Roadside multi-radar adaptive spatial synchronization method, device and equipment

Abstract

This invention provides a method, apparatus, and device for adaptive spatial synchronization of multiple roadside radars. The method includes: obtaining the coordinate errors of successfully associated point pairs according to a preset error calculation method; obtaining the average association error and the mean average error of the current frame in the current iteration based on the coordinate errors and association results; and determining whether the iteration should stop based on the mean average error of the current frame and the mean average error of the previous frame. In this invention, by iteratively updating the current spatial synchronization matrix and calculating the adaptive spatial synchronization results of multiple radars based on the iteration results of the current spatial synchronization matrix, the problem of reduced accuracy of spatial synchronization results caused by environmental influences and slow parameter changes is solved. Furthermore, through automatic iteration, the coverage area of ​​successfully associated point pairs used to calculate the current spatial synchronization matrix gradually approaches the complete sensing overlap area, improving the accuracy of the adaptive spatial synchronization results.

Description

Technical Field

[0001] This invention belongs to the field of radar technology, specifically relating to a roadside multi-radar adaptive spatial synchronization method, device, and equipment. Background Technology

[0002] Millimeter-wave radar, due to its stable detection performance, long operating range, high detection accuracy, and all-weather operation, has been increasingly widely used in the transportation sector. However, in the process of using millimeter-wave radar for real-time monitoring of road traffic environment and events, the limited detection range of a single millimeter-wave radar results in blind spots, making it impossible to achieve complete coverage of the entire road. Therefore, multiple millimeter-wave radars need to be deployed along the route to monitor the entire traffic road. However, since the detection ranges of adjacent millimeter-wave radars have certain overlapping areas, how to achieve unified stitching of the trajectories of multiple millimeter-wave radars has become a crucial issue in the spatial synchronization of multi-radar conversion.

[0003] Existing technologies have proposed solutions for spatial synchronization of multiple radar systems, such as employing strict millimeter-wave radar installation strategies to simplify spatial synchronization calculations. This method involves adjusting the orientation angles of adjacent radars along a relatively straight road to ensure they are at roughly the same angle, thus obtaining their latitude and longitude. After translation, the spatial synchronization effect can be achieved. However, this method places extremely high demands on the installation and debugging of the millimeter-wave radars and is subject to significant road constraints. Some research has proposed using high-precision maps to solve the spatial synchronization problem of multiple radars. This method involves selecting and measuring points on a high-precision map for spatial synchronization. After obtaining the positions of adjacent radars on the high-precision map, the method calculates the distance, offset, and angular deviation between adjacent radars to obtain the multi-radar spatial synchronization result. While this method can achieve high accuracy in installation positions and azimuth angles, resulting in high spatial synchronization accuracy, high-precision maps are prohibitively expensive and cannot be used on road sections without them.

[0004] Therefore, existing multi-radar spatial synchronization methods suffer from poor robustness and high operating costs. Summary of the Invention

[0005] To address the aforementioned problems in the prior art, this invention provides a roadside multi-radar adaptive spatial synchronization method, apparatus, and device.

[0006] The technical problem to be solved by this invention is achieved through the following technical solution:

[0007] In a first aspect, the present invention provides a roadside multi-radar adaptive spatial synchronization method, comprising:

[0008] S101. Obtain the current target coordinate data acquired by any of the radars to be synchronized.

[0009] S102. Multiply the current target coordinate data by the current spatial synchronization matrix of the current iteration to obtain the transformed coordinate data;

[0010] S103. Based on the association stitching algorithm, perform association stitching processing on the current target coordinate data and the transformed coordinate data to obtain the association results and successfully associated point pairs;

[0011] S104. Obtain the coordinate error of the successfully associated point pair according to the preset error calculation method, and obtain the average association error of the current frame and the mean of the average error of the current frame in the current iteration based on the coordinate error and the association result.

[0012] S105. Determine whether the mean average error of the current frame and the mean average error of the previous frame in the previous iteration meet the iteration stopping condition.

[0013] S106. When the condition in S105 is not met, the current spatial synchronization matrix in S102 is updated by using the average correlation error of the current frame and the mean of the average error of the current frame to obtain the updated spatial synchronization matrix, and the updated spatial synchronization matrix is ​​used as the current spatial synchronization matrix in S102.

[0014] S107. Repeat steps S101-S106 until the condition of S105 is met. Take the current spatial synchronization matrix obtained in the most recent execution of S102 as the optimal spatial synchronization matrix, and use the optimal spatial synchronization matrix to obtain the multi-radar adaptive spatial synchronization result.

[0015] Optionally, the roadside multi-radar adaptive spatial synchronization method prior to S101 also includes:

[0016] Based on the preset point pair acquisition method, a preset number of identical target point pairs detected by adjacent radars in the radar to be synchronized are acquired, and the identical target point pairs are combined into an initial point pair set; the identical target point pairs are the coordinate information of adjacent radars detecting the same target at the same time;

[0017] The initial spatial synchronization matrix is ​​calculated based on the initial point pair set and the homography matrix algorithm.

[0018] Optionally, the preset point-to-point acquisition methods include: static acquisition method and dynamic acquisition method;

[0019] The static acquisition method utilizes the coordinate information of a triangular reflector located at the apex of the intersection and splicing area of ​​adjacent radars, which is obtained from adjacent radars.

[0020] The dynamic acquisition method utilizes the coordinate information of the same moving ground target acquired by adjacent radars at the same time.

[0021] Optionally, S104 specifically includes:

[0022] Use Euclidean distance to obtain the coordinate error of successfully associated point pairs;

[0023] The current coordinate error of the current frame is selected from the coordinate errors based on the correlation results;

[0024] The average correlation error of the current frame is obtained by averaging the current coordinate error.

[0025] The mean of the average correlation error of all current frames is calculated to obtain the mean of the average error of the current frame.

[0026] Optionally, the iteration stopping condition is that the absolute value of the difference between the mean average error of the current frame and the mean average error of the previous frame is less than a preset threshold.

[0027] Optionally, S106 specifically includes:

[0028] If the condition of S105 is not met, select new associated point pairs that meet the requirements from the successfully associated point pairs based on the preset point pair selection formula.

[0029] Add the newly added associated point pairs to the initial point pair set to obtain the updated point pair set;

[0030] The current spatial synchronization matrix in S102 is updated based on the set of update point pairs to obtain the updated spatial synchronization matrix.

[0031] The updated spatial synchronization matrix is ​​used as the current spatial synchronization matrix in S102.

[0032] Optionally, the current spatial synchronization matrix in S102 is updated based on the set of updated point pairs to obtain an updated spatial synchronization matrix, including:

[0033] Multiple target point pairs for updating are selected from the set of updated point pairs using a random sampling consensus method.

[0034] Based on the update target point pairs and homography matrix algorithm, the update space synchronization matrix is ​​calculated.

[0035] Optionally, the formula for selecting preset point pairs is expressed as:

[0036] E τ <min(k1*E frame ,k2*E syn ),(0 <k1,k2≤1);

[0037] Among them, E τE represents the coordinate error of the newly added associated point pair. frame E represents the average correlation error of the current frame. syn k1 represents the mean error of the current frame. frame The error coefficient, k2 represents E syn The error coefficient is min(·,·), which represents the minimum value operation.

[0038] Secondly, the present invention provides a roadside multi-radar adaptive spatial synchronization device, which includes: an acquisition unit, a splicing unit, a calculation unit, a judgment unit, an update unit, and an iteration unit.

[0039] The acquisition unit is used to acquire the current target coordinate data acquired by any of the radars to be synchronized.

[0040] Multiply the current target coordinate data by the current spatial synchronization matrix of the current iteration to obtain the transformed coordinate data;

[0041] The stitching unit is used to perform correlation stitching processing on the current target coordinate data and transformed coordinate data based on the correlation stitching algorithm, and to obtain the correlation results and successfully correlated point pairs.

[0042] The calculation unit is used to obtain the coordinate error of successfully associated point pairs according to a preset error calculation method, and to obtain the average association error of the current frame and the mean of the average error of the current frame in the current iteration based on the coordinate error and the association result.

[0043] The judgment unit is used to determine whether the mean average error of the current frame and the mean average error of the previous frame in the previous iteration meet the iteration stop condition.

[0044] The update unit is used to update the current spatial synchronization matrix in the acquisition unit by means of the current frame average association error and the mean of the current frame average error when the condition of the judgment unit is not met, so as to obtain the updated spatial synchronization matrix and use the updated spatial synchronization matrix as the current spatial synchronization matrix in the acquisition unit.

[0045] The iterative unit is used to jump to the acquisition unit until the condition of the judgment unit is met. The current spatial synchronization matrix obtained in the most recent acquisition unit is used as the optimal spatial synchronization matrix, and the optimal spatial synchronization matrix is ​​used to obtain the multi-radar adaptive spatial synchronization result.

[0046] Thirdly, the present invention provides a roadside multi-radar adaptive spatial synchronization device, comprising: a processor, a storage medium and a bus, wherein the storage medium stores machine-readable instructions executable by the processor, and when the roadside multi-radar adaptive spatial synchronization device is running, the processor communicates with the storage medium via the bus, and the processor executes the machine-readable instructions to perform the steps of the roadside multi-radar adaptive spatial synchronization method of the first aspect described above.

[0047] This invention provides a roadside multi-radar adaptive spatial synchronization method, apparatus, and device. The roadside multi-radar adaptive spatial synchronization method includes: S101, acquiring the current target coordinate data acquired by any current radar among the radars to be synchronized; S102, multiplying the current target coordinate data by the current spatial synchronization matrix of the current iteration to obtain transformed coordinate data; S103, performing association and splicing processing on the current target coordinate data and transformed coordinate data based on an association splicing algorithm to obtain the association result and successfully associated point pairs; S104, obtaining the coordinate error of the successfully associated point pairs according to a preset error calculation method, and obtaining the current frame average association error and the mean value of the current frame average error of the current iteration based on the coordinate error and the association result; S105, ... Determine whether the mean average error of the current frame and the mean average error of the previous frame in the previous iteration meet the iteration stopping condition; S106, if the condition in S105 is not met, update the current spatial synchronization matrix in S102 using the mean correlation error of the current frame and the mean average error of the current frame to obtain the updated spatial synchronization matrix, and use the updated spatial synchronization matrix as the current spatial synchronization matrix in S102; S107, repeat steps S101-S106 until the condition in S105 is met, use the current spatial synchronization matrix obtained in the most recent execution of S102 as the optimal spatial synchronization matrix, and use the optimal spatial synchronization matrix to obtain the multi-radar adaptive spatial synchronization result. In this invention, by iteratively updating the current spatial synchronization matrix and calculating the multi-radar adaptive spatial synchronization result based on the iterative result of the current spatial synchronization matrix, the problem of reduced accuracy of spatial synchronization results caused by environmental influences and slow parameter changes is solved. In addition, through automatic iteration, the coverage of the successfully associated point pairs used to calculate the current spatial synchronization matrix gradually approaches the complete sensing overlap area, improving the accuracy and stability of the final multi-radar adaptive spatial synchronization result. Since no high-precision map is required for subsequent processing, the implementation cost of multi-radar spatial synchronization is also reduced.

[0048] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0049] Figure 1 A flowchart illustrating a roadside multi-radar adaptive spatial synchronization method provided in an embodiment of the present invention;

[0050] Figure 2 This is a schematic diagram of the overall process of a roadside multi-radar adaptive spatial synchronization method provided in an embodiment of the present invention;

[0051] Figure 3 This is a schematic diagram of the structure of a roadside multi-radar adaptive spatial synchronization device provided in an embodiment of the present invention;

[0052] Figure 4 This is a schematic diagram of the structure of a roadside multi-radar adaptive spatial synchronization device provided in an embodiment of the present invention. Detailed Implementation

[0053] The present invention will be further described in detail below with reference to specific embodiments, but the implementation of the present invention is not limited thereto.

[0054] To improve the accuracy and stability of the final multi-radar adaptive spatial synchronization results, this invention provides a roadside multi-radar adaptive spatial synchronization method. Figure 1 This is a flowchart illustrating a roadside multi-radar adaptive spatial synchronization method provided in an embodiment of the present invention. Figure 1 As shown, the method includes:

[0055] S101. Obtain the current target coordinate data acquired by any of the radars to be synchronized.

[0056] Optionally, the roadside multi-radar adaptive spatial synchronization method prior to S101 also includes:

[0057] Based on the preset point pair acquisition method, a preset number of identical target point pairs detected by adjacent radars in the radar to be synchronized are acquired, and the identical target point pairs are combined into an initial point pair set; the identical target point pairs are the coordinate information of adjacent radars detecting the same target at the same time;

[0058] The initial spatial synchronization matrix is ​​calculated based on the initial point pair set and the homography matrix algorithm.

[0059] It should be noted that at least four pairs of identical target points are required to use a homography matrix, and these four pairs of identical target points cannot be collinear. Therefore, the preset number is generally greater than or equal to four.

[0060] The initial set of points can be represented as follows:

[0061]

[0062] Where, x nr Let y represent the x-coordinate of the nth point pair of the r-th radar. nr This represents the ordinate of the nth point pair of the r-th radar.

[0063] The initial spatial synchronization matrix H can be expressed as:

[0064]

[0065] Among them, X nr This represents the coordinate information (x) of the nth point pair of the rth radar. nr ,y nr ,1), by jointly calculating and solving the set of initial point pairs, the initial spatial synchronization matrix H can be finally obtained.

[0066] Optionally, the preset point-to-point acquisition methods include: static acquisition method and dynamic acquisition method;

[0067] The static acquisition method utilizes the coordinate information of a triangular reflector located at the apex of the intersection and splicing area of ​​adjacent radars, which is obtained from adjacent radars.

[0068] The dynamic acquisition method utilizes the coordinate information of the same moving ground target acquired by adjacent radars at the same time.

[0069] It should be noted that the static acquisition method has the advantages of stable coordinates and high accuracy, and can cover the entire stitching area as much as possible. Its disadvantages include the need for road closures, complex engineering arrangements, and difficulty in subsequent updates. The dynamic acquisition method does not require road closures, is simpler to select points, and allows for reselection based on results. Its disadvantages include the fact that, due to radar characteristics, the coordinates of moving targets have some error, resulting in slightly lower accuracy.

[0070] S102. Multiply the current target coordinate data by the current spatial synchronization matrix of the current iteration to obtain the transformed coordinate data.

[0071] It should be noted that, in this embodiment, the current target coordinate data can be mapped to another radar coordinate system by calculating the spatial synchronization matrix.

[0072] S103. Based on the association stitching algorithm, perform association stitching processing on the current target coordinate data and the transformed coordinate data to obtain the association results and successfully associated point pairs.

[0073] It should be noted that the main goal of the association stitching algorithm is to associate two target points of the same vehicle detected by two radars (i.e., the current target coordinate data and the transformed coordinate data), that is, to find which data point is the other target point of the same vehicle, and finally obtain a successfully associated point pair {(a1,b1),(a2,b2)}. Here, a1 represents the x-coordinate of the current target coordinate data in the successfully associated point pair, a2 represents the x-coordinate of the transformed coordinate data in the successfully associated point pair, b1 represents the y-coordinate of the current target coordinate data in the successfully associated point pair, and b2 represents the y-coordinate of the transformed coordinate data in the successfully associated point pair.

[0074] S104. Obtain the coordinate error of the successfully associated point pair according to the preset error calculation method, and obtain the average association error of the current frame and the mean value of the average error of the current frame in the current iteration based on the coordinate error and the association result.

[0075] Optionally, S104 specifically includes:

[0076] Use Euclidean distance to obtain the coordinate error of successfully associated point pairs;

[0077] The current coordinate error of the current frame is selected from the coordinate errors based on the correlation results;

[0078] The average correlation error of the current frame is obtained by averaging the current coordinate error.

[0079] The mean of the average correlation error of all current frames is calculated to obtain the mean of the average error of the current frame.

[0080] The current coordinate error E can be expressed as:

[0081]

[0082] Where a'2 is the result of a2 multiplied by the current spatial synchronization matrix, and b'2 represents the result of b2 multiplied by the current spatial synchronization matrix.

[0083] When the association results show that there are multiple successfully associated point pairs in the frame, the average association error E of all successfully associated point pairs in the frame can be calculated. frame , Among them, E i (i = 0 to n) represents the coordinate error of all successfully associated point pairs in this frame.

[0084] The mean of the average correlation error of all current frames is calculated to obtain the mean average error E of the current frame. syn .

[0085]

[0086] Where m represents m frames of data, E frame,i Represents E in the j-th frame frame .

[0087] S105. Determine whether the mean average error of the current frame and the mean average error of the previous frame in the previous iteration satisfy the iteration stopping condition.

[0088] Optionally, the iteration stopping condition is that the absolute value of the difference between the mean average error of the current frame and the mean average error of the previous frame is less than a preset threshold.

[0089] It should be noted that, in this embodiment, a condition for determining whether the current spatial synchronization matrix has been updated can also be added. When the condition in S105 is met, it can also be determined that:

[0090] Condition 1. Update whether the number of point pairs in the point pair set has reached a certain level;

[0091] Condition 2. The number of point pairs in the updated point pair set has not reached a certain number, but exceeds the set update frame threshold.

[0092] If the condition in S105 is met, and either condition 1 or condition 2 is satisfied, then the update process of the current spatial synchronization matrix will be executed.

[0093] Understandably, by adding conditions 1 and 2, frequent updates to the current spatial synchronization matrix can be avoided, greatly reducing the amount of data processed by the device.

[0094] S106. When the condition in S105 is not met, the current spatial synchronization matrix in S102 is updated by using the average correlation error of the current frame and the mean of the average error of the current frame to obtain the updated spatial synchronization matrix, and the updated spatial synchronization matrix is ​​used as the current spatial synchronization matrix in S102.

[0095] Optionally, S106 specifically includes:

[0096] If the condition of S105 is not met, select new associated point pairs that meet the requirements from the successfully associated point pairs based on the preset point pair selection formula.

[0097] Add the newly added associated point pairs to the initial point pair set to obtain the updated point pair set;

[0098] The current spatial synchronization matrix in S102 is updated based on the set of update point pairs to obtain the updated spatial synchronization matrix.

[0099] The updated spatial synchronization matrix is ​​used as the current spatial synchronization matrix in S102.

[0100] Optionally, the current spatial synchronization matrix in S102 is updated based on the set of updated point pairs to obtain an updated spatial synchronization matrix, including:

[0101] Multiple target point pairs for updating are selected from the set of updated point pairs using a random sampling consensus method.

[0102] Based on the update target point pairs and homography matrix algorithm, the update space synchronization matrix is ​​calculated.

[0103] Optionally, the formula for selecting preset point pairs is expressed as:

[0104] E τ <min(k1*Eframe ,k2*E syn ),(0 <k1,k2≤1);

[0105] Among them, E τ E represents the coordinate error of the newly added associated point pair. frame E represents the average correlation error of the current frame. syn k1 represents the mean error of the current frame. frame The error coefficient, k2 represents E syn The error coefficient is min(·,·), which represents the minimum value operation.

[0106] Additionally, it should be noted that in this embodiment, the convergence judgment of the adaptive algorithm can include multiple consecutive judgments on the spatial synchronization matrix error trend. Furthermore, a nonlinear function can be added to the selection rules for newly added associated point pairs. For example, the selection rules for newly added associated point pairs can also adopt:

[0107] E τ <f(E frame E syn );

[0108] Where f represents a nonlinear function.

[0109] S107. Repeat steps S101-S106 until the condition of S105 is met. Take the current spatial synchronization matrix obtained in the most recent execution of S102 as the optimal spatial synchronization matrix, and use the optimal spatial synchronization matrix to obtain the multi-radar adaptive spatial synchronization result.

[0110] It should be noted that most existing technologies do not support the problem of the original spatial synchronization parameters (original spatial synchronization matrix) becoming unusable due to environmental changes. In existing technologies, when the environment changes (due to changes in radar orientation, elevation, and installation height caused by long radar installation time), the radar installation position usually needs to be readjusted, which is time-consuming and labor-intensive. In this embodiment of the invention, the above-mentioned problems are avoided by iteratively optimizing the current spatial synchronization matrix, thus improving the efficiency and accuracy of the adaptive spatial synchronization results.

[0111] To illustrate the overall execution process of the roadside multi-radar adaptive spatial synchronization method provided in this embodiment of the invention, Figure 2 This is a schematic diagram illustrating the overall process of a roadside multi-radar adaptive spatial synchronization method provided in an embodiment of the present invention. Figure 2As shown, the system is first initialized by obtaining an initial set of point pairs using either static or dynamic point selection, and an initial spatial synchronization matrix is ​​derived based on this set. Then, based on the initial spatial synchronization matrix, the target information from multiple radar inputs is transformed for synchronization coordinates. The transformation results are then spatially synchronized with the radar input target information using an association stitching algorithm. Afterward, the coordinate error of the associated point pairs is calculated based on the association results to update the spatial synchronization parameters. Next, it is determined whether the adaptive algorithm is running. If it is, the convergence of the spatial synchronization matrix is ​​checked again (by comparing the average error mean of the current frame with the average error mean of the previous frame, and checking if the absolute value of the difference is less than a preset threshold). If convergence is achieved, the adaptive algorithm stops, and the optimal spatial synchronization matrix is ​​output. If convergence is not achieved, matching point pairs that meet the conditions are selected to update the matching point pair set. Then, based on conditions 1 and 2 mentioned above, it is determined whether the update conditions are met. If either condition 1 or condition 2 is met, the update processing steps for the current spatial synchronization matrix are executed, and the next iteration begins.

[0112] It should be noted that after obtaining the optimal spatial synchronization matrix, steps S101-S103 can be executed again to obtain the final multi-radar adaptive spatial synchronization result.

[0113] This invention provides a roadside multi-radar adaptive spatial synchronization method, including: S101, acquiring the current target coordinate data acquired by any current radar among the radars to be synchronized; S102, multiplying the current target coordinate data by the current spatial synchronization matrix of the current iteration to obtain transformed coordinate data; S103, performing association and splicing processing on the current target coordinate data and transformed coordinate data based on an association splicing algorithm to obtain the association result and successfully associated point pairs; S104, obtaining the coordinate error of the successfully associated point pairs according to a preset error calculation method, and obtaining the average association error of the current frame and the mean of the average error of the current frame of the current iteration based on the coordinate error and the association result; S1 05. Determine whether the mean average error of the current frame and the mean average error of the previous frame in the previous iteration meet the iteration stopping condition; S106. If the condition in S105 is not met, update the current spatial synchronization matrix in S102 using the mean correlation error of the current frame and the mean average error of the current frame to obtain the updated spatial synchronization matrix, and use the updated spatial synchronization matrix as the current spatial synchronization matrix in S102; S107. Repeat steps S101-S106 until the condition in S105 is met, use the current spatial synchronization matrix obtained in the most recent execution of S102 as the optimal spatial synchronization matrix, and use the optimal spatial synchronization matrix to obtain the multi-radar adaptive spatial synchronization result. In this embodiment of the invention, by iteratively updating the current spatial synchronization matrix and calculating the multi-radar adaptive spatial synchronization result based on the iterative result of the current spatial synchronization matrix, the problem of reduced accuracy of spatial synchronization results caused by environmental influences and slow parameter changes is solved. In addition, through automatic iteration, the coverage of the successfully associated point pairs used to calculate the current spatial synchronization matrix gradually approaches the complete sensing overlap area, improving the accuracy and stability of the final multi-radar adaptive spatial synchronization result. Since there is no need to use high-precision maps for subsequent processing, the implementation cost of multi-radar spatial synchronization is also reduced.

[0114] The method provided in this embodiment of the invention can be applied to electronic devices. Specifically, the electronic device can be a desktop computer, a portable computer, a smart mobile terminal, a server, etc., and this embodiment of the invention does not limit the application to such devices.

[0115] Corresponding to a roadside multi-radar adaptive spatial synchronization method, this embodiment of the invention also provides a roadside multi-radar adaptive spatial synchronization device. Figure 3 This is a schematic diagram of the structure of a roadside multi-radar adaptive spatial synchronization device provided in an embodiment of the present invention, as shown below. Figure 3 As shown, it includes:

[0116] The system includes an acquisition unit 301, a splicing unit 302, a calculation unit 303, a judgment unit 304, an update unit 305, and an iteration unit 306.

[0117] The acquisition unit 301 is used to acquire the current target coordinate data acquired by any of the radars to be synchronized.

[0118] Multiply the current target coordinate data by the current spatial synchronization matrix of the current iteration to obtain the transformed coordinate data;

[0119] The splicing unit 302 is used to perform splicing processing on the current target coordinate data and the transformed coordinate data based on the association splicing algorithm, and to obtain the association result and successfully associated point pairs;

[0120] The calculation unit 303 is used to obtain the coordinate error of the successfully associated point pair according to the preset error calculation method, and to obtain the average association error of the current frame and the mean value of the average error of the current frame in the current iteration based on the coordinate error and the association result.

[0121] The judgment unit 304 is used to determine whether the average error of the current frame and the average error of the previous frame in the previous iteration meet the iteration stop condition.

[0122] The updating unit 305 is used to update the current spatial synchronization matrix in the acquisition unit 301 by means of the current frame average association error and the mean of the current frame average error when the condition of the judgment unit 304 is not met, so as to obtain the updated spatial synchronization matrix and use the updated spatial synchronization matrix as the current spatial synchronization matrix in the acquisition unit 301.

[0123] The iteration unit 306 is used to jump to the acquisition unit 301 until the condition of the judgment unit 304 is met, take the current spatial synchronization matrix obtained in the most recent acquisition unit 301 as the optimal spatial synchronization matrix, and use the optimal spatial synchronization matrix to obtain the multi-radar adaptive spatial synchronization result.

[0124] Based on the same inventive concept, embodiments of the present invention also provide a roadside multi-radar adaptive spatial synchronization device. Figure 4 This is a schematic diagram of a roadside multi-radar adaptive spatial synchronization device provided in an embodiment of the present invention. It includes a processor 710, a storage medium 720, and a bus 730. The storage medium 720 stores machine-readable instructions executable by the processor 710. When the roadside multi-radar adaptive spatial synchronization device is running, the processor 710 communicates with the storage medium 720 via the bus 730. The processor 710 executes the machine-readable instructions to perform the steps of the above-described method embodiment. Specific implementation methods and technical effects are similar and will not be repeated here.

[0125] The storage medium may include random access memory (RAM) or non-volatile memory (NVM), such as at least one disk storage device. Optionally, the storage medium may also be at least one storage device located remotely from the aforementioned processor.

[0126] The processors mentioned above can be general-purpose processors, including central processing units (CPUs), network processors (NPs), etc.; they can also be digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0127] It should be noted that the terms "first," "second," etc., are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention. Rather, they are merely examples of apparatuses and methods consistent with some aspects of the invention.

[0128] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to specific features or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Furthermore, those skilled in the art can combine and integrate the different embodiments or examples described in this specification.

[0129] Although the invention has been described herein in conjunction with various embodiments, those skilled in the art will understand and implement other variations of the disclosed embodiments by reviewing the accompanying drawings and the disclosure in carrying out the claimed invention. In the description of the invention, the word "comprising" does not exclude other components or steps, "a" or "an" does not exclude a plurality, and "a plurality" means two or more, unless otherwise explicitly specified. Furthermore, while different embodiments may describe certain measures, this does not mean that these measures cannot be combined to produce good results.

[0130] The above description, in conjunction with specific preferred embodiments, provides a further detailed explanation of the present invention. It should not be construed that the specific implementation of the present invention is limited to these descriptions. For those skilled in the art, various simple deductions or substitutions can be made without departing from the concept of the present invention, and all such modifications and substitutions should be considered within the scope of protection of the present invention.

Claims

1. A roadside multi-radar adaptive spatial synchronization method, characterized in that, include: S101. Obtain the current target coordinate data acquired by any of the radars to be synchronized. S102. Multiply the current target coordinate data by the current spatial synchronization matrix of the current iteration to obtain the transformed coordinate data; S103. Based on the association stitching algorithm, perform association stitching processing on the current target coordinate data and the transformed coordinate data to obtain the association result and successfully associated point pairs; S104. Obtain the coordinate error of the successfully associated point pair according to the preset error calculation method, and obtain the average association error of the current frame and the mean value of the average error of the current frame in the current iteration based on the coordinate error and the association result. S105. Determine whether the mean average error of the current frame and the mean average error of the previous frame in the previous iteration satisfy the iteration stop condition. S106. When the condition in S105 is not met, the current spatial synchronization matrix in S102 is updated by the average correlation error of the current frame and the mean of the average error of the current frame to obtain an updated spatial synchronization matrix, and the updated spatial synchronization matrix is ​​used as the current spatial synchronization matrix in S102. S107. Repeat steps S101-S106 until the condition of S105 is met. Take the current spatial synchronization matrix obtained in the most recent execution of S102 as the optimal spatial synchronization matrix, and use the optimal spatial synchronization matrix to obtain the multi-radar adaptive spatial synchronization result. S105 specifically includes: If condition S105 is met, then the following update trigger conditions are determined: Condition 1: Check whether the number of point pairs in the updated point pair set has reached the preset threshold. Condition 2: The number of point pairs in the updated point pair set has not reached the preset number threshold, and the number of iteration frames has reached the preset update frame number threshold; When the condition of S105 is met, and either condition 1 or condition 2 is satisfied, the update process of the current spatial synchronization matrix is ​​executed. S106 specifically includes: If the condition of S105 is not met, a new associated point pair that meets the requirements is selected from the successfully associated point pairs based on the preset point pair selection formula. The newly added associated point pairs are added to the initial point pair set to obtain the updated point pair set; Based on the set of updated point pairs, the current spatial synchronization matrix in S102 is updated to obtain the updated spatial synchronization matrix; The updated spatial synchronization matrix is ​​used as the current spatial synchronization matrix in S102; The formula for selecting the preset point pair is expressed as follows: ； in, This indicates the coordinate error of the newly added associated point pair. This represents the average correlation error of the current frame. This represents the mean error of the current frame. express The error coefficient, express The error coefficient, This indicates the operation of taking the minimum value.

2. The roadside multi-radar adaptive spatial synchronization method according to claim 1, characterized in that, The roadside multi-radar adaptive spatial synchronization method described above (S101) also includes: Based on the preset point pair acquisition method, a preset number of identical target point pairs detected by adjacent radars in the radar to be synchronized are acquired, and the identical target point pairs are combined into an initial point pair set; the identical target point pairs are the coordinate information of the adjacent radars detecting the same target at the same time. Based on the initial set of point pairs and the homography matrix algorithm, the initial spatial synchronization matrix is ​​calculated.

3. The roadside multi-radar adaptive spatial synchronization method according to claim 2, characterized in that, The preset point pair acquisition method includes: static acquisition method and dynamic acquisition method; The static acquisition method utilizes the coordinate information of the triangular reflector located at the apex of the intersection and splicing area of ​​the adjacent radars, which is obtained from the adjacent radars. The dynamic acquisition method utilizes the coordinate information of the same moving ground target acquired by adjacent radars at the same time.

4. The roadside multi-radar adaptive spatial synchronization method according to claim 1, characterized in that, S104 specifically includes: The coordinate error of the successfully associated point pair is obtained using Euclidean distance; Based on the correlation result, the current coordinate error of the current frame is selected from the coordinate errors; The mean of the current coordinate error is calculated to obtain the average correlation error of the current frame; The mean of the average correlation error of the current frame is calculated by averaging all the current frame average correlation errors.

5. The roadside multi-radar adaptive spatial synchronization method according to claim 1, characterized in that, The iteration stopping condition is that the absolute value of the difference between the average error of the current frame and the average error of the previous frame is less than a preset threshold.

6. The roadside multi-radar adaptive spatial synchronization method according to claim 1, characterized in that, The step of updating the current spatial synchronization matrix in S102 based on the set of updated point pairs to obtain the updated spatial synchronization matrix includes: Multiple target point pairs for updating are selected from the set of updated point pairs using a random sampling consensus method. Based on the updated target point pair and the homography matrix algorithm, the updated space synchronization matrix is ​​calculated.

7. A roadside multi-radar adaptive spatial synchronization device, characterized in that, The roadside multi-radar adaptive spatial synchronization device includes: an acquisition unit, a stitching unit, a calculation unit, a judgment unit, an update unit, and an iteration unit; The acquisition unit is used to acquire the current target coordinate data acquired by any of the radars to be synchronized. Multiply the current target coordinate data by the current spatial synchronization matrix of the current iteration to obtain the transformed coordinate data; The stitching unit is used to perform association stitching processing on the current target coordinate data and the transformed coordinate data based on the association stitching algorithm, and obtain the association result and successfully associated point pairs; The calculation unit is used to obtain the coordinate error of the successfully associated point pair according to a preset error calculation method, and to obtain the average association error of the current frame and the mean of the average error of the current frame in the current iteration based on the coordinate error and the association result. The judgment unit is used to determine whether the average error of the current frame and the average error of the previous frame in the previous iteration satisfy the iteration stop condition. The updating unit is used to update the current spatial synchronization matrix in the acquisition unit by means of the current frame average association error and the mean of the current frame average error when the condition of the judgment unit is not met, so as to obtain an updated spatial synchronization matrix and use the updated spatial synchronization matrix as the current spatial synchronization matrix in the acquisition unit. The iterative unit is used to jump to the acquisition unit until the condition of the judgment unit is met, take the current spatial synchronization matrix obtained in the most recent acquisition unit as the optimal spatial synchronization matrix, and use the optimal spatial synchronization matrix to obtain the multi-radar adaptive spatial synchronization result. In the judgment unit, when the condition of the judgment unit is met, the following update trigger conditions are determined: Condition 1: whether the number of point pairs in the update point pair set has reached a preset number threshold; Condition 2: the number of point pairs in the update point pair set has not reached the preset number threshold, and the number of iteration frames has reached a preset update frame number threshold; when the condition of the judgment unit is met, and either condition 1 or condition 2 is satisfied, the update processing of the current spatial synchronization matrix is ​​executed. The update unit is specifically used to: when the condition of the judgment unit is not met, select a new associated point pair that meets the requirements from the successfully associated point pairs based on the preset point pair selection formula; add the new associated point pair to the initial point pair set to obtain the updated point pair set; update the current spatial synchronization matrix based on the updated point pair set to obtain the updated spatial synchronization matrix; and use the updated spatial synchronization matrix as the current spatial synchronization matrix. The formula for selecting the preset point pair is expressed as follows: ； in, This indicates the coordinate error of the newly added associated point pair. This represents the average correlation error of the current frame. This represents the mean error of the current frame. express The error coefficient, express The error coefficient, This indicates the operation of taking the minimum value.

8. A roadside multi-radar adaptive spatial synchronization device, characterized in that, include: The system includes a processor, a storage medium, and a bus. The storage medium stores machine-readable instructions executable by the processor. When the roadside multi-radar adaptive spatial synchronization device is in operation, the processor communicates with the storage medium via the bus, and the processor executes the machine-readable instructions to perform the steps of the roadside multi-radar adaptive spatial synchronization method as described in any one of claims 1-6.

Images