Parking garage location detection method and system based on auxiliary high reflection plate and vehicle

Abstract

The invention discloses a parking garage location detection method and system based on an auxiliary high reflection plate and a vehicle, the method deploys a high reflection plate in a parking garage location of an intelligent heavy truck, realizes high-precision, high-efficiency and low-cost parking garage location detection in combination with a laser radar and the high reflection plate, reduces the dependence of the intelligent heavy truck on a high-precision map, ground identification and the like, and improves the detection accuracy of the parking garage location. And the storage location does not need to be additionally transformed, the operation efficiency of automatic loading and unloading is improved, and then the automation level and production benefits of an industrial park are improved.

Description

Technical Field

[0001] The present invention belongs to the field of intelligent driving technology and relates to the measurement and positioning of laser radar, and specifically to a parking space detection method, system and vehicle based on an auxiliary high-reflective plate. Background Art

[0002] In parking scenarios, smart driving vehicles need to complete accurate parking operations with a high success rate. The automated storage spaces in industrial warehouses place strict requirements on the parking accuracy of heavy trucks with trailers. Only when the vehicle is parked accurately in the longitudinal, lateral, and heading directions of the storage space can automated loading and unloading operations be achieved.

[0003] Publication No. CN116620311A discloses a parking error detection method, device, vehicle, and storage medium. This method determines the relative position of a target parking space relative to a vehicle. This technology is based on surround-view images and relies on a black-and-white grid and ArUco markers. This technology is limited in practical application by environmental conditions and marker status. Pre-positioning markers increases implementation cost and inconvenience. Furthermore, the image processing algorithm employed is typically computationally complex and is affected by ambient lighting and other conditions, potentially leading to instability and compromising high-precision parking error calculation and overall effectiveness.

[0004] Publication No. CN111257893A discloses a parking space detection method and automated parking method that utilizes ultrasonic sensors and sampling point processing to obtain parking space outlines. This technology uses ultrasonic sensors for parking space detection and obstacle distance measurement. The ultrasonic sensors have low precision, making it difficult to accurately identify parking space corners and boundaries, potentially resulting in insufficient parking accuracy. The low spatial resolution also prevents detailed feature information from being provided, and the system is susceptible to environmental factors such as temperature and humidity, resulting in unstable detection results. Furthermore, the system relies on obstacle distance measurement, failing to provide multi-dimensional environmental perception data and struggling to meet the demands of high-speed, high-precision parking.

[0005] Publication No. CN118061986A discloses a parking space detection and secondary confirmation method and system, a vehicle parking method, and an electronic device. Upon reaching the secondary confirmation position, the vehicle is controlled to continue parking based on the second parking space's global coordinate information. This technology uses a surround-view system and the second parking space's global coordinate information for parking space detection and secondary confirmation. Its reliance on the surround-view system makes it susceptible to lighting, weather, and camera cleanliness, resulting in insufficient detection accuracy and stability. The secondary confirmation adds additional calculation and processing time, while the reliance on the second parking space's global coordinate information increases implementation costs and technical complexity. Furthermore, the phased detection and confirmation process can lead to a lack of consistency in the parking process, making it difficult to adapt to dynamic environmental changes. Summary of the Invention

[0006] In response to the above problems, the main purpose of the present invention is to design a parking space detection method, system and vehicle based on auxiliary high-reflective plates, using laser sensors to detect and output the corner points of the parking space in the heavy-duty truck coordinate system, thereby solving the technical problems of insufficient detection accuracy, poor real-time performance and adaptability, and inability to park continuously in the existing technology.

[0007] In order to achieve the above purpose, the present invention adopts the following technical solutions:

[0008] A parking space detection method based on an auxiliary high-reflective plate is used for scenarios where intelligent heavy trucks are parked in parking spaces that include alignment equipment. The method specifically includes the following steps:

[0009] Install high-reflective plates in the parking spaces for intelligent heavy trucks;

[0010] The intelligent heavy truck receives the parking task and uses LiDAR to obtain point cloud data of the parking space's rear wall and high-reflective plate.

[0011] The point cloud data acquired by the LiDAR is pre-processed to extract the back wall of the storage location containing the high-reflective plate as the point cloud of the target area;

[0012] In the point cloud of the target area, the position and orientation of the high-reflection board are calculated in real time;

[0013] The high-reflection plate is defined as the storage location point in the trailer coordinate system, and the storage location corner point in the trailer coordinate system is generated through the rigid position relationship between the high-reflection plate and the storage location corner point;

[0014] The coordinates of the parking space corner points in the trailer coordinate system are converted to the tractor coordinate system, and the output guides the intelligent heavy-duty truck tractor to park in the parking space.

[0015] As a further description of the present invention, the high reflective plate is arranged on the rear wall of the storage location, specifically including the following steps:

[0016] Confirm the horizontal reference point, and draw a horizontal reference point A at the center of the rear wall of the storage location and the alignment equipment;

[0017] The vertical reference line is confirmed by drawing a vertical reference line L along the longitudinal direction of the rear wall of the storage location, and the vertical reference line L passes through the horizontal reference point A;

[0018] Align the center of the high-reflection plate with the horizontal reference point A and install it axisymmetrically along the vertical reference line L.

[0019] As a further description of the present invention, one laser radar is provided and located in the middle of the rear of the trailer of the smart heavy truck, which is used to obtain point cloud data of the high-reflective plate and the rear wall of the storage space during the entire parking process of the smart heavy truck.

[0020] As a further description of the present invention, the process of point cloud preprocessing is:

[0021] Based on the point cloud data obtained by the lidar, the trailer positioning is used to obtain the position P of the trailer origin in the global coordinate system trailer , determine the rough pose P of the high-reflection board in the global coordinate system through a static data reflector ;

[0022] According to the trailer origin pose P trailer and high-reverse board posture P reflector Relative transformation, converting the point cloud in the trailer coordinate system to the high-reflection board coordinate system;

[0023] In the high-reflection board coordinate system, a screening range is preset, and the rough center of the high-reflection board is used as the origin to extract the point cloud data within the range to obtain the point cloud of the target area.

[0024] As a further description of the present invention, the position calculation of the high reflective plate includes the following steps:

[0025] For the points inside the high-reflection board, set the intensity threshold and point number constraints;

[0026] Sort the intensity values ​​of the point cloud in the target area and extract the points whose intensity values ​​exceed the set intensity threshold;

[0027] Combined with the point number constraint, the inner points of the high-reflection board are screened out;

[0028] The coordinates of the inner points of the screened high-reflection board are averaged or a straight line is fitted to the high-reflection board to obtain the coordinates (x, y) of the center of the high-reflection board in the trailer coordinate system.

[0029] As a further description of the present invention, the calculation of the orientation of the high-reflection board includes the following steps:

[0030] Through the relative transformation of the trailer posture and the high-reflection board posture, the point cloud in the trailer coordinate system is converted to the high-reflection board coordinate system;

[0031] The point cloud of the target area is divided into blocks, and the plane point cloud on the storage location wall is obtained by performing eigenvalue decomposition and screening on each point cloud block;

[0032] Project the wall point cloud onto the ground, and use the least squares method to perform straight line fitting to obtain the fitted rear wall straight line of the storage location;

[0033] The vertical direction vector of the rear wall of the storage location is used as the direction of the high reflective board.

[0034] As a further description of the present invention, generating a storage location corner point in a trailer coordinate system includes the following steps:

[0035] The coordinates of the storage corner point in the trailer coordinate system are obtained by the position and orientation of the high-reflection plate in the trailer coordinate system, as well as the rigid position relationship between the high-reflection plate and the storage corner point. The expression is:

[0036]

[0037] in, is the coordinate of corner point A relative to the center of the high reflective board, l and w are the known length and width of the storage location, is the coordinate of the center of the known high-reflection plate in the trailer coordinate system;

[0038] Similarly, obtain the coordinates of all storage location corner points in the trailer coordinate system;

[0039] All the obtained storage location corner points are used as the storage locations in the trailer coordinate system.

[0040] As a further description of the present invention, the conversion of the storage location corner point in the trailer coordinate system to the tractor coordinate system includes the following steps:

[0041] Time synchronization of lidar observations, trailer pose, and tractor pose;

[0042] Using the relative positions of the tractor and trailer at the same moment, calculate the transformation matrix from the trailer coordinate system to the tractor coordinate system;

[0043] The coordinates of the storage corner points in the trailer coordinate system are converted to the tractor coordinate system through the transformation matrix;

[0044] Output the storage location coordinates in the tractor coordinate system in real time.

[0045] A parking space detection system based on auxiliary high-reflective plates includes a data acquisition module, a preprocessing module, a storage space generation module, and a coordinate conversion module. When the system is executed, high-reflective plates and alignment equipment are placed in the parking spaces, and the high-reflective plates are used as markers for auxiliary positioning.

[0046] The data acquisition module is used to obtain point cloud data of the storage location back wall and high-reflective plate through laser radar;

[0047] The pre-processing module is used to process the point cloud data acquired by the data acquisition module and extract the back wall of the high-reflective board storage location as the target area point cloud;

[0048] The storage location generation module is used to define the high-reflection plate as a storage location point in the trailer coordinate system, and generate the storage location corner point in the trailer coordinate system based on the rigid position relationship between the high-reflection plate and the storage location corner point;

[0049] The coordinate conversion module is used to convert the coordinates of the parking space corner points in the trailer coordinate system to the tractor coordinate system to realize the detection of the parking space.

[0050] A vehicle comprises the above-mentioned detection system and executes the above-mentioned detection method to achieve parking of the vehicle.

[0051] Compared with the prior art, the technical effects of the present invention are:

[0052] The present invention provides a parking space detection method, system and vehicle based on auxiliary high-reflective plates. The method deploys high-reflective plates in the parking spaces of intelligent heavy trucks, obtains point cloud data of the parking space back wall and the high-reflective plates through laser radar, extracts the target area point cloud containing the parking space back wall of the high-reflective plate through point cloud preprocessing, and calculates the position and orientation of the high-reflective plates in real time in the target area point cloud; defines the high-reflective plates as the parking point positions in the trailer coordinate system, generates the parking space corner points in the trailer coordinate system through the rigid position relationship between the high-reflective plates and the parking space corner points, converts the coordinates of the parking space corner points in the trailer coordinate system into the tractor coordinate system, and outputs the method to guide the intelligent heavy truck tractor into the parking space. The present invention combines laser radar and high-reflective plates to achieve high-precision, efficient and low-cost parking space detection, reduces the dependence of intelligent heavy trucks on high-precision maps, ground signs, etc., and does not require additional modification of the parking spaces, improves the efficiency of automated loading and unloading operations, and is conducive to improving the automation level and production efficiency of industrial parks. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 Schematic diagram of the overall method flow of the present invention;

[0054] Figure 2 This is a schematic diagram of the installation of a high reflective board of the present invention;

[0055] Figure 3 Schematic diagram of the trailer rear laser radar and its field of view of the present invention;

[0056] Figure 4 This is a schematic diagram of the point cloud of the warehouse back wall including the high-reflective board of the present invention;

[0057] Figure 5 This is a schematic diagram of the target area point cloud extraction effect of the present invention;

[0058] Figure 6 Generate a schematic diagram for the storage location and navigation of the present invention;

[0059] Figure 7 Generate a schematic diagram for the storage location corner points of the present invention;

[0060] Figure 8 This is a schematic diagram of the trailer-to-tractor garage conversion of the present invention;

[0061] Figure 9 This is a schematic diagram of the visualization effect of the storage location detection of the present invention. DETAILED DESCRIPTION

[0062] The present invention is described in detail below with reference to the accompanying drawings:

[0063] In one embodiment of the present invention, a parking space detection method based on an auxiliary high-reflective plate is disclosed, referring to Figure 1-9 As shown, this method is used in a scenario where an intelligent heavy truck is parked in a storage space, and the storage space includes a positioning device, and specifically includes the following steps:

[0064] Install high-reflective plates in the parking spaces for intelligent heavy trucks;

[0065] The intelligent heavy truck receives the parking task and uses LiDAR to obtain point cloud data of the parking space's rear wall and high-reflective plate.

[0066] The point cloud data acquired by the LiDAR is pre-processed to extract the back wall of the storage location containing the high-reflective plate as the point cloud of the target area;

[0067] In the point cloud of the target area, the position and orientation of the high-reflection board are calculated in real time;

[0068] The high-reflection plate is defined as the storage location point in the trailer coordinate system, and the storage location corner point in the trailer coordinate system is generated through the rigid position relationship between the high-reflection plate and the storage location corner point;

[0069] The coordinates of the parking space corner points in the trailer coordinate system are converted to the tractor coordinate system, and the output guides the intelligent heavy-duty truck tractor to park in the parking space.

[0070] Specifically, in this embodiment, the high-reflective plate is made of a plate with high laser reflectivity, and its size is usually no more than 45*75cm. The specific size is determined by the site layout. The high-reflective plate is placed on the back wall of the storage location, including the following steps:

[0071] Confirm the horizontal reference point, and draw a horizontal reference point A at the center of the rear wall of the storage location and the alignment equipment;

[0072] The vertical reference line is confirmed by drawing a vertical reference line L along the longitudinal direction of the rear wall of the storage location, and the vertical reference line L passes through the horizontal reference point A;

[0073] Align the center of the high-reflection plate with the horizontal reference point A and install it axisymmetrically along the vertical reference line L.

[0074] High-reflective plates do not require custom production and are low-cost, easy to purchase and maintain. They can generate high-intensity points in the point cloud scanned by the LiDAR. At long distances, their intensity values ​​are reduced, but as long as they are different from the ordinary point clouds nearby, they are sufficient. High-reflective plates need to be aligned with the warehouse docking equipment to ensure good initial accuracy. Specifically, Figure 2The figure shows a precise and convenient method for installing a high-reflection plate. The plate's position is precisely determined in two steps. Step one is lateral reference point alignment. Using a T-ruler, the center of the docking station and the rear wall of the storage area are aligned horizontally, marking the horizontal installation center point A for the high-reflection plate. Step two is vertical reference line confirmation. A vertical baseline L is drawn along the plumb line, passing the plumb line perpendicularly through point A. The high-reflection plate is then installed axially symmetrically along vertical line L, completing the installation. This convenient operation ensures the high-reflection plate is installed with high precision on the rear wall of the storage area, providing a solid foundation for subsequent storage area detection using point clouds.

[0075] like Figure 3 As shown, since the loading, unloading and parking of smart heavy trucks are mainly backward movements, a laser radar is installed at the rear of the trailer of the smart heavy truck to obtain point cloud data of the high-reflective plate and the rear wall of the storage space during the entire parking process of the smart heavy truck; specifically, one laser radar is set and located in the middle position of the rear of the trailer of the smart heavy truck. The height and angle of the radar enable the effective observation point cloud of the high-reflective plate and the rear wall of the storage space to be obtained during the entire parking process, providing good and stable input for subsequent point cloud preprocessing.

[0076] In this embodiment, the position of the rough center of the high-reflection plate in the radar coordinate system can be known through static data once. Specifically, the point cloud preprocessing process is as follows:

[0077] Based on the point cloud data obtained by the lidar, the trailer positioning is used to obtain the position P of the trailer origin in the global coordinate system trailer , determine the rough pose P of the high-reflection board in the global coordinate system through a static data reflector ;

[0078] According to the trailer origin pose P trailer and high-reverse board posture P reflector Relative transformation, converting the point cloud in the trailer coordinate system to the high-reflection board coordinate system;

[0079] In the high-reflection board coordinate system, a screening range is preset, and the rough center of the high-reflection board is used as the origin to extract the point cloud data within the range to obtain the point cloud of the target area.

[0080] In this embodiment, by presetting a smaller range, a smaller rear wall point cloud containing a high-reflection plate area is selected through range screening, so that the point cloud of the high-reflection plate area can be stably obtained even at a larger attitude angle of the trailer. Figure 4 The white coarse-grained point cloud in .

[0081] During the parking process of heavy trucks, the point cloud of the target area changes from sparse to dense. In order to stably extract the inner points of the high-reflection board, the point cloud of this area is sorted according to the intensity value, and filtered by the dual constraints of intensity value and number of points. Finally, the inner points of the high-reflection board are stably extracted, such as Figure 5The dense white point cloud in the middle. The point cloud of the high-reflective plate is relatively short. By taking the average of the coordinates of all the internal points of the high-reflective plate, or performing a linear fit on the high-reflective plate, the precise coordinates of the high-reflective plate center in the trailer coordinate system can be obtained. Specifically, in this embodiment, the position calculation of the high-reflective plate includes the following steps:

[0082] For the points inside the high-reflection board, set the intensity threshold and point number constraints;

[0083] Sort the intensity values ​​of the point cloud in the target area and extract the points whose intensity values ​​exceed the set intensity threshold;

[0084] Combined with the point number constraint, the inner points of the high-reflection board are screened out;

[0085] The coordinates of the inner points of the screened high-reflection board are averaged or a straight line is fitted to the high-reflection board to obtain the coordinates (x, y) of the center of the high-reflection board in the trailer coordinate system.

[0086] like Figure 6 As shown, the heading of the high-reflective board is consistent with the heading of the storage location. Since the high-reflective board area is small, the heading can be calculated using the wall where the high-reflective board is located. Specifically, the direction calculation of the high-reflective board includes the following steps:

[0087] The point cloud in the trailer coordinate system is converted to the high-reflection board coordinate system through the relative transformation of the trailer posture and the high-reflection board posture. The rough point cloud of the rear wall area of ​​the storage location can also be screened to remove the noise point cloud that does not belong to the wall plane in this area. The point cloud of the target area is divided into blocks, and the eigenvalue decomposition and screening of each point cloud are performed to obtain the plane point cloud on the storage location wall, and the point cloud that does not belong to the wall is removed.

[0088] The wall point cloud after noise removal is projected onto the ground, and a linear fit is performed using the least squares method to obtain the fitted rear wall straight line of the storage location. During the fitting process, a small number of outliers (noise points on the straight line) are removed.

[0089] The vertical direction vector of the rear wall of the storage location is used as the direction of the high reflective board, such as Figure 6 As shown by the arrow in .

[0090] like Figure 7 As shown, by real-time calculation of the position of the center of the high-reflective board and the heading of the high-reflective board, the longitudinal, lateral, and heading data of the high-reflective board in the heavy truck trailer coordinate system are accurately obtained, and the high-reflective board can be abstracted as a storage location under the trailer system. Since the positional relationship between the high-reflective board and the four corner points of the overall storage location is rigidly fixed, the position of the high-reflective board in the trailer system and this positional relationship can be used to obtain the position of the storage location corner point in the trailer coordinate system. Specifically, in this embodiment, generating the storage location corner point in the trailer coordinate system includes the following steps:

[0091] The coordinates of the storage location corner point in the trailer coordinate system are obtained through the position and orientation of the high-reflection plate in the trailer coordinate system, as well as the rigid position relationship between the high-reflection plate and the storage location corner point. Taking the corner point A in the lower right corner of the storage location as an example, the expression is:

[0092]

[0093] in, is the coordinate of corner point A relative to the center of the high reflective board, l and w are the known length and width of the storage location, is the coordinate of the center of the known high-reflection plate in the trailer coordinate system;

[0094] Similarly, obtain the coordinates of all storage location corner points in the trailer coordinate system;

[0095] All the obtained storage location corner points are used as the storage locations in the trailer coordinate system.

[0096] like Figure 8 As shown in the figure, since the whole vehicle control of intelligent heavy trucks is usually based on the tractor, it is necessary to convert the obtained storage location in the trailer coordinate system to the tractor system and output the parking in real time; specifically, the conversion of the storage location corner point in the trailer coordinate system to the tractor coordinate system includes the following steps:

[0097] Time synchronization of lidar observations, trailer pose, and tractor pose;

[0098] Using the relative positions of the tractor and trailer at the same moment, calculate the transformation matrix from the trailer coordinate system to the tractor coordinate system;

[0099] The coordinates of the parking space corner points in the trailer coordinate system are converted to the tractor coordinate system through the transformation matrix, and finally the parking space expression in the front vehicle coordinate system is obtained;

[0100] Output the storage location coordinates in the tractor coordinate system in real time.

[0101] In this embodiment, through the precise installation of the high-reflective plate and point cloud detection processing, the coordinate system of the tractor is output in real time, which reduces the dependence on high-precision maps and ground signs during parking, and has good output consistency in the case of multiple vehicles. Figure 9 As shown in the figure, the square box is the warehouse location alignment equipment, and the middle vertical line is the output warehouse location horizontal direction. The vertical line accurately passes through the central axis of the alignment equipment, that is, the key horizontal direction accuracy is high; the horizontal line and arrow in the figure express the fitting line of the warehouse location back wall and the output warehouse location heading. It can be seen that the warehouse location heading is also relatively accurate; the left and right vertical lines are the lines connecting the two warehouse location corner points on the left and right respectively. Finally, the four warehouse location corner points are output and updated in real time.

[0102] In another embodiment of the present invention, a parking space detection system based on auxiliary high-reflective boards is disclosed. The system is used to implement the above-mentioned parking space detection method and includes a data acquisition module, a preprocessing module, a parking space generation module, and a coordinate conversion module. When the system is executed, high-reflective boards and alignment equipment are arranged in the parking spaces, and the high-reflective boards are used as markers to assist in positioning.

[0103] The data acquisition module is used to obtain point cloud data of the storage location back wall and high-reflective plate through laser radar;

[0104] The pre-processing module is used to process the point cloud data acquired by the data acquisition module and extract the back wall of the high-reflective board storage location as the target area point cloud;

[0105] The storage location generation module is used to define the high-reflection plate as a storage location point in the trailer coordinate system, and generate the storage location corner point in the trailer coordinate system based on the rigid position relationship between the high-reflection plate and the storage location corner point;

[0106] The coordinate conversion module is used to convert the coordinates of the parking space corner points in the trailer coordinate system to the tractor coordinate system to realize the detection of the parking space.

[0107] The above embodiments disclose the detection method and system of the present invention. Compared with the prior art, the present invention has the following advantages:

[0108] 1. This invention provides accurate detection of storage locations through an efficient and convenient high-reflective plate installation method, obtaining precise storage location corner points without the need for construction such as storage location marking;

[0109] 2. This invention uses a high-reflective plate in conjunction with a laser radar to achieve high-precision parking for vehicles at different locations, reducing reliance on map markers, improving loading and unloading efficiency, and increasing the economic benefits of each vehicle per unit time.

[0110] 3. The implementation of the present invention does not involve the transformation of the storage site, does not limit the size and form of the storage site, and has the characteristics of good generalization and low cost.

[0111] In another embodiment of the present invention, a vehicle is disclosed. The vehicle includes the above-mentioned detection system and executes the above-mentioned detection method to achieve parking of the vehicle.

[0112] The above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the scope of the present invention. Other modifications or equivalent substitutions made to the technical solutions of the present invention by ordinary technicians in this field should be included in the scope of the claims of the present invention as long as they do not depart from the spirit and scope of the technical solutions of the present invention.

Claims

1. A parking space detection method based on auxiliary high-reflective board, characterized in that: This method is used in scenarios where a smart heavy truck is parked in a storage space that includes alignment equipment. It specifically includes the following steps: Install high-reflective plates in the parking spaces for intelligent heavy trucks; The intelligent heavy truck receives the parking task and uses LiDAR to obtain point cloud data of the parking space's rear wall and high-reflective plate. The point cloud data acquired by the LiDAR is pre-processed to extract the back wall of the storage location containing the high-reflective plate as the point cloud of the target area; In the point cloud of the target area, the position and orientation of the high-reflection board are calculated in real time; The high-reflection plate is defined as the storage location point in the trailer coordinate system, and the storage location corner point in the trailer coordinate system is generated through the rigid position relationship between the high-reflection plate and the storage location corner point; The coordinates of the parking space corner points in the trailer coordinate system are converted to the tractor coordinate system, and the output guides the intelligent heavy-duty truck tractor to park in the parking space.

2. The parking space detection method based on auxiliary high-reflective board according to claim 1 is characterized in that: The high reflective plate is placed on the back wall of the storage location, which includes the following steps: Confirm the horizontal reference point, and draw a horizontal reference point A at the center of the rear wall of the storage location and the alignment equipment; The vertical reference line is confirmed by drawing a vertical reference line L along the longitudinal direction of the rear wall of the storage location, and the vertical reference line L passes through the horizontal reference point A; Align the center of the high-reflection plate with the horizontal reference point A and install it axisymmetrically along the vertical reference line L.

3. The parking space detection method based on auxiliary high-reflective board according to claim 1 is characterized in that: The laser radar is provided in one piece and is located in the middle of the rear of the trailer of the intelligent heavy truck, and is used to obtain point cloud data of the high-reflective plate and the rear wall of the parking space during the entire parking process of the intelligent heavy truck.

4. The parking space detection method based on auxiliary high-reflective board according to claim 1 is characterized in that: The process of point cloud preprocessing is: Based on the point cloud data obtained by the lidar, the trailer positioning is used to obtain the position P of the trailer origin in the global coordinate system trailer , determine the rough pose P of the high-reflection board in the global coordinate system through a static data reflector ; According to the trailer origin pose P trailer and high-reverse board posture P reflector Relative transformation, converting the point cloud in the trailer coordinate system to the high-reflection board coordinate system; In the high-reflection board coordinate system, a screening range is preset, and the rough center of the high-reflection board is used as the origin to extract the point cloud data within the range to obtain the point cloud of the target area.

5. The parking space detection method based on auxiliary high-reflective board according to claim 4 is characterized in that: The position calculation of the high-reflection board includes the following steps: For the points inside the high-reflection board, set the intensity threshold and point number constraints; Sort the intensity values ​​of the point cloud in the target area and extract the points whose intensity values ​​exceed the set intensity threshold; Combined with the point number constraint, the inner points of the high-reflection board are screened out; The coordinates of the inner points of the screened high-reflection board are averaged or a straight line is fitted to the high-reflection board to obtain the coordinates (x, y) of the center of the high-reflection board in the trailer coordinate system.

6. The parking space detection method based on auxiliary high-reflective board according to claim 5, characterized in that: The calculation of the orientation of the high-reflection board includes the following steps: Through the relative transformation of the trailer's posture and the high-reflection board's posture, the point cloud in the trailer coordinate system is converted to the high-reflection board coordinate system; the point cloud in the target area is divided into blocks, and the plane point cloud on the storage wall is obtained by eigenvalue decomposition and screening of each point cloud block; The wall point cloud is projected onto the ground, and a linear fit is performed using the least squares method to obtain the fitted straight line of the storage location's rear wall. The perpendicular direction vector of the straight line of the storage location's rear wall is used as the orientation of the high-reflection board.

7. The parking space detection method based on auxiliary high-reflective board according to claim 1, characterized in that: Generating the corner points of the trailer coordinate system includes the following steps: The coordinates of the storage corner point in the trailer coordinate system are obtained by the position and orientation of the high-reflection plate in the trailer coordinate system, as well as the rigid position relationship between the high-reflection plate and the storage corner point. The expression is: in, is the coordinate of corner point A relative to the center of the high reflective board, l and w are the known length and width of the storage location, is the coordinate of the center of the known high-reflection plate in the trailer coordinate system; Similarly, obtain the coordinates of all storage location corner points in the trailer coordinate system; All the obtained storage location corner points are used as the storage locations in the trailer coordinate system.

8. The parking space detection method based on auxiliary high-reflective board according to claim 1, characterized in that: The conversion of the storage location corner points in the trailer coordinate system to the tractor coordinate system includes the following steps: Time synchronization of lidar observations, trailer pose, and tractor pose; Using the relative positions of the tractor and trailer at the same moment, the transformation matrix from the trailer coordinate system to the tractor coordinate system is calculated; the coordinates of the storage corner points in the trailer coordinate system are converted to the tractor coordinate system through the transformation matrix; Output the storage location coordinates in the tractor coordinate system in real time.

9. A parking space detection system based on an auxiliary high-reflective plate according to any one of claims 1 to 8, characterized in that: The system includes a data acquisition module, a pre-processing module, a storage location generation module, and a coordinate conversion module. When the system is executed, high-reflective plates and alignment equipment are placed in the parking spaces, and the high-reflective plates are used as markers to assist in positioning. The data acquisition module is used to obtain point cloud data of the storage location back wall and high-reflective plate through laser radar; The pre-processing module is used to process the point cloud data acquired by the data acquisition module and extract the back wall of the high-reflective board storage location as the target area point cloud; The storage location generation module is used to define the high-reflection plate as a storage location point in the trailer coordinate system, and generate the storage location corner point in the trailer coordinate system based on the rigid position relationship between the high-reflection plate and the storage location corner point; The coordinate conversion module is used to convert the coordinates of the parking space corner points in the trailer coordinate system to the tractor coordinate system to realize the detection of the parking space.

10. A vehicle, characterized in that: The vehicle includes the detection system described in claim 9 and executes the detection method described in any one of claims 1 to 8 to achieve parking of the vehicle.

Images