Doppler velocity filtering multipath false target suppression method, device and medium

Abstract

The invention discloses a Doppler velocity filtering multipath false target suppression method and device, and a medium, and belongs to the technical field of millimeter wave radar target detection. The method comprises the following steps: acquiring radar point cloud data; performing dynamic and static point cloud data separation on the radar original echo data to obtain a static point cloud set and a dynamic point cloud set; clustering the static point cloud set, and performing linear boundary fitting according to a clustering result to obtain a linear equation of a reflection boundary; doppler velocities of corresponding first-order and second-order multipath false targets are respectively estimated for trace points in the dynamic point cloud set, and corresponding Doppler velocity search intervals are obtained; traversing the dynamic point cloud set, and performing global point cloud Doppler velocity filtering on the obtained search interval to obtain a filtered dynamic point cloud set; and performing multi-dimensional information judgment on the filtered dynamic point cloud set, and identifying a multi-path false target so as to realize multi-path false target suppression. According to the method, Doppler information is effectively utilized, so that a multipath false target is effectively identified and suppressed.

Description

Technical Field

[0001] The present invention relates to the technical field of millimeter wave radar target detection, and in particular to a Doppler velocity filtering multipath false target suppression method, equipment and medium. Background Art

[0002] Millimeter-wave radar is widely used in the detection, positioning and identification of targets in complex environments. The path where electromagnetic waves are directly reflected from the radar and received by the radar after being emitted to the target is called the direct-line path. However, due to the presence of various static reflective surfaces in the environment, the electromagnetic waves are reflected multiple times by the reflective surfaces during propagation, resulting in the presence of multipath signal components in addition to the direct-line path signal when detecting and identifying real targets. Since the energy of millimeter-wave radar is greatly attenuated after multiple reflections, only the first and second reflections are considered. When detecting real targets, the false alarms caused by the first-order and second-order multipath false targets formed by the static reflective surfaces in the environment by the electromagnetic wave signals cause great interference to the identification and tracking of real targets. Therefore, it is of great significance to identify and suppress multipath false targets caused by static reflective surfaces under millimeter-wave radar detection.

[0003] In recent years, many research institutions at home and abroad have carried out research on multipath false target identification and suppression methods under vehicle-mounted millimeter-wave radar and traffic radar detection. The first existing technical solution proposes a multipath suppression method based on the correlation between true and false target trajectories. This method uses the correlation between the tracked target and its mirror false target trajectory, evaluates the correlation between candidate track pairs through three criteria, and classifies the false target tracks, thereby completing the suppression of multipath false targets. The evaluation criteria of this method rely on the track correlation of the matching pairs of real targets and false targets. When the amount of point cloud data is large, due to the high complexity of the algorithm, its computational complexity is large and it is not suitable for multi-target scenes in complex scenes. The second existing technical solution proposes a multipath model and false target suppression method based on millimeter-wave road monitoring radar. This method extracts eight features such as point track distance, azimuth, speed, signal-to-noise ratio based on the detection point cloud data and classifies them based on a fully connected neural network, achieving a multipath false target recognition rate of 95.74%, and an overall recognition rate of 90.70%, achieving effective detection and positioning of vehicle targets. The above methods need to obtain the true value through the assistance of sensors such as cameras, and complete the data set preparation through manual annotation methods. The workload is large and there are certain errors. The complexity of the multi-dimensional feature data set and the neural network structure are both restrictive factors for the real-time performance of road radar detection. Therefore, it is of great value to study the identification and suppression methods of multipath false targets caused by static reflective surfaces in millimeter-wave radar detection. Summary of the invention

[0004] In order to solve one of the technical problems existing in the prior art to at least a certain extent, the object of the present invention is to provide a Doppler velocity filtering multipath false target suppression method, device and medium based on reflection boundary fitting.

[0005] The first technical solution adopted by the present invention is:

[0006] A Doppler velocity filtering multipath false target suppression method comprises the following steps:

[0007] Get radar point cloud data;

[0008] Based on Doppler velocity information, the radar original echo data is separated into static and dynamic point cloud data to obtain static point cloud sets and dynamic point cloud sets;

[0009] Cluster the static point cloud set, perform linear boundary fitting based on the clustering results, and obtain the straight line equation of the reflection boundary;

[0010] Based on the multipath propagation model, the Doppler velocities of the first-order multipath false targets and the second-order multipath false targets are estimated for the points in the dynamic point cloud set, and the corresponding Doppler velocity search interval is obtained.

[0011] Traversing the dynamic point cloud set, performing global point cloud Doppler velocity filtering on the obtained search interval, and obtaining a filtered dynamic point cloud set;

[0012] Multi-dimensional information judgment is performed on the filtered dynamic point cloud set to identify multipath false targets in order to achieve multipath false target suppression.

[0013] Furthermore, the expression of the radar point cloud data is:

[0014]

[0015] In the formula, p g is the g-th point in the point cloud set Ω; x g ,y g represents the position coordinates of the g-th point; r g Indicates the radial distance of the g-th point; represents the Doppler velocity of the g-th point; v g represents the speed of the g-th point; θ g Represents the azimuth of the g-th point; rcs g represents the scattering area of ​​the g-th point; snr g represents the signal-to-noise ratio of the g-th point; power g Represents the signal power of the g-th trace; N is the number of traces.

[0016] Furthermore, clustering the static point cloud set and performing linear boundary fitting according to the clustering result to obtain the straight line equation of the reflection boundary includes:

[0017] The static point cloud set is clustered using a density-based clustering method, and the reflection boundary is fitted using Hough transform in each cluster to obtain the straight line equation of the reflection boundary.

[0018] Furthermore, the method of clustering the static point cloud set using a density-based clustering method and fitting the reflection boundary using Hough transform in each cluster includes:

[0019] Perform DBSCAN processing on the static point cloud of each frame to obtain M static point cloud clusters;

[0020] For all data points (x i ,y i )Use the following formula to get the Hesse normal:

[0021] ρ=x i cosθ+y i sinθ

[0022] In the formula, ρ represents the distance from the origin to the point (x i ,y i ), θ represents the rotation angle of the vertical line, and according to the distribution characteristics of the static point cloud, the corresponding range of ρ is determined (ρ min ,ρ max ), the range of θ (θ min ,θ max );

[0023] Get all the curves corresponding to the data points in the cluster to form a set C;

[0024] Estimate the intersection point (θ) of the curve set C corresponding to each cluster point set j ,ρ j );

[0025] Finally, the parameters of the fitted line are estimated:

[0026] Assuming that in a Cartesian coordinate system, the equation of the fitted straight line of the reflection boundary is:

[0027] Ax+By+C=0

[0028] Based on the estimated intersection coordinates (θ j ,ρ j ), convert the Hesse normal parameters to general line equation parameters:

[0029]

[0030] For all data points (x i ,y i ), based on the fitting straight line equation, all static point cloud clusters are traversed to obtain the fitting result of each point cloud cluster, thereby realizing the estimation of the static reflection boundary.

[0031] Furthermore, the Doppler velocity of the second-order multipath false target is estimated as follows:

[0032] Assume that the target point coordinates in the dynamic point cloud set D are (x t ,y t ), the parameters A, B, C are obtained based on the fitted straight line equation, and the second-order false target coordinates (x ts ,y ts ):

[0033]

[0034] Known target point (x t ,y t ) About the second-order false target point (x) fitting the reflection boundary ts ,y ts ), and obtain the theoretical second-order false target point (x ts ,y ts ) is:

[0035]

[0036] According to the differential relationship between Doppler velocity and radial distance in radar signal processing, the theoretical second-order false target point (x ts ,y ts ) is:

[0037]

[0038] The Doppler velocity of the first-order multipath false target is estimated as follows:

[0039] According to the multipath propagation geometric model, the real target point trace generates first-order false targets M1,,M2 respectively about the static reflection boundary; Assuming that the coordinates of the target point trace in the dynamic point cloud set D are (x t ,y t ), the coordinates of the second-order false target point are (x ts ,y ts ), according to the electromagnetic wave multipath propagation model, the theoretical radial distance of the first-order false targets M1 and M2 is equal and recorded as r M :

[0040]

[0041] where r ts is the theoretical radial distance of the second-order false target, r t is the radial distance of the target point;

[0042] At the same time, the theoretical Doppler velocity of the first-order false targets M1 and M2 is:

[0043]

[0044] In the formula, (x t ,y t ), (x ts ,y ts ) are the target point coordinates and the second-order false target coordinates.

[0045] Further, obtaining the corresponding Doppler velocity search interval includes:

[0046] Assume that the dynamic target point is p i ∈D, D is a dynamic point cloud set, p i ={x i ,y i ,r i ,v Di ,v i ,θ i ,rcs i ,snr i ,power i}, calculate the target point trace p respectively i Position coordinates of the theoretical second-order false target S obtained by fitting the reflection boundary Radial distance Radial velocity Search interval And the radial distance of the first-order false targets M1 and M2 Radial velocity Search interval Based on the search interval Get the Doppler velocity search interval δ of the target point i for:

[0047]

[0048] Furthermore, the global point cloud Doppler velocity filtering is performed on the obtained search interval to obtain a filtered dynamic point cloud set, including:

[0049] Doppler velocity filtering in the dynamic point cloud set D will satisfy the search interval δ i The point set of

[0050]

[0051] For the point trace p in the dynamic point cloud set D i , get the point trace set about this point Then all the points in the set satisfy the Doppler velocity constraint.

[0052] Furthermore, multi-dimensional information decision making is realized by constructing a multi-dimensional information decision judger, specifically:

[0053] The 3D feature vector is defined as follows:

[0054]

[0055] Among them, rcs is the scattering area of ​​the point cloud data, snr is the point cloud signal-to-noise ratio, and power is the signal power;

[0056] Define the weight vector in the 3D information decision maker for:

[0057]

[0058] The weighted sum of the output op of the decision maker is calculated as:

[0059]

[0060] The decision maker is used to calculate the output of the decision maker in the point cloud traces of the dynamic target and the point cloud set traces after Doppler velocity filtering, thereby realizing the identification of multipath false targets.

[0061] The second technical solution adopted by the present invention is:

[0062] An electronic device comprises a processor and a memory, wherein the memory stores at least one instruction, at least one program, a code set or an instruction set, and the at least one instruction, at least one program, the code set or the instruction set is loaded and executed by the processor to implement a Doppler velocity filtering multipath false target suppression method as described above.

[0063] The third technical solution adopted by the present invention is:

[0064] A computer-readable storage medium stores at least one instruction, at least one program, a code set or an instruction set, wherein the at least one instruction, the at least one program, the code set or the instruction set is loaded and executed by a processor to implement a Doppler velocity filtering multipath false target suppression method as described above.

[0065] The fourth technical solution adopted by the present invention is:

[0066] A computer program product or a computer program includes computer instructions stored in a computer-readable storage medium. A processor of a computer device can read the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes the above method.

[0067] The beneficial effects of the present invention are as follows: the present invention proposes a Doppler velocity filtering multipath false target suppression scheme based on reflection boundary fitting, which can more accurately fit the static reflection surface of the multipath signal based on the boundary fitting algorithm of the minimum variance Hough transform of the static radar point cloud data in millimeter wave radar detection, and filter the Doppler velocity dimension of the dynamic point cloud based on the multipath propagation model, effectively utilize Doppler information, thereby effectively identifying and suppressing multipath false targets, and suppressing false targets based on the multidimensional information of the point cloud data without causing the loss of the real target. Compared with the multipath suppression method mentioned in the technical background, the present invention can make full use of the point cloud data information to filter in the Doppler velocity dimension in complex environment radar detection, and has been verified in the measured data, and can effectively realize the identification and suppression of multipath false targets. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the embodiments of the present invention or the drawings of related technical solutions in the prior art are introduced below. It should be understood that the drawings introduced below are only for the convenience of clearly describing some embodiments of the technical solutions of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0069] Figure 1 It is a flow chart of a Doppler velocity filtering multipath false target suppression method based on reflection boundary fitting in an embodiment of the present invention;

[0070] Figure 2 A schematic diagram of a radar detection scene in an embodiment of the present invention;

[0071] Figure 3 A schematic diagram of a static road guardrail reflection geometric model in an embodiment of the present invention;

[0072] Figure 4 This is a schematic diagram of Doppler velocity under radar detection in an embodiment of the present invention; wherein, Figure 4 (a) is the side view of radar detection. Figure 4 (b) is a top view of radar detection;

[0073] Figure 5 This is a measured scene diagram in an embodiment of the present invention;

[0074] Figure 6 Schematic diagram of a multipath false area of ​​a static reflection surface in an embodiment of the present invention;

[0075] Figure 7 Schematic diagram of the measured data processing results in an embodiment of the present invention; wherein, Figure 7 (a) is a single-frame point cloud in the measured scene; Figure 7 (b) is the single-frame point cloud recognition result in the measured scene; Figure 7 (c) Single-frame point cloud suppression result in the measured scene. DETAILED DESCRIPTION

[0076] The embodiments of the present invention are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and are not to be construed as limitations of the present invention. For the step numbers in the following embodiments, they are only provided for the convenience of explanation, and the order between the steps is not limited in any way, and the execution order of each step in the embodiment can be adaptively adjusted according to the understanding of those skilled in the art.

[0077] In the description of the present invention, it should be understood that descriptions involving orientations, such as up, down, front, back, left, right, etc., and orientations or positional relationships indicated are based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present invention.

[0078] In the description of the present invention, "several" means one or more, "more" means more than two, "greater than", "less than", "exceed" etc. are understood as not including the number itself, and "above", "below", "within" etc. are understood as including the number itself. If there is a description of "first" or "second", it is only used for the purpose of distinguishing the technical features, and cannot be understood as indicating or implying the relative importance or implicitly indicating the number of the indicated technical features or implicitly indicating the order of the indicated technical features.

[0079] In the description of the present invention, unless otherwise clearly defined, terms such as setting, installing, connecting, etc. should be understood in a broad sense, and technicians in the relevant technical field can reasonably determine the specific meanings of the above terms in the present invention based on the specific content of the technical solution.

[0080] In view of the existing technical problems, the present invention provides a Doppler velocity filtering multipath false target suppression scheme based on reflection boundary fitting, which is suitable for the identification and suppression of first-order and second-order multipath false targets caused by static reflection surfaces in millimeter wave radar detection. Firstly, dynamic and static point clouds are separated based on the characteristics of radar point cloud data, and reflection boundary fitting is performed on static point clouds; secondly, modeling is performed based on a typical multipath propagation geometric model, and the Doppler velocity characteristics and geometric model of the real target are used to obtain the Doppler velocity of the first-order multipath false target and the second-order false target respectively; finally, the dynamic point cloud is filtered based on the Doppler velocity and the multi-domain information is jointly judged, thereby realizing the identification and suppression of multipath false targets.

[0081] Figure 2 is a schematic diagram of the millimeter wave radar detection scene, and the geometric model of the scene diagram is abstracted as follows Figure 3 As shown, Figure 5 The following is a diagram of the actual measured scene. The radar O is located directly above the middle of the road, and the target T is located between the left railing P1 and the right railing P2. Since the railing is a strong reflective surface, in addition to the direct detection signal of the target, the electromagnetic wave signal also has multipath false targets formed by the primary and secondary reflection signals. The propagation model is established based on the reflection characteristics of electromagnetic waves, which mainly includes the direct path, the primary reflection of the railings on both sides, and the secondary reflection path.

[0082] The right railing serves as a strong reflection boundary, and the specific propagation path of the electromagnetic wave signal is:

[0083] Direct line of sight path (l1): O→T→O, the electromagnetic wave signal is transmitted from radar O to the real target T and then returns to radar O.

[0084] First-order multipath (l2): O→P2→T→O, the electromagnetic wave signal is transmitted from radar O to the right railing P2, reflected to the real target T and then returned to radar O. Under this path, a first-order false target M is formed on the line connecting radar O and the right railing P2. 21 .

[0085] First-order multipath (l3): O→T→P2→O, the electromagnetic wave signal is transmitted from radar O to the real target T, reflected to the right railing P2, and then returned to radar O. This path forms a first-order false target M at the same azimuth as the real target T. 22 .

[0086] Second-order multipath (l4): O→P2→T→P2→O, the electromagnetic wave signal is transmitted from the radar O to the real target T, reflected to the right railing P2, then reflected to the real target T, and then returned to the radar O. This path generates a second-order false target S2 at the position where the real target is mirror-symmetrical about the right railing P2.

[0087] The left railing serves as a strong reflection boundary, and the specific propagation path of the electromagnetic wave signal is:

[0088] Direct line of sight path (l5): O→T→O, the electromagnetic wave signal is transmitted from radar O to the real target T and then returns to radar O.

[0089] First-order multipath (l6): O→P1→T→O, the electromagnetic wave signal is transmitted from radar O to the left railing P1, reflected by the real target T and then returned to radar O. Under this path, a first-order false target M is formed on the line connecting radar O and the left railing P1. 12 .

[0090] First-order multipath (l7): O→T→P1→O, the electromagnetic wave signal is transmitted from radar O to the real target T, reflected to the left railing P1, and then returned to radar O. This path forms a first-order false target M at the same azimuth as the real target T. 11 .

[0091] Second-order multipath (l8): O→P1→T→P1→O, the electromagnetic wave signal is transmitted from the radar O to the real target T, reflected to the left railing P1, then reflected to the real target T, and then returned to the radar O. This path generates a second-order false target S1 at the position where the real target is mirror-symmetrical about the left railing P1.

[0092] Taking into account the strong energy attenuation of millimeter-wave radar signals after multiple reflections, the energy of high-order path signals of multiple reflections is weak and can be ignored. Only the real targets formed by the first and second reflections and the first-order and second-order false targets are considered.

[0093] Based on the modeling of electromagnetic wave propagation signals in the above radar detection scenario, the contents of the embodiments of the present invention are specifically as follows.

[0094] Example 1

[0095] like Figure 1 As shown, this embodiment provides a Doppler velocity domain filtering multipath false target suppression method based on reflection boundary fitting, comprising the following steps:

[0096] S1: Raw radar echo signal processing.

[0097] According to the original radar echo, fast Fourier transform is performed in the fast time dimension and slow time dimension respectively, and constant false alarm detection and arrival angle estimation are performed to obtain the target distance, Doppler velocity, angle and other information, and then obtain point cloud data.

[0098] The top view of the millimeter wave radar detection scene is as follows: Figure 4 As shown in (b), assuming that the radar installation height is h, the radial distance of the target in the radar point cloud is measured to be R, and the radial velocity is v D', the angle between the radial distance and the radar normal is θ, and the scattering area rcs of the target and the signal-to-noise ratio snr of the point cloud can also be obtained. The coordinate transformation from three-dimensional space to two-dimensional space is realized for the radar point cloud, and the radial distance R in three-dimensional space is projected into two-dimensional space as follows:

[0099]

[0100]

[0101] Where r is the radial distance projected into two-dimensional space, and α is the defined altitude angle.

[0102] Based on the azimuth angle θ of the target point cloud detected by the radar and the radial distance r projected into the two-dimensional space, the position coordinates (x, y) of the target are obtained as follows:

[0103]

[0104] Since the millimeter wave radar has a certain installation height h, the radial velocity v in three-dimensional space D' Projected into two-dimensional space:

[0105] v D =v D' sinα

[0106] where v D is the Doppler velocity in two-dimensional space. Given the moving direction of the target under millimeter-wave radar detection, the moving velocity v of the target is:

[0107]

[0108] Assume that the point cloud data set is Ω, which has N points recorded as:

[0109] Ω={p g |x g ,y g ,r g ,v Dg ,v g ,θ g ,rcs g ,snr g ,power g}(g=1,2,……N)

[0110] Among them, p g is the g-th point in the point cloud set Ω, x g ,y g ,r g ,v Dg ,v g ,θ g ,rcsg ,snr g ,power g They respectively represent the position coordinates, radial velocity, movement velocity, azimuth angle, scattering area, signal-to-noise ratio, and signal power of the g-th point.

[0111] S2: Separation of static and dynamic point clouds. The radar point cloud data obtained in step S1 is separated into static and dynamic point cloud data based on Doppler velocity information to obtain static and dynamic point cloud sets respectively.

[0112] In some embodiments, the radar point cloud set Ω obtained by step S1 is assumed to be a static point cloud set S and contains L points, and a dynamic point cloud set D and contains H points. Considering that there is a certain error in radar speed measurement, the Doppler speed in the point cloud set is 0, and the Doppler speed interval Δv is set. D , according to the empirical value, it is set to 0.3m / s. Separate the dynamic and static point clouds. The separation basis of dynamic and static point clouds is:

[0113]

[0114] The dynamic point cloud set D and the static point cloud set S constitute the point cloud set Ω:

[0115] Ω=S∪D

[0116] S3: Static point cloud clustering and boundary fitting. Density-based clustering is implemented for static point cloud sets, and linear boundary fitting is performed to obtain the straight line equation of the reflection boundary.

[0117] Based on step S2, a static point cloud set S is obtained. In this embodiment of the present invention, DBSCAN clustering is used for the point cloud data in the set S, where the field search radius Eps is set to 4m and the minimum number of points in the neighborhood Minpts is set to 1. DBSCAN processing is performed on each frame of static point cloud. Assume that M static point cloud clusters are obtained, which are recorded as cluster i (i=1,2,3....M). For each static point cloud cluster i The data points (x i ,y i ), and perform boundary fitting. As a common straight line detection algorithm, the least square method has certain drawbacks, such as unsatisfactory curve fitting effect and high sensitivity to noise points. Therefore, the embodiment of the present invention proposes a boundary fitting method based on the minimum variance Hough transform, which can effectively perform linear fitting of the reflection boundary.

[0118] First, coordinate transformation is performed for all data points (x i ,y i )Use the following formula to get the Hesse normal:

[0119] ρ=x i cosθ+y i sinθ

[0120] Where ρ represents the origin to the point (x i ,y i ), θ represents the rotation angle of the vertical line, and according to the distribution characteristics of the static point cloud, the corresponding range of ρ is determined (ρ min ,ρ max ), the range of θ (θ min ,θ max ).

[0121] Secondly, get the data points (x i ,y i ) form a set C. In the Cartesian coordinate system, the data points that are approximately on a straight line and the curves corresponding to the points in the polar coordinate system intersect at one point (ρ j ,θ j ), which represents the corresponding parameter of the fitted line.

[0122] Again, find the intersection of the curve set C corresponding to each cluster point set. Assume that cluster i i There are m points under this condition. First, in the polar coordinate system, the interval (θ min ,θ max ) discretize to get N discrete values, and calculate each discrete value The ρ of all Hesse curves under (n=1,....,N), further calculate the variance of P, and select the angle corresponding to the minimum variance as the angle estimate θ j , (theoretically, at the intersection of the Hesse curve, the above variance value is 0). Further, estimate ρ j , the method is as follows:

[0123] Record the corresponding curve coordinates at this angle as ρ 1j ,ρ 2j ......ρ mj , find the approximate coordinates of the intersection of the curve family (θ j ,ρ j ):

[0124]

[0125] At this point, the intersection coordinates in the polar coordinate system are estimated: (θ j ,ρ j ).

[0126] Finally, the parameters of the fitting line are estimated. Assume that in the Cartesian coordinate system, the equation of the reflection boundary approximation line is:

[0127] Ax+By+C=0

[0128] Based on the above estimation, the intersection coordinates are estimated (θ j ,ρ j ), convert the Hesse normal parameters to general line equation parameters:

[0129]

[0130] For all data points (x i ,y i ), the fitting straight line equation is obtained based on the above S, and all static point cloud clusters are traversed to obtain the fitting result of each point cloud cluster, so as to estimate the static reflection boundary, and then estimate the Doppler velocity of multipath false targets based on the fitting boundary.

[0131] S4: Doppler velocity estimation of multipath false targets.

[0132] See also Figure 6 Based on the multipath propagation model, the Doppler velocity of the corresponding first-order multipath false target and the Doppler velocity of the second-order multipath false target are estimated for the target points in the dynamic point cloud, and the corresponding Doppler velocity search interval is obtained. The details are as follows:

[0133] 1) Second-order false target Doppler velocity estimation

[0134] Assume that the target point coordinates in the dynamic point cloud set D are (x t ,y t ), based on the parameters A, B, C obtained from the linear equation fitted in step S3, and the second-order false target coordinates (x ts ,y ts ):

[0135]

[0136] Known target point (x t ,y t ) About the second-order false target point (x) fitting the reflection boundary ts ,y ts ), we can get the theoretical second-order false target point (x ts ,y ts ) is:

[0137]

[0138] According to the differential relationship between Doppler velocity and radial distance in radar signal processing, the theoretical second-order false target point (x ts ,y ts ) is:

[0139]

[0140] Since there is a certain error in radar signal detection, the theoretical second-order false target point Doppler velocity is obtained. Set the search interval δ1, its threshold radius 0.1m / s.

[0141] 2) First-order false target Doppler velocity estimation

[0142] According to the multipath propagation geometric model, the real target point traces generate first-order false targets M1, M2 respectively about the static reflection boundary. Assume that the coordinates of the target point traces in the dynamic point cloud set D are (x t ,y t ), the coordinates of the second-order false target point obtained in 1) are (x ts ,y ts ), according to the electromagnetic wave multipath propagation model, the theoretical radial distance of the first-order false targets M1 and M2 is equal and recorded as r M :

[0143]

[0144] where r ts is the theoretical radial distance of the second-order false target obtained in 1), r t is the radial distance of the target point.

[0145] At the same time, the theoretical Doppler velocity of the first-order false targets M1 and M2 is:

[0146]

[0147] in are the radial distances of the first-order false targets, (x t ,y t ), (x ts ,y ts ) are the target point trace and the coordinates of the second-order false target. Since there is a certain error in radar signal detection, the Doppler velocity of the theoretical first-order false target M1, M2 is obtained Set the search interval δ2, whose threshold radius is Set to 0.1m / s.

[0148] S5: Doppler velocity filtering of dynamic point cloud. Perform step S4 on each point in the dynamic point cloud set, perform global point cloud Doppler velocity filtering on the obtained search interval, and obtain a dynamic point cloud set that meets the conditions.

[0149] Assume that the dynamic target point is p i ∈D,p i ={x i ,y i ,r i ,v Di ,v i ,θ i ,rcs i ,snr i ,power i}, according to the method in step S4, the target point track p is calculated by using the radial velocity, radial distance, azimuth and other information of the target point track i The position coordinates of the theoretical second-order false target S obtained by fitting the reflection boundary in step S3 Radial distance Radial velocity

[0150] Search interval And the radial distance of the first-order false targets M1 and M2 Radial velocity Search interval Based on the search interval Get the Doppler velocity search interval δ of the target point i for:

[0151]

[0152] Doppler velocity filtering in the dynamic point cloud set D will satisfy the search interval δ i The point set of

[0153]

[0154] For the point trace p in the dynamic point cloud set D i , get the point trace set about this point Then all the points in the set satisfy the constraint of Doppler velocity, and the points are judged in step S6.

[0155] Since millimeter-wave radar detection has the characteristics of high resolution and wide field of view, in the measured data processing, the left and right lane targets are processed separately according to the Doppler velocity direction of the dynamic point cloud target, and multipath propagation geometric modeling is performed on the vehicle targets on each side and the nearby road guardrails.

[0156] S6: Multi-dimensional information joint decision. Multi-dimensional information judgment is performed on the original single point trace in step S5 and the dynamic point cloud set that meets the conditions. The dynamic point cloud point traces that meet the conditions are regarded as multipath false targets. Steps S4 to S6 are repeated to complete the identification and suppression of multipath false targets in all dynamic point clouds.

[0157] The multi-dimensional information decision-making judger uses systems and models that process and analyze multi-dimensional information to make decisions and classify prediction results. Each dimension usually represents a factor or feature that affects the decision. The accuracy of the decision is improved by making a joint decision based on multi-dimensional information. This embodiment proposes a Doppler velocity filtering multipath false target suppression method based on reflection boundary fitting. Based on the above steps, a filtered point set is obtained, which contains the scattering area rcs, signal-to-noise ratio snr, and signal power information of the point cloud. According to the detection characteristics of electromagnetic waves for targets, the target scattering area rcs, signal-to-noise ratio snr, and signal power of the multipath false targets generated by multiple reflections are much smaller than the corresponding characteristic information of the real target. Based on this feature, the dynamic point trace p in step S5 is i The obtained Doppler velocity filtered point set Make further judgments based on multi-dimensional information.

[0158] Construct a multidimensional information decision maker κ and define a multidimensional information vector containing n factors or features for:

[0159]

[0160] The x i is a certain dimension of information that represents the data. At the same time, the n-dimensional weight vector corresponding to the multidimensional information vector is defined for:

[0161]

[0162] where b i is the weight value corresponding to the i-th feature.

[0163] Define the decision maker output op can be expressed as a weighted sum:

[0164]

[0165] where x i is a dimension information representing the data, b i is the weight value corresponding to the i-th feature.

[0166] Based on millimeter wave radar detection, point cloud data is obtained, including the scattering area rcs, signal-to-noise ratio snr, and signal power of the radar point cloud data point trace. A 3D information decision maker is defined, in which the feature vector for:

[0167]

[0168] Define the weight vector in the 3D information decision maker for:

[0169]

[0170] Among them, b1, b2, and b3 represent the weight factors corresponding to the components of the scattering area rcs, the signal-to-noise ratio snr, and the signal power power in the vector x, respectively.

[0171] The output op of the decision maker is defined as a weighted sum:

[0172]

[0173] Assume that for the point p in step S5 i , the obtained filtered point set The points in are:

[0174] p j ={x j ,y j ,r j ,v Dj ,v j ,θ j ,rcs j ,snr j ,power j}

[0175] In this embodiment, based on the empirical value, the weight vectors in the decision maker are set to 0.4, 0.3, and 0.3 respectively. i , p j Use the decision maker to get the output op i , op j If the following relationship is satisfied, then the point trace p in the set is considered j The target point p i Multipath false target points are removed:

[0176] op i >op j

[0177] Repeat step S4 for each point in the dynamic point cloud set D to obtain the Doppler velocity estimation value of the multipath false target, and complete the Doppler velocity filtering of the moving target point based on step S5, and perform multi-dimensional information joint judgment of step S6 on the point set that meets the conditions, so as to complete the identification and suppression of multipath false targets.

[0178] See also Figure 7 Through the measured data of millimeter wave radar, the method provided in this embodiment is based on the multipath false targets caused by static reflection boundaries, and filtering in the Doppler velocity dimension is used to realize the identification and suppression of multipath false targets, while ensuring that the real targets are not lost, verifying the correctness and effectiveness of the present invention.

[0179] Example 2

[0180] An embodiment of the present invention further provides an electronic device, the electronic device comprising a processor and a memory, the memory storing at least one instruction, at least one program, a code set or an instruction set, the at least one instruction, the at least one program, the code set or the instruction set being loaded and executed by the processor to implement the following Figure 1 A Doppler velocity filtering multipath false target suppression method is shown.

[0181] It is understood that the memory may include random access memory (RAM) or read-only memory (RAM). Optionally, the memory includes a non-transitory, computer-readable, storage, medium. The memory may be used to store instructions, programs, codes, code sets, or instruction sets. The memory may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function, instructions for implementing the above-mentioned various method embodiments, etc.; the data storage area may store data created according to the use of the server, etc.

[0182] The processor may include one or more processing cores. The processor uses various interfaces and lines to connect the various parts of the entire server, and executes various functions of the server and processes data by running or executing instructions, programs, code sets or instruction sets stored in the memory, and calling data stored in the memory. Optionally, the processor can be implemented in at least one hardware form of digital signal processing (Digital, Signal, Processing, DSP), field programmable gate array (Field-Programmable, Gate, Array, FPGA), and programmable logic array (Programmable, Logic, Array, PLA). The processor can integrate one or a combination of a central processing unit (Central, Processing, Unit, CPU) and a modem. Among them, the CPU mainly processes the operating system and application programs; the modem is used to process wireless communications. It is understandable that the above-mentioned modem may not be integrated into the processor, but implemented separately through a chip.

[0183] Since the electronic device is an electronic device corresponding to a Doppler velocity filtering multipath false target suppression method in an embodiment of the present invention, and the principle of solving the problem by the electronic device is similar to that of the method, the implementation of the electronic device can refer to the implementation process of the above method embodiment, and the repeated parts will not be repeated.

[0184] Example 3

[0185] The embodiment of the present invention further provides a computer-readable storage medium, wherein the storage medium stores at least one instruction, at least one program, code set or instruction set, and the at least one instruction, the at least one program, the code set or instruction set is loaded and executed by a processor to implement the following Figure 1 A Doppler velocity filtering multipath false target suppression method is shown.

[0186] A person of ordinary skill in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium, and the storage medium includes a read-only memory (Read-Only, Memory, ROM), a random access memory (Random, Access, Memory, RAM), a programmable read-only memory (Programmable, Read-only, Memory, PROM), an erasable programmable read-only memory (Erasable Programmable, Read, Only, Memory, EPROM), a one-time programmable read-only memory (One-time, Programmable Read-Only, Memory, OTPROM), an electronically erasable rewritable read-only memory (Electrically-Erasable Programmable, Read-Only, Memory, EEPROM), a compact disc (Compact, Disc, Read-Only, Memory, CD-ROM) or other optical disc storage, magnetic disk storage, magnetic tape storage, or any other computer-readable medium that can be used to carry or store data.

[0187] Since the storage medium is a storage medium corresponding to a Doppler velocity filtering multipath false target suppression method in an embodiment of the present invention, and the principle of solving the problem by the storage medium is similar to that of the method, the implementation of the storage medium can refer to the implementation process of the above method embodiment, and the repeated parts will not be repeated.

[0188] Example 4

[0189] In some possible implementations, various aspects of the method of the embodiment of the present invention can also be implemented in the form of a program product, which includes a program code. When the program product is run on a computer device, the program code is used to enable the computer device to execute the steps of a Doppler velocity filtering multipath false target suppression method according to various exemplary embodiments of the present application described above in this specification. Among them, the executable computer program code or "code" for executing each embodiment can be written in a high-level programming language such as C, C++, C#, Smalltalk, Java, JavaScript, Visual Basic, structured query language (e.g., Transact-SQL), Perl, or in various other programming languages.

[0190] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0191] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.

[0192] The above embodiments are only for illustrating the technical concept and features of the present invention, and their purpose is to enable ordinary technicians in the field to understand the content of the present invention and implement it accordingly, and they cannot be used to limit the protection scope of the present invention. Any equivalent changes or modifications made based on the essence of the content of the present invention should be included in the protection scope of the present invention.

Claims

1. A Doppler velocity filtering multipath false target suppression method, characterized in that: The following steps are involved: Get radar point cloud data; Separate the dynamic and static point cloud data of the radar original echo data to obtain static point cloud set and dynamic point cloud set; Cluster the static point cloud set, perform linear boundary fitting based on the clustering results, and obtain the straight line equation of the reflection boundary; Based on the multipath propagation model, the Doppler velocities of the first-order multipath false targets and the second-order multipath false targets are estimated for the points in the dynamic point cloud set, and the corresponding Doppler velocity search interval is obtained. Traversing the dynamic point cloud set, performing global point cloud Doppler velocity filtering on the obtained search interval, and obtaining a filtered dynamic point cloud set; Multi-dimensional information judgment is performed on the filtered dynamic point cloud set to identify multipath false targets in order to achieve multipath false target suppression.

2. A Doppler velocity filtering multipath false target suppression method according to claim 1, characterized in that: The expression of the radar point cloud data is: Ω={p g |x g ,y g ,r g ,v Dg ,v g ,θ g ,rcs g ,snr g ,power g }(g=1,2,……N) In the formula, p g is the g-th point in the point cloud set Ω; x g ,y g represents the position coordinates of the g-th point; r g Indicates the radial distance of the g-th point; represents the Doppler velocity of the g-th point; v g represents the speed of the g-th point; θ g Represents the azimuth of the g-th point; rcs g represents the scattering area of ​​the g-th point; snr g represents the signal-to-noise ratio of the g-th point; power g Represents the signal power of the g-th trace; N is the number of traces.

3. A Doppler velocity filtering multipath false target suppression method according to claim 1, characterized in that: The method of clustering the static point cloud set and performing linear boundary fitting according to the clustering result to obtain the straight line equation of the reflection boundary includes: The static point cloud set is clustered using a density-based clustering method, and the reflection boundary is fitted using Hough transform in each cluster to obtain the straight line equation of the reflection boundary.

4. A Doppler velocity filtering multipath false target suppression method according to claim 3, characterized in that: The method of clustering a static point cloud set using a density-based clustering method and fitting a reflection boundary using a Hough transform in each cluster includes: Perform DBSCAN processing on the static point cloud of each frame to obtain M static point cloud clusters; For all data points (x i ,y i )Use the following formula to get the Hesse normal: p=x i cosθ+y i sinth In the formula, ρ represents the distance from the origin to the point (x i ,y i ), θ represents the perpendicular distance of the vertical line; Get all the curves corresponding to the data points in the cluster to form a set C; Estimate the intersection point (θ) of the curve set C corresponding to each cluster point set j ,ρ j ); Finally, the parameters of the fitted line are estimated: Assuming that in a Cartesian coordinate system, the equation of the fitted straight line of the reflection boundary is: Ax+By+C=0 Based on the estimated intersection coordinates (θ j ,ρ j ), convert the Hesse normal parameters to general line equation parameters: For all data points (x i ,y i ), based on the fitting straight line equation, all static point cloud clusters are traversed to obtain the fitting result of each point cloud cluster, thereby realizing the estimation of the static reflection boundary.

5. A Doppler velocity filtering multipath false target suppression method according to claim 1, characterized in that: The Doppler velocity of the second-order multipath false target is estimated as follows: Assume that the target point coordinates in the dynamic point cloud set D are (x t ,y t ), the parameters A, B, C are obtained based on the fitted straight line equation, and the second-order false target coordinates (x ts ,y ts ): According to the target point (x t ,y t ) About the second-order false target point (x) fitting the reflection boundary ts ,y ts ), obtain the second-order false target point (x ts ,y ts ) is: According to the differential relationship between Doppler velocity and radial distance in radar signal processing, the second-order false target point (x ts ,y ts ) is: The Doppler velocity of the first-order multipath false target is estimated as follows: According to the multipath propagation geometric model, the real target point traces generate first-order false targets M1 and M2 respectively about the static reflection boundary; Assume that the target point coordinates in the dynamic point cloud set D are (x t ,y t ), the coordinates of the second-order false target point are (x ts ,y ts ), according to the electromagnetic wave multipath propagation model, the theoretical radial distance of the first-order false targets M1 and M2 is equal and recorded as r M : where r ts is the theoretical radial distance of the second-order false target, r t is the radial distance of the target point; At the same time, the theoretical Doppler velocity of the first-order false targets M1 and M2 is: In the formula, (x t ,y t ), (x ts ,y ts ) are the target point coordinates and the second-order false target coordinates.

6. A Doppler velocity filtering multipath false target suppression method according to claim 5, characterized in that: The obtaining of the corresponding Doppler velocity search interval includes: Assume that the dynamic target point is p i ∈D, D is a dynamic point cloud set, Calculate the target point p respectively i The position coordinates of the theoretical second-order false target S obtained by fitting the reflection boundary Radial distance Radial velocity Search interval And the radial distance of the first-order false targets M1, M2 Radial velocity Search interval Based on the search interval Get the Doppler velocity search interval δ of the target point i for:

7. A Doppler velocity filtering multipath false target suppression method according to claim 6, characterized in that: The step of performing global point cloud Doppler velocity filtering on the obtained search interval to obtain a filtered dynamic point cloud set includes: Doppler velocity filtering in the dynamic point cloud set D will satisfy the search interval δ i The point set of For the point trace p in the dynamic point cloud set D i , get the point trace set about this point Then all the points in the set satisfy the Doppler velocity constraint.

8. A Doppler velocity filtering multipath false target suppression method according to claim 1, characterized in that: Multi-dimensional information decision making is achieved by constructing a multi-dimensional information decision judger, specifically: The 3D feature vector is defined as follows: Among them, rcs is the scattering area of ​​the point cloud data, snr is the point cloud signal-to-noise ratio, and power is the signal power; Define the weight vector in the 3D information decision maker for: The weighted sum of the output op of the decision maker is calculated as: The decision maker is used to calculate the output of the decision maker in the point cloud traces of the dynamic target and the point cloud set traces after Doppler velocity filtering, thereby realizing the identification of multipath false targets.

9. An electronic device, characterized in that: The electronic device includes a processor and a memory, wherein the memory stores at least one instruction, at least one program, a code set or an instruction set, and the at least one instruction, the at least one program, the code set or the instruction set is loaded and executed by the processor to implement the method described in any one of claims 1 to 8.

10. A computer-readable storage medium, characterized in that: The storage medium stores at least one instruction, at least one program, a code set or an instruction set, and the at least one instruction, the at least one program, the code set or the instruction set is loaded and executed by the processor to implement the method according to any one of claims 1 to 8.

Images