Millimeter-wave Radar Target Association Method and System Based on Multidimensional Similarity

Abstract

The present invention relates to the technical field of target detection, and provides a millimeter-wave radar target association method and system based on multi-dimensional similarity. The method includes: calculating a deviation distance between detection information and trajectory information for target state parameters; calculating a similarity between the detection information and the trajectory information of the target state parameters based on a probability distribution of the target state parameters on the deviation distance; performing matching based on the joint similarity of at least two of the target state parameters to obtain a target association result; the probability distribution is a piecewise function on the deviation distance. By calculating the similarity between the detection information and the trajectory information of the target state parameters through a piecewise probability distribution on the deviation distance, and performing matching based on the joint similarity of at least two target state parameters on this basis, the present invention can obtain a more accurate target association result with less computational resource requirements without introducing too many hyperparameters.

Description

Technical Field

[0001] The present invention relates to the technical field of target detection, and particularly to a millimeter-wave radar target association method and system based on multi-dimensional similarity. Background Art

[0002] Autopilot, also known as driverless, computer-driven, or wheeled mobile robots, is a cutting-edge technology that relies on computer and artificial intelligence technologies to complete a complete, safe, and effective driving without human operation.

[0003] In the 21st century, due to the continuous increase in automobile users, problems such as traffic congestion and safety accidents faced by road traffic have become increasingly serious. With the support of vehicle networking technology and artificial intelligence technology, autopilot technology can coordinate travel routes and planned times, thereby greatly improving travel efficiency and reducing energy consumption to a certain extent. Autopilot can also help avoid safety hazards such as drunk driving and fatigue driving, reduce driver errors, and improve safety. Therefore, autopilot has become a research and development focus in various countries in recent years.

[0004] As an automated vehicle, an autonomous vehicle can sense its environment and navigate without human operation. As a feasible hardware for autonomous driving environment perception, on-vehicle millimeter-wave radar can collect point cloud data of obstacles during driving. Further, based on the point cloud data, the states of obstacles, such as the positions, speeds, and sizes of multiple targets, can be analyzed.

[0005] For the target association task under the multi-target tracking framework, existing technologies usually use traditional distance metrics, such as IOU (Intersection over Union), center point distance, corner point (correlation point) distance, multi-dimensional Euclidean distance, Mahalanobis distance, etc., as the basis for target matching.

[0006] Due to problems such as many false detections in the millimeter-wave radar detection results, low distinguishability between false detection targets and real targets, large position errors, jumps, and jumps in target size parameters, the above existing technology methods cannot accurately and stably complete the matching.

[0007] In addition, there are also improved solutions that cascade the above existing technology methods or perform preliminary filtering by setting preconditions. However, due to the addition of many artificially set hyperparameters, the model robustness of such improved solutions still has relatively large problems.

[0008] Therefore, how to more accurately achieve target association has become an urgent technical problem in the industry. Summary of the Invention

[0009] The present invention provides a millimeter-wave radar target association method and system based on multi-dimensional similarity, which is used to solve the defects of many false detections and poor model robustness in the prior art and achieve more accurate target association.

[0010] The present invention provides a millimeter-wave radar target association method based on multi-dimensional similarity, including:

[0011] Calculate the deviation distance between the detection information and the trajectory information for the target state parameters;

[0012] Based on the probability distribution of the target state parameters on the deviation distance, calculate the similarity between the detection information and the trajectory information of the target state parameters;

[0013] Perform matching based on the joint similarity of at least two of the target state parameters to obtain a target association result;

[0014] The probability distribution is a piecewise function on the deviation distance.

[0015] According to the millimeter-wave radar target association method based on multi-dimensional similarity provided by the present invention, the step of performing matching based on the joint similarity of at least two of the target state parameters to obtain a target association result includes:

[0016] Calculate an intermediate value of the joint similarity based on the similarity between the detection information and the trajectory information of at least two of the target state parameters;

[0017] Correct the intermediate value of the joint similarity according to the prior probability distribution of the target radar cross section to obtain the joint similarity of at least two of the target state parameters;

[0018] Obtain a target association result according to the joint similarity.

[0019] According to the millimeter-wave radar target association method based on multi-dimensional similarity provided by the present invention, the step of calculating an intermediate value of the joint similarity based on the similarity between the detection information and the trajectory information of at least two of the target state parameters includes:

[0020] Correct the similarity between the detection information and the trajectory information of the target state parameters based on a stability coefficient to obtain a corrected similarity value;

[0021] Calculate an intermediate value of the joint similarity according to the corrected similarity values of at least two of the target state parameters.

[0022] According to the millimeter-wave radar target association method based on multi-dimensional similarity provided by the present invention, the probability distribution is a Gaussian distribution within a set segment; the similarity between the detection information and the trajectory information of the target state parameters is calculated based on the mean and standard deviation of the Gaussian distribution.

[0023] According to a millimeter-wave radar target association method based on multi-dimensional similarity provided by the present invention, the first standard deviation is greater than the second standard deviation;

[0024] The first standard deviation is the standard deviation of the Gaussian distribution of the set state parameters of the target in the overtaking area at the deviation distance;

[0025] The second standard deviation is the standard deviation of the Gaussian distribution of the set state parameters of the target in the non-overtaking area at the deviation distance;

[0026] The overtaking area is the area outside the lane where the reference vehicle is located and the distance from the reference vehicle is less than the set value; the reference vehicle is the vehicle where the millimeter-wave radar is located.

[0027] According to a millimeter-wave radar target association method based on multi-dimensional similarity provided by the present invention, the step of obtaining the target association result according to the joint similarity includes:

[0028] Obtaining a target detection distance matrix and a target trajectory distance matrix according to the joint similarity of at least two of the target state parameters;

[0029] Performing Hungarian matching on the target detection distance matrix and the target trajectory distance matrix to obtain a target association result.

[0030] The present invention also provides a millimeter-wave radar target association system based on multi-dimensional similarity, including:

[0031] A deviation module for calculating the deviation distance between the detection information and the trajectory information for the target state parameters;

[0032] A similarity module for calculating the similarity between the detection information and the trajectory information of the target state parameters based on the probability distribution of the target state parameters at the deviation distance;

[0033] A matching module for performing matching based on the joint similarity of at least two of the target state parameters to obtain a target association result;

[0034] The probability distribution is a piecewise function at the deviation distance.

[0035] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, the steps of any one of the above-mentioned millimeter-wave radar target association methods based on multi-dimensional similarity are implemented.

[0036] The present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the millimeter-wave radar target association method based on multi-dimensional similarity as described in any one of the above are implemented.

[0037] The present invention also provides a computer program product, including a computer program. When the computer program is executed by a processor, the steps of the millimeter-wave radar target association method based on multi-dimensional similarity as described in any one of the above are implemented.

[0038] The millimeter-wave radar target association method and system based on multi-dimensional similarity provided by the present invention calculate the similarity between the target state parameter detection information and the trajectory information through the piecewise probability distribution of the deviation distance, and on this basis, combine the joint similarity of at least two target state parameters for matching, which can obtain a more accurate target association result with less computational resource requirements without introducing too many hyperparameters. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0040] Figure 1 is a schematic flowchart of the millimeter-wave radar target association method based on multi-dimensional similarity provided by the present invention;

[0041] Figure 2 is a schematic structural diagram of the millimeter-wave radar target association system based on multi-dimensional similarity provided by the present invention;

[0042] Figure 3 is a schematic structural diagram of the electronic device provided by the present invention.

[0043] Reference numerals:

[0044] 1: deviation module; 2: deviation module; 3: matching module;

[0045] 310: processor; 320: communication interface; 330: memory;

[0046] 340: communication bus. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0047] To make the objectives, technical solutions, and advantages of the present invention clearer, the following will clearly and completely describe the technical solutions in the present invention in conjunction with the accompanying drawings in the present invention. Apparently, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present invention without creative efforts belong to the scope of protection of the present invention.

[0048] The following will describe Figure 1 the method for associating millimeter-wave radar targets based on multi-dimensional similarity of the present invention.

[0049] As Figure 1 shown, an embodiment of the present invention provides a method for associating millimeter-wave radar targets based on multi-dimensional similarity, including:

[0050] Step 101: Calculate the deviation distance between the detection information and the trajectory information for the target state parameters.

[0051] Step 103: Calculate the similarity between the detection information and the trajectory information of the target state parameters based on the probability distribution of the target state parameters on the deviation distance.

[0052] Step 105: Perform matching based on the joint similarity of at least two of the target state parameters to obtain the target association result.

[0053] The probability distribution is a piecewise function on the deviation distance.

[0054] In a preferred embodiment, for different target state parameters, within the same deviation distance segmentation interval, the probability distributions are independent. On this basis, matching based on the joint similarity of multi-dimensional (i.e., at least two) target state parameters can obtain more accurate results.

[0055] In another preferred embodiment, the probability distribution of the static target state parameters is different from that of the moving target state parameters. This embodiment can separately consider static targets and moving targets independently, so that the final association (matching) result is more accurate. At the same time, considering that the influence of static targets on the vehicle driving strategy is higher, the independent consideration scheme can further enhance the safety of subsequent driving strategies.

[0056] The beneficial effects of this embodiment are as follows:

[0057] By calculating the similarity between the detection information and the trajectory information of the target state parameters through the piecewise probability distribution on the deviation distance, and performing matching based on the joint similarity of at least two target state parameters on this basis, it is possible to obtain a more accurate target association result with less computational resource requirements without introducing too many hyperparameters.

[0058] According to the above embodiments, in this embodiment:

[0059] The step of matching based on the combined similarity of at least two of the target state parameters to obtain a target association result includes:

[0060] Calculating an intermediate combined similarity value based on the similarity between the detection information and the trajectory information of at least two of the target state parameters;

[0061] Correcting the intermediate combined similarity value according to the prior probability distribution of the target radar cross section to obtain the combined similarity of at least two of the target state parameters;

[0062] Obtaining a target association result according to the combined similarity.

[0063] The step of calculating an intermediate combined similarity value based on the similarity between the detection information and the trajectory information of at least two of the target state parameters includes:

[0064] Correcting the similarity between the detection information and the trajectory information of the target state parameters based on a stability coefficient to obtain a similarity correction value; the stability coefficient measures the degree of stable trajectory tracking and can ensure that trajectories with stable tracking can be preferentially matched.

[0065] Calculating an intermediate combined similarity value according to the similarity correction values of at least two of the target state parameters.

[0066] In this embodiment, since the radar cross section (RCS) is an attribute of the object itself, using the RCS value to correct the combined similarity can reduce the associated matching of trajectory information and misdetected targets (such as mirror targets, targets generated by multipath reflection, etc.).

[0067] Specifically, the statistical characteristics of RCS basically satisfy a bimodal distribution. Therefore, a bimodal distribution can be used as the prior probability of the radar cross section RCS (in a preferred scheme, it can also be described by an inverse gamma distribution). Thus, a probability density function can be constructed or generated through statistics to obtain the prior probability of the object.

[0068] The prior probability of RCS can reduce the associated similarity of misdetections (usually with abnormal RCS).

[0069] In addition, in order to make the targets with stable tracking more prioritized during association (to avoid new-introduced targets from preemptively matching and causing mutations), this embodiment also adopts the technical means of a stability coefficient.

[0070] Specifically, the stability coefficient approaches 1 as the number of successful matches increases. For example, the initial base value of the stability coefficient p can be set to p = 0.1, and after each successful association, p is updated to p = p^0.8.

[0071] According to any of the above embodiments, in this embodiment:

[0072] The probability distribution is a Gaussian distribution within a set segment; the similarity between the target state parameter detection information and the trajectory information is calculated based on the mean and standard deviation of the Gaussian distribution.

[0073] The first standard deviation is greater than the second standard deviation;

[0074] The first standard deviation is the standard deviation of the Gaussian distribution of the set state parameters of the target within the overtaking area at the deviation distance;

[0075] The second standard deviation is the standard deviation of the Gaussian distribution of the set state parameters of the target within the non - overtaking area at the deviation distance;

[0076] The overtaking area is the area outside the lane where the reference vehicle is located and at a distance less than a set value from the reference vehicle; the reference vehicle is the vehicle where the millimeter - wave radar is located.

[0077] That is to say, this embodiment focuses on processing the overtaking area, and the specific description is as follows.

[0078] Problems existing in the overtaking area:

[0079] The azimuth angle is very large, which may cause a large speed error;

[0080] It is at the handover point of the rear - facing radar (boundary - angle radar), and the position accuracy of the radar point is poor and the position is inaccurate;

[0081] The multipath interference in this area is serious, there are many false detections with speed, it is easy to introduce new detections, and there is a lack of historical information;

[0082] The state deviation conditions in the positive and negative directions of the overtaking area are different, and more refined parameter adjustment is required. Therefore, in this embodiment, for the state quantity with a large error in a specific area, its standard deviation is appropriately increased to make its similarity distribution smoother and reduce the false association caused by the mutation of a certain dimension.

[0083] Further, the step of obtaining the target association result according to the joint similarity includes:

[0084] Obtaining a target detection distance matrix and a target trajectory distance matrix according to the joint similarity of at least two of the target state parameters;

[0085] Performing Hungarian matching on the target detection distance matrix and the target trajectory distance matrix to obtain the target association result.

[0086] Among them, in a preferred embodiment, the distance elements in the target detection distance matrix and the target trajectory distance matrix take values of 1 - all_prob, where all_prob refers to the joint similarity (between the detection value and the trajectory value of the specified state parameter of the target).

[0087] It should be noted that in the embodiments of the present invention, the detection information (detection value) refers to the return value of the in - vehicle millimeter - wave radar at a set time; the trajectory information (trajectory value) refers to the historical information obtained based on the target detection / matching / correlation results before the set time.

[0088] The beneficial effects of this embodiment are as follows:

[0089] By processing the parameters of the overtaking area, a more practical probability distribution is obtained, thereby obtaining a more accurate target association result.

[0090] According to any of the above - mentioned embodiments, embodiments with feasible calculation methods will be provided below.

[0091] This embodiment includes the following steps:

[0092] 1. Calculate the distance between each single - dimension detection (detection information) and tracking (trajectory information), such as: center_x_distance = trk_x - det_x

[0093] 2. Obtain the distribution of each dimension in each distance interval according to the prior statistics

[0094] 3. Calculate the similarity of each dimension. For example, the similarity of the target center point position x is: x_prob = exp(-0.5 * pow(center_x_distance - x_mean, 2) / pow(x_std, 2))

[0095] 4. Calculate the multi - dimension joint similarity: all_prob = vel_prob * x_prob * y_prob * ext_x_prob * ext_y_prob * point_num_prob

[0096] 5. Since RCS is an attribute of the object itself, this can be used to reduce the association matching between trk and false detections (mirrors, multipaths):

[0097] At this time, the joint similarity is: all_prob = vel_prob * x_prob * y_prob * ext_x_prob * ext_y_prob * point_num_prob * rcs_prob

[0098] 6. At the same time, in order to make the target of stable tracking more prioritized during association, multiply each similarity by the stability coefficient of the trajectory, and this coefficient approaches 1 as the survival time of the trajectory increases.

[0099] The base of the stability coefficient p is 0.1, and each time the association is successful, it is updated to p = p^0.8.

[0100] At this time, the combined similarity is: all_prob = vel_prob * x_prob * y_prob * ext_x_prob * ext_y_prob * point_num_prob * rcs_prob * p

[0101] 7. Calculate the association distance = 1 - all_prob

[0102] 8. Repeat the above steps until the distance matrix corresponding to all detections and tracking targets is obtained, and then the final matching result is obtained through Hungarian matching.

[0103] In this embodiment, pow() is an exponential function; center_x_distance is the distance between the detection information and the trajectory information in the x - coordinate dimension; trk_x is the value of the x - coordinate in the trajectory information; det_x is the value of the x - coordinate in the detection information; gaussion N~(mean, std) is a Gaussian distribution with mean as mean and std as standard deviation; x_prob is the similarity in the x - coordinate dimension; x_mean is the mean of the Gaussian distribution in the x - coordinate dimension; x_std is the standard deviation of the Gaussian distribution in the x - coordinate dimension; all_prob is the combined similarity; vel_prob is the velocity similarity; y_prob is the similarity in the y - coordinate dimension; ext_x_prob is the similarity in the component dimension of the target size on the x - axis; ext_y_prob is the similarity in the component dimension of the target size on the y - axis; point_num_prob is the similarity of the number of points (the number of original radar points forming the detection).

[0104] Next, the millimeter - wave radar target association device based on multi - dimensional similarity provided by the present invention will be described. The millimeter - wave radar target association device described below can be correspondingly referred to the millimeter - wave radar target association method described above.

[0105] The embodiment of the present invention further provides a millimeter - wave radar target association system based on multi - dimensional similarity, including:

[0106] Deviation module 1, used to calculate the deviation distance between the detection information and the trajectory information for the target state parameters;

[0107] A similarity module 2, configured to calculate the similarity between the target state parameter detection information and the trajectory information based on the probability distribution of the target state parameter on the deviation distance;

[0108] A matching module 3, configured to perform matching based on the joint similarity of at least two of the target state parameters to obtain a target association result;

[0109] The probability distribution is a piecewise function on the deviation distance.

[0110] Further, the matching module 3 includes:

[0111] An intermediate value sub-module, configured to calculate an intermediate joint similarity value based on the similarity between the detection information and the trajectory information of at least two of the target state parameters;

[0112] A joint similarity sub-module, configured to correct the intermediate joint similarity value according to the prior probability distribution of the target radar cross section to obtain the joint similarity of at least two of the target state parameters;

[0113] An association sub-module, configured to obtain a target association result according to the joint similarity.

[0114] The intermediate value sub-module includes:

[0115] A stability coefficient unit, configured to correct the similarity between the detection information and the trajectory information of the target state parameter based on a stability coefficient to obtain a similarity correction value;

[0116] A corrected intermediate value unit, configured to calculate an intermediate joint similarity value according to the similarity correction values of at least two of the target state parameters.

[0117] The association sub-module includes:

[0118] A distance matrix unit, configured to obtain a target detection distance matrix and a target trajectory distance matrix according to the joint similarity of at least two of the target state parameters;

[0119] A Hungarian matching unit, configured to perform Hungarian matching on the target detection distance matrix and the target trajectory distance matrix to obtain a target association result.

[0120] Figure 3 Illustrates a schematic physical structure diagram of an electronic device, such as Figure 3As shown, the electronic device may include: a processor 310, a communications interface 320, a memory 330, and a communication bus 340. Among them, the processor 310, the communications interface 320, and the memory 330 complete communication with each other through the communication bus 340. The processor 310 may call the logical instructions in the memory 330 to execute the millimeter-wave radar target association method based on multi-dimensional similarity. The method includes: calculating the deviation distance between the detection information and the trajectory information for the target state parameters; calculating the similarity between the detection information and the trajectory information of the target state parameters based on the probability distribution of the target state parameters on the deviation distance; performing matching based on the joint similarity of at least two of the target state parameters to obtain a target association result; the probability distribution is a piecewise function on the deviation distance.

[0121] In addition, when the logical instructions in the above-mentioned memory 330 are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs that can store program codes.

[0122] On the other hand, the present invention also provides a computer program product. The computer program product includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the millimeter-wave radar target association method based on multi-dimensional similarity provided by the above-mentioned various methods. The method includes: calculating the deviation distance between the detection information and the trajectory information for the target state parameters; calculating the similarity between the detection information and the trajectory information of the target state parameters based on the probability distribution of the target state parameters on the deviation distance; performing matching based on the joint similarity of at least two of the target state parameters to obtain a target association result; the probability distribution is a piecewise function on the deviation distance.

[0123] In another aspect, the present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements the method for millimeter-wave radar target association based on multi-dimensional similarity provided by the above-mentioned various methods. The method includes: calculating the deviation distance between detection information and trajectory information for target state parameters; calculating the similarity between the detection information and trajectory information of the target state parameters based on the probability distribution of the target state parameters on the deviation distance; performing matching based on the joint similarity of at least two target state parameters to obtain a target association result; the probability distribution is a piecewise function on the deviation distance.

[0124] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art can understand and implement it without creative labor.

[0125] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the essence of the above technical solution, or the part that contributes to the prior art, can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0126] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A millimeter-wave radar target association method based on multi-dimensional similarity, characterized in that Including: Calculating a deviation distance between detection information and trajectory information for a target state parameter; Calculating a similarity between the detection information and the trajectory information of the target state parameter based on a probability distribution of the target state parameter on the deviation distance; Performing matching based on a joint similarity of at least two of the target state parameters to obtain a target association result; The probability distribution is a piecewise function on the deviation distance; The step of performing matching based on a joint similarity of at least two of the target state parameters to obtain a target association result includes: calculating an intermediate value of the joint similarity according to similarities between detection information and trajectory information of at least two of the target state parameters; Correcting the intermediate value of the joint similarity according to a prior probability distribution of a target radar cross section to obtain a joint similarity of at least two of the target state parameters; and obtaining a target association result according to the joint similarity.

2. The method for millimeter-wave radar target association based on multi-dimensional similarity according to claim 1, wherein The step of calculating an intermediate value of the joint similarity according to similarities between detection information and trajectory information of at least two of the target state parameters includes: Correcting the similarity between the detection information and the trajectory information of the target state parameter based on a stability coefficient to obtain a corrected similarity value; Calculating an intermediate value of the joint similarity according to corrected similarity values of at least two of the target state parameters.

3. The method for millimeter-wave radar target association based on multi-dimensional similarity according to claim 1, wherein The probability distribution is a Gaussian distribution within a set segment; and the similarity between the detection information and the trajectory information of the target state parameter is calculated based on a mean value and a standard deviation of the Gaussian distribution.

4. The method for millimeter-wave radar target association based on multi-dimensional similarity according to claim 3, characterized in that, A first standard deviation is greater than a second standard deviation; The first standard deviation is a standard deviation of a Gaussian distribution of a set state parameter of a target within an overtaking area on the deviation distance; The second standard deviation is a standard deviation of a Gaussian distribution of a set state parameter of a target within a non-overtaking area on the deviation distance; The overtaking area is an area outside the lane where a reference vehicle is located and at a distance less than a set value from the reference vehicle; and the reference vehicle is the vehicle where a millimeter-wave radar is located.

5. The method for millimeter-wave radar target association based on multi-dimensional similarity according to claim 1, characterized in that The step of obtaining a target association result according to the joint similarity includes: Obtaining a target detection distance matrix and a target trajectory distance matrix according to the joint similarity of at least two of the target state parameters; Performing Hungarian matching on the target detection distance matrix and the target trajectory distance matrix to obtain a target association result.

6. A millimeter-wave radar target association system based on multi-dimensional similarity, characterized in that, Including: A deviation module configured to calculate a deviation distance between detection information and trajectory information for a target state parameter; A similarity module configured to calculate a similarity between the detection information and the trajectory information of the target state parameter based on a probability distribution of the target state parameter on the deviation distance; A matching module configured to perform matching based on a joint similarity of at least two of the target state parameters to obtain a target association result, including: calculating an intermediate value of the joint similarity according to similarities between detection information and trajectory information of at least two of the target state parameters; correcting the intermediate value of the joint similarity according to a prior probability distribution of a target radar cross section to obtain a joint similarity of at least two of the target state parameters; and obtaining a target association result according to the joint similarity; The probability distribution is a piecewise function on the deviation distance.

7. An electronic device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein, When the processor executes the program, the steps of the millimeter-wave radar target association method based on multi-dimensional similarity according to any one of claims 1 to 5 are implemented.

8. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, the steps of the millimeter-wave radar target association method based on multi-dimensional similarity according to any one of claims 1 to 5 are implemented.

9. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, the steps of the millimeter-wave radar target association method based on multi-dimensional similarity according to any one of claims 1 to 5 are implemented.

Images