamically combining grid status and obstacle distance. Ultimately, it ensures that parking is only deemed possible when both the interior of the parking space and the aisle space meet safety conditions, providing a safe and efficient decision-making basis for autonomous parking.

[0118] In one embodiment of this application, after determining the parking availability information of the parking space to be observed, the historical parking space information can be updated based on the parking availability information of the parking space to be observed. In one example, if the parking availability information of the parking space to be observed is available, the parking space information of the parking space to be observed is added to the historical parking space information.

[0119] For example, if a parking space is available, its information, such as location coordinates, size, type, and surrounding environment, can be added to the historical parking space database. This continuously enriches the available parking space resource data, allowing for quick retrieval of suitable parking spaces and route planning when vehicles re-enter the area, thus improving parking efficiency and safety.

[0120] In one embodiment of this application, in order to further enhance the user's parking experience and achieve a more intelligent and convenient parking operation, the system can output historical parking space information that is available for parking, respond to the user's selection operation of any historical parking space that is available for parking, and control the vehicle to park in any historical parking space.

[0121] In one example, filtered and confirmed historical parking space information can be presented to users in an intuitive and clear manner. The presentation format is diverse: it can be displayed on an in-vehicle screen as a graphical map, clearly marking the specific location of each historical parking space with specific icons for easy viewing; or it can be presented as a list, detailing key information for each available parking space, such as space number, distance from the vehicle, and parking type (perpendicular, parallel, etc.), allowing users to quickly filter according to their needs and preferences.

[0122] After browsing this historical parking space information, users can select any available parking space based on their own situation and preferences. For example, users can make a selection through touch buttons on the in-vehicle display, voice commands, or corresponding function buttons on a mobile application connected to the vehicle.

[0123] Once the user completes their selection, the system uses historical parking space information, combined with the vehicle's current real-time status (such as position, speed, and direction) and surrounding environmental information (such as obstacle distribution and road conditions), to plan an optimal parking path for the vehicle using a path planning algorithm. This optimal parking path not only considers factors such as shortest distance and simplest operation, but also fully ensures safety during the journey, avoiding collisions with surrounding obstacles.

[0124] Subsequently, the system can control various actuators of the vehicle, such as the steering system, braking system, and power system, to park the vehicle in any historical parking space selected by the user according to the optimal parking path. Throughout the parking process, the system can monitor the vehicle's status and changes in the surrounding environment in real time. In case of emergencies, the system can adjust the control strategy in a timely manner to ensure the safe and smooth completion of the parking process, providing users with an efficient, worry-free, and safe parking experience.

[0125] The parking space availability determination method provided in this application acquires details of the currently detected target parking space and obstacle distribution information based on the vehicle's sensors. By comparing this information with historical parking space information, it efficiently locates the parking space to be observed, avoiding invalid processing. If the parking space to be observed is within a preset observation area, its grid observation attributes are precisely defined using a grid map, discretizing complex information for easier analysis. Finally, by comprehensively considering grid attributes and obstacle information, the parking space availability status is accurately and reliably determined, effectively overcoming the shortcomings of existing methods such as unreasonable observation areas and insufficient judgment accuracy. This significantly improves the reliability of parking space release determination and its adaptability to different environments.

[0126] In one example, reference Figure 7 The diagram shows the flow of a parking space availability determination method provided in another embodiment of this application. The order of the steps in this flow can be changed according to different needs, and some steps can be omitted.

[0127] Step S701: Initial classification and processing of parking spaces.

[0128] Parking spaces are pre-screened to determine subsequent processing strategies. If a parking space is a historical space classified as "observe_bad," its original parking availability status (valid_status) is retained. If a new parking space (e.g., the target parking space) is identified as "observe_bad," its data is discarded to avoid interference from unreliable data. For parking spaces outside the current observation area, historical parking spaces retain their historical status, while new parking spaces are also discarded due to a lack of valid observation data. This ensures the rational allocation of processing resources and control over data quality.

[0129] Step S702: Calculate the basic parameters of the available parking spaces.

[0130] For valid parking spaces that pass the initial screening, determine the range of interest (e.g., the entrance area slot_area_roi) and the median line of the parking space to provide basic geometric information for subsequent spatial analysis.

[0131] Step S703: Determine the grid observation attributes of the parking space.

[0132] Parking spaces are categorized based on observations from the grid map. For example, if a parking space is determined to be completely free (observe_perfect_free), it is further verified whether the space between obstacles on both sides of the parking space is less than a preset threshold. If the threshold is not met, it indicates that there is sufficient space, and the parking space counter (map_valid_counter) is incremented by 1.

[0133] If the parking space status is uncertain (observe_uncertain), calculate the minimum distance from the obstacles on the left and right sides of the parking space to the center point (left_dis, right_dis), and adjust the counter according to the comparison result of the sum of the two and the parking space threshold: if the sum is less than the threshold, the counter is decremented by 1, otherwise it is incremented by 1.

[0134] If a parking space is marked as completely not free (observe_perfect_no_free), the available parking counter value is directly decremented, reflecting the status of the parking space being occupied or unavailable.

[0135] Step S704: Determine the conditions for releasing parking space.

[0136] The parking space release condition is determined based on the value of the parking space counter. When the counter value reaches or exceeds the release threshold (release_counter), and the distance to at least one obstacle exceeds a preset threshold, the parking space validity flag (map_obs_space_valid) is set to true. At the same time, another space validity flag (fs_obs_space_valid) is calculated based on the obstacle point information. The parking space is considered to meet the release requirements only when both flags are valid.

[0137] Step S705, determine the aisle space.

[0138] Calculate the parking space entrance area (slot_entrance_roi) and the parking space centerline, and evaluate the distance from grid obstacles or actual obstacle points to the centerline. If any distance is less than the safety threshold, the parking space is deemed unsuitable for parking; otherwise, the aisle space meets the requirements.

[0139] Step S706, Parking availability decision.

[0140] A parking space is marked as available only when both the parking space and the aisle space meet the release conditions (valid_status=available); otherwise, the parking space is considered unavailable.

[0141] Step S707: Update historical parking space information.

[0142] For available parking spaces, their detailed information (including location, type, status, etc.) is entered into the historical parking space information database for future reference. For unavailable parking spaces, they are marked as invalid or their original status is retained according to specific logical needs, ensuring the continuity and availability of historical data.

[0143] Based on the above embodiments, invalid data is effectively filtered through a pre-screening mechanism in the initial classification stage, ensuring that processing resources are tilted towards high-quality parking spaces. Subsequently, the basic geometric parameters of parking spaces are calculated to lay the data foundation for spatial analysis. In the grid observation attribute judgment stage, a dynamic counter adjustment strategy is adopted to comprehensively reflect the relationship between parking space vacancy and obstacle space. Combined with dual-condition verification of parking space release (counter threshold and obstacle distance) and aisle space safety verification, a multi-dimensional safety judgment mechanism is formed. The final decision strictly follows the principle of "dual satisfaction of space and aisle", ensuring both parking feasibility and traffic safety. For the judgment results, a dynamic update strategy of historical parking space information is used to achieve continuous accumulation of parking space knowledge and intelligent elimination of invalid data, significantly improving the adaptive capability and long-term operational reliability of the judgment system. The entire process, through resource optimization, multi-source data fusion, and progressive verification, improves system operating efficiency and robustness while ensuring judgment accuracy.

[0144] Figure 8 This is a schematic diagram of the vehicle structure provided in the embodiments of this application. The above-described parking space availability determination method is applied to vehicle 120, specifically it can be on-board equipment, controller, etc., of the vehicle. Vehicle 120 can also be a cloud server, physical server, etc., that is communicatively connected to the vehicle. The vehicle 120 may include: a memory 1201, a processor 1202, and a computer program stored on the memory 1201 and executable on the processor 1202. When the processor 1202 executes the program, it implements the parking space availability determination method provided in the above embodiments.

[0145] Furthermore, the vehicle also includes a communication interface 1203 for communication between the memory 1201 and the processor 1202.

[0146] The memory 1201 is used to store computer programs that can run on the processor 1202.

[0147] The memory 1201 may include high-speed RAM memory, and may also include non-volatile memory, such as at least one disk storage device.

[0148] If the memory 1201, processor 1202, and communication interface 1203 are implemented independently, then the communication interface 1203, memory 1201, and processor 1202 can be interconnected via a bus to complete communication between them. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc.

[0149] Optionally, in a specific implementation, if the memory 1201, processor 1202, and communication interface 1203 are integrated on a single chip, then the memory 1201, processor 1202, and communication interface 1203 can communicate with each other through an internal interface.

[0150] The processor 1202 may be a central processing unit (CPU), an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of this application.

[0151] This embodiment also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described parking space availability determination method.

[0152] This embodiment also provides a computer program product, including a computer program, which is executed to implement the parking space availability determination method of the above embodiment.

[0153] Those skilled in the art will understand that all or part of the steps of the methods described in the above embodiments can be implemented by a program instructing related hardware, and the program can be stored in a computer-readable storage medium. When executed, the program includes one or a combination of the steps of the method embodiments.

[0154] Furthermore, the functional units in the various embodiments of this application can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0155] The storage medium mentioned above can be a read-only memory, a disk, or an optical disk, etc. Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of this application.

Claims

1. A method for determining the availability of parking spaces, characterized in that, The method includes: Based on the detection information from the vehicle's sensors, determine the parking space information and obstacle information of the target parking space; Based on the comparison results between the target parking space information and the historical parking space information, the parking space to be observed is determined; If the detection information determines that the parking space to be observed is located within the observation area, the grid observation attributes of the parking space to be observed are determined based on the preset grid map; Based on the grid observation attributes and the obstacle information of the parking space to be observed, the parking availability information of the parking space to be observed is determined.

2. The parking space availability determination method as described in claim 1, characterized in that, The target parking space refers to the parking space currently detected by the sensor; the parking space information of the target parking space includes the parking space number of the target parking space; the historical parking spaces refer to parking spaces that the sensor has previously detected; the parking space information of the historical parking spaces includes the parking space number of the historical parking spaces; determining the parking space to be observed based on the comparison result between the parking space information of the target parking space and the parking space information of the historical parking spaces includes: The parking space number of the target parking space is compared with the parking space number of the historical parking space. If the parking space number of the target parking space does not match the parking space number of the historical parking space, the corresponding target parking space is determined as the parking space to be observed.

3. The parking space availability determination method as described in claim 1, characterized in that, The parking space information of the historical parking spaces includes the detection status information and parking availability information of the historical parking spaces; the method further includes: If the detection status information of the historical parking space indicates that the historical parking space is in an undetectable state, or the position of the wheel chock in the historical parking space does not meet the preset requirements, the historical parking space is determined to be in an unparkable state. If the parking availability information of the historical parking space indicates that the historical parking space is not unparkable, then the historical parking space is used as the target parking space.

4. The parking space availability determination method as described in claim 1, characterized in that, The sensor includes an ultrasonic sensor and / or a laser sensor, and the method further includes determining whether the parking space to be observed is located within the observation area based on the detection information, including: Based on the coordinates of the corner point of the parking space to be observed in the vehicle coordinate system, determine the coordinates of the center point of the entrance of the parking space to be observed in the vehicle coordinate system. If the detection information from the ultrasonic sensor indicates that the coordinates of the center point are within the detection range of the ultrasonic sensor, then the parking space to be observed is determined to be within the observation area; or, If the detection information from the laser sensor indicates that the coordinates of the center point are within the detection range of the laser sensor, it is determined that the parking space to be observed is located within the observation area.

5. The parking space availability determination method as described in claim 4, characterized in that, The sensor also includes multiple image sensors, and the step of determining whether the parking space to be observed is located within the observation area based on the detection information further includes: If the detection information from the plurality of image sensors indicates that the coordinates of the center point are within the detection range of the plurality of image sensors, it is determined that the parking space to be observed is located within the observation area; or, If the parking space to be observed is a line parking space, select an image sensor combination from the plurality of image sensors according to the center point coordinates; The fan-shaped observation range is determined based on the image sensor combination. If the coordinates of the center point are within the fan-shaped observation range, it is determined whether the parking space to be observed is located within the observation area.

6. The parking space availability determination method as described in claim 1, characterized in that, The parking space information of the target parking space includes the coordinates of the corner point of the target parking space in the vehicle coordinate system corresponding to the vehicle; the obstacle information includes at least one of the following: the coordinates of the obstacle in the vehicle coordinate system, and the obstacle type; the grid observation attributes include poor observation, completely empty, completely not empty, and uncertain; the determination of the grid observation attributes of the parking space to be observed based on the preset grid map includes at least one of the following methods: The vehicle's position in the preset grid map is determined, and the parking space area of ​​the parking space to be observed in the preset grid map is determined based on the vehicle's position and the coordinates of the corner point of the parking space to be observed in the vehicle coordinate system. If the parking space area indicates that any corner point of the parking space to be observed is outside the preset grid map, the grid observation attribute of the parking space to be observed is determined to be poor observation; Based on the category label of each grid in the preset grid map, count the number of grids in each category, where the category labels include free, obstacle, and unknown; The grid observation attributes are determined based on the number of grids in each category and the total number of grids corresponding to the parking space area.

7. The parking space availability determination method as described in claim 6, characterized in that, The step of determining the grid observation attributes based on the number of grids in each category and the total number of grids corresponding to the parking space area includes: If all grids within the parking space area are of unknown category, the grid observation attribute is determined to be poor observation. If there are no obstacle-type grids within the parking space area, the grid observation attribute is determined to be completely empty; If there are grids of obstacle type in the parking space area, and the sum of the number of grids of obstacle type and the number of grids of unknown type exceeds the proportion of the total number of grids, the grid observation attribute is determined to be completely not idle. If there are grids of obstacle type in the parking space area, but the sum of the number of grids of obstacle type and the number of grids of unknown type does not exceed the first proportion threshold, the grid observation attribute is determined to be uncertain; If there are no grids of the obstacle category in the parking space area, and the proportion of the number of grids of the unknown category to the total number exceeds a preset second proportion threshold, the grid observation attribute is determined to be poor observation. If there are no obstacle category grids in the parking space area, and the number of empty category grids accounts for more than the first proportion threshold, the grid observation attribute is determined to be completely empty. If there are no obstacle category grids within the parking space area, but the proportion of the number of vacant category grids to the total number of grids does not exceed the first proportion threshold, the grid observation attribute is determined to be uncertain.

8. The parking space availability determination method as described in claim 1, characterized in that, The step of determining the parking availability information of the parking space to be observed based on the grid observation attributes and the obstacle information of the parking space to be observed includes: During the vehicle's operation, the grid observation attributes of the parking space to be observed are determined multiple times based on the sampling rate of the sensor. Based on the obstacle information, determine the minimum distance between the obstacles on the left and right sides of the parking space to be observed and the center point of the parking space; Based on the repeatedly determined grid observation attributes and the minimum distance, update the parking count value corresponding to the preset parking counter; Based on the available parking count and the minimum distance, the release status information of the parking space to be observed and the release status information of the parking space aisle space are determined. Based on the releaseable state information of the parking space to be observed and the releaseable state information of the parking aisle space, the parking availability information of the parking space to be observed is determined.

9. The parking space availability determination method as described in claim 8, characterized in that, The grid observation attributes include completely idle, completely non-idle, and uncertain. Updating the preset parking counter value based on the repeatedly determined grid observation attributes and the minimum distance includes: If the grid observation attribute is completely idle, increase the available count value by a preset value; If the grid observation attribute is completely not idle, reduce the available count value by the preset value; If the grid observation attribute is uncertain and the minimum distance is less than a preset distance threshold, the parking count value is reduced by the preset value; If the grid observation attribute is uncertain and the minimum distance is greater than or equal to the distance threshold, the parking count value is increased by the preset value.

10. The parking space availability determination method as described in claim 8, characterized in that, The process of determining the release status information of the parking space to be observed and the release status information of the parking aisle space based on the available parking count value and the minimum distance includes: If the available parking count is greater than a preset count threshold, and at least one of the minimum distances is greater than a preset distance threshold, the parking space release status information is determined to be releaseable. Based on the coordinates of the corner point of the parking space to be observed in the vehicle coordinate system, determine the position of the entrance area of ​​the parking space to be observed and the centerline position of the parking space to be observed. If the distance between the location of an obstacle in the parking entrance area and the location of the centerline is greater than or equal to a preset safe passageway distance threshold, the releaseable status information of the parking space passageway is determined to be releaseable.

11. The parking space availability determination method as described in claim 8, characterized in that, The process of determining the parking availability status of the parking space to be observed based on the releaseable status information of the parking space and the releaseable status information of the parking space aisle space includes: If the releasable state information of the parking space and the parking space aisle space are both releasable, then the parking availability state information of the parking space to be observed is determined to be parking available.

12. The parking space availability determination method as described in claim 1, characterized in that, The method further includes: Output historical parking space information, respond to user selection of any historical parking space, and control the vehicle to park in any historical parking space.

13. A vehicle, characterized in that, The vehicle includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When executed by the processor, the computer program implements the parking space availability determination method as described in any one of claims 1 to 12.

Images