A point cloud data clustering method, device, equipment and medium

Abstract

The embodiment of the application provides a point cloud data clustering method, device, equipment and medium, in the method, based on the point cloud data collected by a radar, a first distance and a first angle between the radar and a target object are determined, for each preset direction, a first clustering range of the target object in the preset direction is determined according to the first distance, the first angle, an angle resolution of the radar and a target preset clustering range determination function corresponding to the preset direction, and according to each first clustering range, a first target clustering range of the target object is determined, and the point cloud data corresponding to each point in the first target clustering range is determined as the point cloud data corresponding to the target object. The purpose that the clustering range dynamically changes with the distance is achieved when the point cloud data is clustered.

Description

Technical Field

[0001] This application relates to the field of trusted artificial intelligence technology, and in particular to a point cloud data clustering method, apparatus, device and medium. Background Technology

[0002] Radar, as a common sensor, such as millimeter-wave radar, has advantages such as strong anti-interference capability, high detection accuracy, and protection of user privacy, and therefore has been widely used in many fields, such as human vital sign monitoring, human motion trajectory tracking, and intelligent driving. However, radar still has some problems. For example, when tracking human motion trajectories, the radar output is point cloud data. The density of this point cloud data varies with distance, producing a denser near-field and sparser far-field effect. Data of different densities can lead to significantly different clustering results. Therefore, how to cluster data of different densities is an important research direction in millimeter-wave radar data processing.

[0003] Currently, the most common point cloud data clustering method is to use the Density-Based Spatial Clustering of Applications with Noise algorithm to cluster point cloud data and extract features from the clustering target. This method can handle noisy data well, but its disadvantage is that the clustering effect is not ideal for point cloud data with different density distributions.

[0004] Therefore, how to achieve dynamic changes in clustering range with distance has become an urgent problem to be solved. Summary of the Invention

[0005] This application provides a point cloud data clustering method, apparatus, device, and medium to solve the problem of unsatisfactory clustering results for data with different density distributions in the prior art.

[0006] Firstly, this application provides a point cloud data clustering method, the method comprising:

[0007] Based on the point cloud data collected by the radar, a first distance and a first angle between the radar and the target are determined;

[0008] For each preset direction, the first clustering range of the target object in that preset direction is determined based on the first distance, the first angle, the angular resolution of the radar, and the target preset clustering range determination function corresponding to that preset direction;

[0009] Based on each of the first cluster ranges, determine each point located within the first target cluster range, and determine the point cloud data corresponding to each point within the first target cluster range as the point cloud data corresponding to the target object.

[0010] In one possible implementation, determining the first clustering range of the target object in the preset direction based on the first distance, the first angle, the angular resolution of the radar, and the target preset clustering range determination function corresponding to the preset direction includes: determining the first clustering range of the target object in the preset direction based on the first distance, the first angle, the angular resolution of the radar, a preset value of a second parameter corresponding to the preset direction, and the target preset clustering range determination function; or determining the value range of the second parameter, arbitrarily selecting a value from the value range as the first value of the second parameter, and determining the first clustering range of the target object in the preset direction based on the first value of the second parameter, the first distance, the first angle, the angular resolution of the radar, the target preset clustering range determination function, and the preset direction.

[0011] In one possible implementation, determining the first clustering range of the target object in each preset direction based on the first distance, the first angle, the angular resolution of the radar, and the target preset clustering range determination function corresponding to the preset direction includes: randomly selecting a preset number of first numerical groups, wherein the first numerical groups include a second value of a second parameter arbitrarily selected from the value range of the second parameter corresponding to each preset direction; processing the preset number of first numerical groups based on a genetic algorithm to obtain a preset number of second numerical groups; selecting one second numerical group as the optimal numerical group from the preset number of second numerical groups; and determining the first clustering range of the target object in each preset direction based on the third value of the second parameter corresponding to the preset direction in the optimal numerical group, the first distance, the first angle, the angular resolution of the radar, and the target preset clustering range determination function for the preset direction.

[0012] In one possible implementation, the step of selecting a second value group as the optimal value group from the preset number of second value groups involves: determining the target optimization value corresponding to each of the preset number of second value groups based on the preset number of second value groups and the preset target optimization algorithm; determining the optimal target optimization value based on the preset number of target optimization values ​​and the simulated annealing algorithm; and taking the second value group corresponding to the optimal target optimization value as the optimal value group.

[0013] In one possible implementation, determining the target optimization value corresponding to each of the preset number of second numerical groups based on the preset number of second numerical groups and the preset target optimization algorithm includes: for each second numerical group, determining the second clustering range of the target object in each preset direction based on the parameter value of the second parameter corresponding to each preset direction included in the second numerical group, the first distance, the first angle, the angular resolution of the radar, and the target preset clustering range determination function for each preset direction; determining each point of the target object located within the second target clustering range based on each second clustering range; determining the weight coefficient corresponding to each point within the second target clustering range based on the first position coordinates of the cluster center of the second target clustering range, the second position coordinates of each point within the second target clustering range, the first distance, and the median value of the weight coefficient; determining the point cloud data density and data similarity corresponding to the second numerical group based on the total number of points contained in the second target clustering range, the weight coefficient, the preset point cloud data density determination function, and the preset data similarity determination function; and determining the sum of the point cloud data density and the data similarity as the target optimization value corresponding to the second numerical group.

[0014] Secondly, this application provides a point cloud data clustering device, the device comprising:

[0015] The first determining module is used to determine the first distance and the first angle between the radar and the target based on the point cloud data collected by the radar.

[0016] The second determining module is used to determine the first clustering range of the target object in the preset direction for each preset direction based on the first distance, the first angle, the angular resolution of the radar, and the target preset clustering range determination function corresponding to the preset direction.

[0017] The third determining module is used to determine each point within the first target cluster range based on each of the first cluster ranges, and to determine the point cloud data corresponding to each point within the first target cluster range as the point cloud data corresponding to the target object.

[0018] In one possible implementation, the second determining module is further configured to determine the first clustering range of the target object in the preset direction based on the first distance, the first angle, the angular resolution of the radar, the preset value of the second parameter corresponding to the preset direction, and the target preset clustering range determining function; or to determine the value range of the second parameter, arbitrarily select a value in the value range as the first value of the second parameter, and determine the first clustering range of the target object in the preset direction based on the first value of the second parameter, the first distance, the first angle, the angular resolution of the radar, the target preset clustering range determining function in the preset direction.

[0019] In one possible implementation, the device further includes:

[0020] The selection module is used to randomly select a preset number of first numerical groups, wherein the first numerical group includes a second value of a second parameter arbitrarily selected from the value range of the second parameter corresponding to each preset direction; the preset number of first numerical groups are processed based on a genetic algorithm to obtain a preset number of second numerical groups; and a second numerical group is selected from the preset number of second numerical groups as the optimal numerical group.

[0021] The second determining module is further configured to, for each preset direction, determine the first clustering range of the target object in the preset direction based on the third value of the second parameter corresponding to the preset direction in the optimal value group, the first distance, the first angle, the angular resolution of the radar, and the target preset clustering range determination function in the preset direction.

[0022] In one possible implementation, the selection module is specifically used to determine the target optimization values ​​corresponding to the preset number of second value groups according to the preset number of second value groups and the preset target optimization algorithm; determine the optimal target optimization value according to the preset number of target optimization values ​​and the simulated annealing algorithm; and take the second value group corresponding to the optimal target optimization value as the optimal value group.

[0023] In one possible implementation, the selection module is specifically configured to, for each second numerical group, determine the second clustering range of the target object in each preset direction based on the parameter values ​​of the second parameters corresponding to each preset direction included in the second numerical group, the first distance, the first angle, the angular resolution of the radar, and the target preset clustering range determination function for each preset direction; determine each point of the target object located within the second target clustering range based on each second clustering range; determine the weight coefficient corresponding to each point within the second target clustering range based on the first position coordinates of the cluster center of the second target clustering range, the second position coordinates of each point within the second target clustering range, the first distance, and the median value of the weight coefficient; determine the point cloud data density and data similarity corresponding to the second numerical group based on the total number of points contained in the second target clustering range, the weight coefficients, the preset point cloud data density determination function, and the preset data similarity determination function; and determine the sum of the point cloud data density and the data similarity as the target optimization value corresponding to the second numerical group.

[0024] Thirdly, this application also provides an electronic device, which includes at least a processor and a memory, wherein the processor is configured to execute a computer program stored in the memory to implement the steps of any of the point cloud data clustering methods described above.

[0025] Fourthly, this application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of any of the point cloud data clustering methods described above.

[0026] This application provides a point cloud data clustering method, apparatus, device, and medium. The method, based on point cloud data acquired by radar, determines a first distance and a first angle between the radar and a target object. For each preset direction, a first clustering range of the target object is determined according to the first distance, the first angle, the radar's angular resolution, and a target preset clustering range determination function corresponding to that preset direction. Furthermore, based on each first clustering range, a first target clustering range of the target object is determined, and the point cloud data corresponding to each point within the first target clustering range is identified as the point cloud data corresponding to the target object. Because in this application embodiment, the first distance and the first angle between the radar and the target object are determined based on the point cloud data acquired by radar, and the determined first distance, first angle, and radar's angular resolution are input into the target preset clustering range determination function corresponding to each preset direction to determine the first clustering range of the target object in each preset direction, the first target clustering range of the target object can be determined based on each determined first clustering range. This achieves the goal of dynamically changing the clustering range with distance during point cloud data clustering. Attached Figure Description

[0027] To more clearly illustrate the technical solutions of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0028] Figure 1 This is one of the schematic diagrams of a point cloud data clustering process provided in some embodiments of this application;

[0029] Figure 2 A second schematic diagram illustrating a point cloud data clustering process provided for some embodiments of this application;

[0030] Figure 3 A third schematic diagram illustrating a point cloud data clustering process provided for some embodiments of this application;

[0031] Figure 4 A schematic diagram illustrating a process for determining an optimal set of values, provided for some embodiments of this application;

[0032] Figure 5 A fourth schematic diagram illustrating a point cloud data clustering process provided for some embodiments of this application;

[0033] Figure 6 A schematic diagram illustrating a process for determining a target optimization value, provided for some embodiments of this application;

[0034] Figure 7 Fifth of some embodiments of this application provides a schematic diagram of a point cloud data clustering process;

[0035] Figure 8a One of the structural schematic diagrams of a point cloud data clustering device provided in some embodiments of this application;

[0036] Figure 8b A second schematic diagram of a point cloud data clustering device provided in some embodiments of this application;

[0037] Figure 9 This is a schematic diagram of an electronic device structure provided for some embodiments of this application. Detailed Implementation

[0038] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art are within the scope of protection of this application.

[0039] This application provides a point cloud data clustering method, apparatus, device, and medium. The method, based on point cloud data acquired by radar, determines a first distance and a first angle between the radar and a target object. For each preset direction, it determines a first clustering range of the target object in that preset direction based on the first distance, the first angle, the radar's angular resolution, and a target preset clustering range determination function corresponding to that preset direction. Furthermore, based on each first clustering range, it determines a first target clustering range of the target object, and identifies the point cloud data corresponding to each point within the first target clustering range as the point cloud data corresponding to the target object.

[0040] Figure 1 A schematic diagram of a point cloud data clustering process is provided for some embodiments of this application. The process includes the following steps:

[0041] S101: Based on the point cloud data collected by the radar, determine the first distance and the first angle between the radar and the target.

[0042] The point cloud data clustering method provided in some embodiments of this application is applied to electronic devices, such as smart terminals, PCs, or servers.

[0043] In order to achieve dynamic changes in the clustering range with distance, thereby improving the effect of point cloud data clustering, in this embodiment of the application, a first distance and a first angle between the radar and the target are determined based on the point cloud data collected by the radar.

[0044] Taking millimeter-wave radar as an example, the transmitted signal of a millimeter-wave radar is a continuous frequency modulated wave. After receiving the echo signal, the millimeter-wave radar processes it through a mixer to obtain an intermediate frequency (IF) signal. In this embodiment, the IF signal output by the millimeter-wave radar can be acquired, and the acquired IF signal can be preprocessed to obtain the coordinate representation of the point cloud data of the points collected by the millimeter-wave radar. When preprocessing the IF signal, the acquired IF signal can be processed based on the fast Fourier transform (FFT) to obtain the distance r, velocity v, and angle corresponding to each point collected by the millimeter-wave radar, where the angle includes the horizontal angle θ and the elevation angle. Specifically, firstly, a range-dimensional FFT can be performed on the intermediate frequency signal to obtain the distance *r* between a point on the target and the radar. Then, based on this FFT, velocity-dimensional FFT and angle-dimensional FFT can be performed to obtain the velocity *v* and angle of the point on the target relative to the radar, where the angle includes the horizontal angle θ and the elevation angle.

[0045] The distance r, velocity v, horizontal angle θ, and elevation angle corresponding to each point acquired by the millimeter-wave radar were determined. Then, for each point acquired by the millimeter-wave radar, the velocity and distance from the origin in each preset direction can be determined based on trigonometric functions. The preset directions can be the X-axis, Y-axis, and Z-axis directions of a predefined coordinate system. Specifically, the distances of the point from the origin in the X-axis, Y-axis, and Z-axis directions can be expressed as:

[0046]

[0047]

[0048]

[0049] Where, r x r is the distance from the origin in the X-axis direction. y r is the distance from the origin in the Y-axis direction. z Let r be the distance from the origin along the Z-axis, and r represent the straight-line distance between the point and the radar. In other words, performing a range-dimension FFT on the intermediate frequency signal yields the distance r between the point on the target and the radar. θ is the elevation angle corresponding to the point obtained by the millimeter-wave radar based on FFT, and θ is the horizontal angle corresponding to the point obtained by the millimeter-wave radar based on FFT.

[0050] The velocities of this point in the X-axis, Y-axis, and Z-axis directions can be expressed as:

[0051]

[0052]

[0053]

[0054] Among them, v x Let v be the velocity component in the X-axis direction. y Let v be the velocity component in the Y-axis direction. z Let v be the velocity component along the Z-axis, representing the velocity of that point relative to the radar. θ is the elevation angle corresponding to the point obtained by the millimeter-wave radar based on FFT, and θ is the horizontal angle corresponding to the point obtained by the millimeter-wave radar based on FFT.

[0055] The velocity v of the point in each preset direction was determined. x v y v z and the distance r from the origin x r y rz Then, the point cloud data of that point can be represented as P(r x ,r y ,r z ,v x ,v y ,v z ).

[0056] Once the point cloud data for each point is determined, the first distance and the first angle between the radar and the target can also be determined. The first distance can be the average distance determined based on the point cloud data of each point on the target. The first angle can include the first horizontal angle and the first pitch angle. The first horizontal angle and the first pitch angle can be determined based on the horizontal angle and the pitch angle corresponding to each point on the target.

[0057] S102: For each preset direction, determine the first clustering range of the target object in that preset direction based on the first distance, the first angle, the angular resolution of the radar, and the target preset clustering range determination function corresponding to that preset direction.

[0058] Since the density of 3D point cloud data varies with distance, distance can be introduced as an influencing parameter when establishing the clustering range to improve the point cloud data clustering effect. In this embodiment, for each preset direction, which can be the X-axis, Y-axis, and Z-axis, a first clustering range of the target object in that preset direction can be determined based on a determined first distance, a first angle, the radar's angular resolution, and a target preset clustering range determination function corresponding to that preset direction. Because the coordinates of the point cloud data are determined by a coordinate system, the clustering range in each preset direction of that coordinate system is determined.

[0059] In this embodiment, the radar's angular resolution includes horizontal angular resolution and elevation angular resolution, which can be determined based on the radar's initial parameters. Specifically, the horizontal angular resolution can be expressed by the following formula:

[0060]

[0061] Where Δθ is the horizontal angular resolution of the radar, θ FOV N represents the radar's horizontal field of view. θ This represents the number of antennas in the horizontal direction of the radar.

[0062] Pitch angle resolution can be expressed using the following formula:

[0063]

[0064] For the radar's elevation angle resolution, The radar's visible range in the vertical direction. This represents the number of antennas in the vertical direction of the radar.

[0065] In this embodiment, a target preset clustering range determination function is stored for each preset direction. For example, the target preset clustering range determination function stored in advance for the X-axis direction can be expressed as:

[0066]

[0067] Where a is the first clustering range along the X-axis, f x Here, r is the preset second parameter corresponding to the X-axis direction, and r is the first distance. Let Δθ be the radar's elevation angle resolution, Δθ be the radar's horizontal angle resolution, and θ be the first horizontal angle included in the first angle.

[0068] The function for determining the pre-saved target clustering range in the Y-axis direction can be expressed as:

[0069] b = f y ·r·cos(Δθ)·sin(Δθ)·cos(θ)

[0070] Where b is the first clustering range along the Y-axis, f y The preset second parameter is the one corresponding to the Y-axis direction, r is the first distance, Δθ is the horizontal angular resolution of the radar, and θ is the first horizontal angle included in the first angle.

[0071] The function for determining the pre-saved target clustering range along the Z-axis can be expressed as:

[0072]

[0073] Where c is the first clustering range along the Z-axis, f z The second preset parameter is the one corresponding to the Z-axis direction, and r is the first distance. This represents the radar's elevation angle resolution.

[0074] S103: Based on each of the first clustering ranges, determine each point located within the first target clustering range, and determine the point cloud data corresponding to each point within the first target clustering range as the point cloud data corresponding to the target object.

[0075] After determining the first clustering range in each preset direction, the first target clustering range of the target object can be determined based on each first clustering range.

[0076] Specifically, in this embodiment, the average value in each preset direction can be calculated based on the coordinates of each point cloud data. The determined average value in each preset direction is then used as the cluster center in that preset direction. For ease of description, the cluster center in the X-axis direction can be represented as... Cluster centers along the Y-axis can be represented as The cluster centers along the Z-axis can be represented as After determining the cluster centers along the X, Y, and Z axes, each point within the first target cluster range can be determined based on the first cluster range and cluster centers along the X, Y, and Z axes, according to the following formula. The coordinates of each point within the first target cluster range satisfy the following formula, and each point within the first target cluster range determines an ellipsoidal first target cluster range:

[0077]

[0078] Where a represents the first cluster range along the X-axis, b represents the first cluster range along the Y-axis, and c represents the first cluster range along the Z-axis. The cluster centers are located along the X-axis. The cluster centers are located along the Y-axis. Let x, y, and z be the cluster centers along the Z-axis, and let x, y, and z be the coordinates of points within the first target cluster.

[0079] After identifying each point within the first target cluster range, the point cloud data corresponding to each point within the first target cluster range can be determined as the point cloud data corresponding to the target object. Of course, to further improve the point cloud data clustering effect, after identifying each point within the first target cluster range, further data clustering can be performed on the point cloud data corresponding to each point within the first target cluster range, which is not limited in this embodiment.

[0080] In this embodiment, based on the point cloud data collected by the radar, the first distance and the first angle between the radar and the target are determined, and the determined first distance, the first angle, and the angular resolution of the radar are input into the target preset clustering range determination function corresponding to each preset direction to determine the first clustering range of the target in each preset direction. Thus, based on each determined first clustering range, each point of the target located within the first target clustering range can be determined, thereby achieving the purpose of dynamically changing the clustering range with distance when clustering point cloud data.

[0081] To further improve the effect of point cloud data clustering, based on the above embodiments, in this embodiment, determining the first clustering range of the target object in the preset direction according to the first distance, the first angle, the angular resolution of the radar, and the target preset clustering range determination function corresponding to the preset direction includes:

[0082] Based on the first distance, the first angle, the angular resolution of the radar, the preset value of the second parameter corresponding to the preset direction, and the target preset clustering range determination function, the first clustering range of the target object in the preset direction is determined; or

[0083] The range of values ​​for the second parameter is determined, and any value within the range is taken as the first value of the second parameter. Based on the first value of the second parameter, the first distance, the first angle, the angular resolution of the radar, and the target preset clustering range determination function in the preset direction, the first clustering range of the target is determined.

[0084] To further improve the point cloud data clustering effect, in this embodiment, a preset value for a second parameter corresponding to each preset direction can be pre-configured. This second parameter is the preset second parameter in the target preset clustering range determination function corresponding to each preset direction. Specifically, the second preset parameter f corresponding to the X-axis direction... x The preset value can be 1, and the corresponding second preset parameter f in the Y-axis direction y The preset value can be 1, and the corresponding second preset parameter f in the Z-axis direction is 1. z The default value can be 1.

[0085] After determining the first distance and the first angle between the radar and the target, for each preset direction, the first cluster range of the target can be determined based on the first distance, the first angle, the radar's angular resolution, the preset value of the second parameter corresponding to the preset direction, and the target preset cluster range determination function corresponding to the preset direction.

[0086] In one possible implementation, a range of values ​​for the second parameter can be determined for each preset direction, and any value within that range can be taken as the first value of the second parameter for that preset direction.

[0087] Specifically, in this embodiment, a reference value can be configured for each preset direction. This reference value can be the maximum value of the target object in that direction. For example, assuming the target object is a person and the maximum height of a person is 2 meters, then the maximum value in the Z-axis direction is 2 meters. When determining the range of values ​​for the second parameter, the range of values ​​for the second parameter in the preset direction can be determined based on the reference value corresponding to the preset direction and the size of the search unit set for that direction.

[0088] Specifically, the value range for each preset direction can be determined based on the reference value corresponding to each preset direction and the preset second parameter. When determining the value range of the second parameter corresponding to the X-axis direction, the following formula can be used:

[0089]

[0090] Among them, X max Here, f is the reference value along the X-axis, and a′ is the search unit corresponding to the X-axis. For example, if the reference value along the X-axis is 2m and the corresponding search unit is 0.5m, then the range of the second parameter along the X-axis is f. x ∈[1, 40].

[0091] When determining the range of values ​​for the second parameter corresponding to the Y-axis direction, the following formula can be used:

[0092]

[0093] Among them, Y max b is the reference value in the Y-axis direction, and b′ is the search unit corresponding to the Y-axis direction.

[0094] When determining the range of values ​​for the second parameter corresponding to the Z-axis direction, the following formula can be used:

[0095]

[0096] Among them, Z max c is the reference value in the Z-axis direction, and c′ is the search unit corresponding to the Z-axis direction.

[0097] The size of the search unit corresponding to each preset direction can be set separately, and the search units corresponding to each preset direction can be the same or different.

[0098] After determining the range of values ​​for the second parameter in each preset direction, any value can be selected as the first value of the second parameter within the range of values ​​for each preset direction. Based on the first value of the second parameter, the first distance, the first angle, the angular resolution of the radar, and the target preset clustering range determination function in the preset direction, the first clustering range of the target object in that direction is determined.

[0099] Assuming that the formula determines the range of values ​​for the second parameter corresponding to the X-axis direction, the determined range of values ​​for the second parameter corresponding to the X-axis direction is f. x ∈[1,4]. Any value within this range is chosen as the first value of the second parameter, which is 2. This first value of 2 is then used to determine the target's preset clustering range along the X-axis. The preset second parameter f x The function is updated to determine the target preset clustering range. Based on the first distance, the first angle, the radar's angular resolution, and the updated target preset clustering range determination function, the first clustering range of the target is determined in the X-axis direction.

[0100] Figure 2 This is a second schematic diagram illustrating a point cloud data clustering process provided for some embodiments of this application. For example... Figure 2 As shown, the process includes the following steps:

[0101] S201: Based on the point cloud data collected by the radar, determine the first distance and the first angle between the radar and the target.

[0102] S202a: Determine the first clustering range of the target object in the preset direction based on the first distance, the first angle, the angular resolution of the radar, the preset value of the second parameter corresponding to the preset direction, and the target preset clustering range determination function.

[0103] S202b: Determine the value range of the second parameter, arbitrarily select a value from the value range as the first value of the second parameter, and determine the first clustering range of the target in the preset direction based on the first value of the second parameter, the first distance, the first angle, the angular resolution of the radar, and the target preset clustering range determination function in the preset direction.

[0104] S203: Based on each of the first clustering ranges, determine each point located within the first target clustering range, and determine the point cloud data corresponding to each point within the first target clustering range as the point cloud data corresponding to the target object.

[0105] To further improve the effect of point cloud data clustering, based on the above embodiments, in this embodiment, determining the first clustering range of the target object in the preset direction for each preset direction, according to the first distance, the first angle, the angular resolution of the radar, and the target preset clustering range determination function corresponding to the preset direction, includes:

[0106] A first set of values ​​of a predetermined number is randomly selected. The first set of values ​​includes a second value of a second parameter that is arbitrarily selected from the range of values ​​of the second parameter corresponding to each predetermined direction.

[0107] The first set of values ​​of a preset number is processed using a genetic algorithm to obtain a second set of values ​​of a preset number.

[0108] Select one of the preset number of second value groups as the optimal value group;

[0109] For each preset direction, the first clustering range of the target object in that preset direction is determined based on the third value of the second parameter corresponding to that preset direction in the optimal value group, the first distance, the first angle, the angular resolution of the radar, and the target preset clustering range determination function for that preset direction.

[0110] To further improve the clustering effect of point cloud data, in this embodiment, a preset number of first value groups can be randomly selected. Each first value group includes a second value of a second parameter randomly selected from the value range of the second parameter corresponding to each preset direction. In other words, the first value group includes the second value of any second parameter selected from the value range corresponding to each preset direction. Since the preset directions in this embodiment include the X-axis, Y-axis, and Z-axis directions, each first value group contains three second values, and each second value corresponds to one preset direction.

[0111] In this embodiment of the application, a preset number of first numerical groups can be processed using a genetic algorithm to obtain a preset number of second numerical groups. Specifically, the genetic algorithm can perform selection, crossover, and mutation operations on the preset number of first numerical groups to generate the preset number of second numerical groups.

[0112] Specifically, the specific steps of selection, crossover, and mutation operations in genetic algorithms can be performed using existing technologies, and will not be elaborated here.

[0113] After obtaining a preset number of second numerical sets, one of these preset number of second numerical sets can be arbitrarily selected as the optimal numerical set. Then, for each preset direction, based on the third value of the second parameter corresponding to that preset direction in the optimal numerical set, the first distance, the first angle, the radar's angular resolution, and the target preset clustering range determination function for that preset direction, the first clustering range of the target is determined in that preset direction.

[0114] Figure 3 This is the third schematic diagram illustrating a point cloud data clustering process provided in some embodiments of this application. Figure 3 As shown, the process includes the following steps:

[0115] S301: Based on the point cloud data collected by the radar, determine the first distance and the first angle between the radar and the target.

[0116] S302: Randomly select a preset number of first value groups, wherein the first value group includes a second value of a second parameter arbitrarily selected from the value range of the second parameter corresponding to each preset direction.

[0117] S303: Process the preset number of first numerical groups based on a genetic algorithm to obtain a preset number of second numerical groups; select one of the preset number of second numerical groups as the optimal numerical group.

[0118] S304: For each preset direction, determine the first clustering range of the target object in the preset direction based on the third value of the second parameter corresponding to the preset direction in the optimal value group, the first distance, the first angle, the angular resolution of the radar, and the target preset clustering range determination function in the preset direction.

[0119] S305: Based on each of the first clustering ranges, determine each point located within the first target clustering range, and determine the point cloud data corresponding to each point within the first target clustering range as the point cloud data corresponding to the target object.

[0120] In this embodiment, a genetic algorithm is used to process the first numerical group to obtain a second numerical group, and a second numerical group is selected from the second numerical group as the optimal numerical group, thereby determining the first clustering range of the target object in the preset direction, which improves the effect of point cloud data clustering.

[0121] To further improve the point cloud data clustering effect, based on the above embodiments, in this embodiment, a second value group is selected as the optimal value group from the preset number of second value groups:

[0122] Based on the preset number of second numerical groups and the preset target optimization algorithm, determine the target optimization values ​​corresponding to the preset number of second numerical groups respectively;

[0123] Based on the preset number of target optimization values ​​and the simulated annealing algorithm, the optimal target optimization value is determined, and the second value group corresponding to the optimal target optimization value is taken as the optimal value group.

[0124] To further improve the clustering effect of point cloud data, in this embodiment, a preset number of target optimization values ​​corresponding to each of the second numerical groups can be determined according to a preset target optimization algorithm. Specifically, the preset target optimization algorithm can be the Metropolis criterion. Based on the preset number of second numerical groups and the Metropolis criterion, the values ​​in each second numerical group that conform to the Metropolis criterion are obtained, and the values ​​that conform to the Metropolis criterion are taken as the target optimization values ​​corresponding to the second numerical group.

[0125] After obtaining the target optimization values ​​corresponding to the second set of values, the target optimization values ​​of the preset number and the simulated annealing algorithm are used to process the preset number of target optimization values ​​to obtain the optimal target optimization values. The second set of values ​​corresponding to the optimal target optimization values ​​is then taken as the optimal set of values.

[0126] The specific steps of the Metropolis criterion and simulated annealing algorithm can be performed using existing technologies and will not be elaborated here.

[0127] Figure 4 This is a schematic diagram illustrating a process for determining an optimal set of values, provided for some embodiments of this application. For example... Figure 4 As shown, the process includes the following steps:

[0128] S401: Based on the preset number of second numerical groups and the preset target optimization algorithm, determine the target optimization values ​​corresponding to the preset number of second numerical groups respectively.

[0129] S402: Based on the preset number of target optimization values ​​and the simulated annealing algorithm, determine the optimal target optimization value, and take the second value group corresponding to the optimal target optimization value as the optimal value group.

[0130] In this embodiment of the application, selecting the optimal value group for a preset number of second value groups can further improve the effect of point cloud data clustering.

[0131] Figure 5 This is the fourth schematic diagram illustrating a point cloud data clustering process provided for some embodiments of this application. Figure 5 As shown, the process includes the following steps:

[0132] S501: Based on the point cloud data collected by the radar, determine the first distance and the first angle between the radar and the target.

[0133] S502: Randomly select a preset number of first value groups, the first value group including a second value of a second parameter arbitrarily selected from the value range of the second parameter corresponding to each preset direction.

[0134] S503: Process the preset number of first numerical groups based on a genetic algorithm to obtain a preset number of second numerical groups; determine the target optimization values ​​corresponding to the preset number of second numerical groups according to the preset number of second numerical groups and a preset target optimization algorithm; determine the optimal target optimization value according to the preset number of target optimization values ​​and a simulated annealing algorithm, and take the second numerical group corresponding to the optimal target optimization value as the optimal numerical group.

[0135] S504: For each preset direction, the first clustering range of the target object in the preset direction is determined according to the third value of the second parameter corresponding to the preset direction in the optimal value group, the first distance, the first angle, the angular resolution of the radar, and the target preset clustering range determination function in the preset direction.

[0136] S505: Based on each of the first clustering ranges, determine each point located within the first target clustering range, and determine the point cloud data corresponding to each point within the first target clustering range as the point cloud data corresponding to the target object.

[0137] To further improve the point cloud data clustering effect, based on the above embodiments, in this embodiment, the target optimization values ​​corresponding to the preset number of second numerical groups are determined according to the preset number of second numerical groups and the preset target optimization algorithm, including:

[0138] For each second numerical group, based on the parameter values ​​of the second parameters corresponding to each preset direction included in the second numerical group, the first distance, the first angle, the angular resolution of the radar, and the target preset clustering range determination function for each preset direction, the second clustering range of the target object in each preset direction is determined; based on each second clustering range, each point of the target object located within the second target clustering range is determined; based on the first position coordinates of the cluster center of the second target clustering range, the second position coordinates of each point within the second target clustering range, the first distance, and the median value of the weight coefficients, the weight coefficients corresponding to each point within the second target clustering range are determined; based on the total number of points contained within the second target clustering range, the weight coefficients, the preset point cloud data density determination function, and the preset data similarity determination function, the point cloud data density and data similarity corresponding to the second numerical group are determined; the sum of the point cloud data density and the data similarity is determined as the target optimization value corresponding to the second numerical group.

[0139] To further improve the effect of point cloud data clustering, in this embodiment of the application, for the obtained second numerical group of a preset number, the parameter value of the second parameter corresponding to each preset direction included in the second numerical group can be obtained, specifically including: the parameter value of the second parameter f_x corresponding to the X-axis direction, the parameter value of the second parameter f_y corresponding to the Y-axis direction, and the parameter value of the second parameter f_z corresponding to the Z-axis direction.

[0140] After determining the first distance and the first angle between the radar and the target, for each preset direction, the second clustering range of the target can be determined based on the first distance, the first angle, the radar's angular resolution, the parameter values ​​of the second numerical group corresponding to the preset direction, and the target preset clustering range determination function corresponding to the preset direction.

[0141] After determining the second cluster range in each preset direction, the target object can be located at each point in the second target cluster range based on the points and cluster centers within the second cluster range in each preset direction.

[0142] Specifically, in this embodiment, the average value in each preset direction can be calculated based on the coordinates of each point cloud data. The determined average value in each preset direction is then used as the cluster center in the corresponding preset direction. After determining the cluster centers in the X, Y, and Z axis directions, each point within the second target cluster range can be determined based on the second cluster range and cluster centers in the X, Y, and Z axis directions, according to the following formula. The coordinates of each point within the second target cluster range satisfy the following formula, and each point within the second target cluster range determines an ellipsoidal second target cluster range:

[0143]

[0144] Where a represents the second clustering range along the X-axis, b represents the second clustering range along the Y-axis, and c represents the second clustering range along the Z-axis. Cluster centers along the X-axis. The cluster centers are located along the Y-axis. Let x, y, and z be the cluster centers along the Z-axis, and let x, y, and z be the coordinates of points within the second target cluster range.

[0145] Then, based on the first position coordinates of the cluster center within the second target clustering range, the second position coordinates of each point within the second target clustering range, the first distance, and the median value of the weight coefficients, the weight coefficients corresponding to each point within the second target clustering range are determined.

[0146] Specifically, in this embodiment, the median weight coefficient of each point within the second target cluster range can be calculated based on the first position coordinates of the cluster center within the second target cluster range, the second position coordinates of each point within the second target cluster range, and the first distance. For ease of description, x can be used to represent the first position coordinates of the cluster center within the second target cluster range. n This represents the second position coordinate of the nth point within the second target cluster range, where r is the first distance. The median value of the weight coefficient of the nth point within the second target cluster range is expressed as:

[0147]

[0148] Where x represents the first position coordinate of the cluster center within the second target clustering range, x n Let r represent the second position coordinate of the nth point within the second target cluster range, r be the first distance corresponding to the nth point, and h represent the median weight coefficient of the nth point within the second target cluster range. Both the first and second position coordinates correspond to three-dimensional values, and the median weight coefficient corresponding to that point is determined by the sum of the squared differences between the values ​​in the corresponding dimensions.

[0149] After obtaining the median weight coefficient of each point within the second target cluster range, the second position coordinate x of the nth point within the second target cluster range can be determined based on the total number N of points within the second target cluster range. n The median value of the weight coefficient for each point is used to determine the weight coefficient corresponding to each point within the second target clustering range.

[0150] Specifically, you can use w n The weight coefficient corresponding to the nth point within the second target cluster range is expressed as:

[0151]

[0152] Where, x n Let r represent the second position coordinate of the nth point within the second target cluster range, r be the first distance corresponding to the nth point, h represent the median value of the weight coefficient of the nth point within the second target cluster range, and w represent the second position coordinate of the nth point within the second target cluster range. n This represents the weight coefficient corresponding to the nth point within the second target cluster range, where N represents the total number of points included within the second target cluster range.

[0153] Then, based on the total number N of points included within the second target clustering range and the weight coefficient w n And the following preset point cloud data density determination functions are used to determine the point cloud data density ρ corresponding to the second set of values:

[0154]

[0155] Among them, w n Let represent the weight coefficient corresponding to the nth point within the second target cluster range, N represent the total number of points included within the second target cluster range, and x represent the coordinates of the first position of the cluster center within the second target cluster range. n This represents the second position coordinate of the nth point within the second target cluster range.

[0156] Based on the total number N of points contained in the second target cluster and the weight coefficient corresponding to each point, as well as the preset data similarity determination function, the data similarity corresponding to the second numerical group is determined. The data similarity represents the similarity between all points in the second target cluster and the cluster center corresponding to the second numerical group. The velocity information of the point cloud data is taken into account, while the influence of the human torso and noise points is eliminated.

[0157] The preset data similarity determination function is expressed as follows:

[0158]

[0159] Among them, w n Let N represent the weight coefficient corresponding to the nth point within the second target cluster range, N represent the total number of points included within the second target cluster range, and V represent the first velocity corresponding to the cluster center within the second target cluster range. n sim(o) represents the second velocity corresponding to the nth point within the second target cluster range. n ,o) represents the data similarity corresponding to the second numerical group.

[0160] The sum of the point cloud data density and data similarity corresponding to the second numerical group is determined as the target optimization value corresponding to the second numerical group.

[0161] Figure 6 This is a schematic diagram illustrating a process for determining a target optimization value, provided for some embodiments of this application. For example... Figure 6 As shown, the process includes the following steps:

[0162] S601: For each second numerical group, based on the parameter value of the second parameter corresponding to each preset direction included in the second numerical group, the first distance, the first angle, the angular resolution of the radar, and the target preset clustering range determination function for each preset direction, determine the second clustering range of the target object in each preset direction.

[0163] S602: Based on each of the second clustering ranges, determine each point within the second target clustering range where the target object is located.

[0164] S603: Based on the first position coordinates of the cluster center of the second target cluster range, the second position coordinates of each point within the second target cluster range, the first distance, and the median value of the weight coefficient, determine the weight coefficient corresponding to each point within the second target cluster range.

[0165] S604: Determine the point cloud data density and data similarity corresponding to the second numerical group based on the total number of points contained within the second target cluster range, the weight coefficient, the preset point cloud data density determination function, and the preset data similarity determination function.

[0166] S605: The sum of the point cloud data density and the data similarity is determined as the target optimization value corresponding to the second numerical group.

[0167] In the application embodiment, the effect of point cloud data clustering is further improved by determining the target optimization value corresponding to the second numerical group.

[0168] Figure 7 This is the fifth schematic diagram illustrating a point cloud data clustering process provided in some embodiments of this application. Figure 7 As shown, the process includes the following steps:

[0169] S701: Based on the point cloud data collected by the radar, determine the first distance and the first angle between the radar and the target.

[0170] S702: Randomly select a preset number of first value groups, wherein the first value group includes a second value of a second parameter arbitrarily selected from the value range of the second parameter corresponding to each preset direction.

[0171] S703: Process the preset number of first numerical groups based on a genetic algorithm to obtain a preset number of second numerical groups; determine the target optimization values ​​corresponding to the preset number of second numerical groups according to the preset number of second numerical groups and the preset target optimization algorithm.

[0172] S704: For each second numerical group, based on the parameter value of the second parameter corresponding to each preset direction included in the second numerical group, the first distance, the first angle, the angular resolution of the radar, and the target preset clustering range determination function for each preset direction, determine the second clustering range of the target object in each preset direction.

[0173] S705: Based on each of the second clustering ranges, determine each point within the second target clustering range where the target object is located; based on the first position coordinates of the cluster center of the second target clustering range, the second position coordinates of each point within the second target clustering range, the first distance, and the median value of the weight coefficients, determine the weight coefficient corresponding to each point within the second target clustering range.

[0174] S706: Based on the total number of points contained within the second target clustering range, the weight coefficient, the preset point cloud data density determination function, and the preset data similarity determination function, determine the point cloud data density and data similarity corresponding to the second numerical group; determine the sum of the point cloud data density and the data similarity as the target optimization value corresponding to the second numerical group.

[0175] S707: For each preset direction, the first clustering range of the target object in the preset direction is determined according to the third value of the second parameter corresponding to the preset direction in the optimal value group, the first distance, the first angle, the angular resolution of the radar, and the target preset clustering range determination function in the preset direction.

[0176] S708: Based on each of the first clustering ranges, determine each point located within the first target clustering range, and determine the point cloud data corresponding to each point within the first target clustering range as the point cloud data corresponding to the target object.

[0177] Based on the above embodiments, this application also provides a point cloud data clustering device. Figure 8a This is one of the structural schematic diagrams of a point cloud data clustering device provided in some embodiments of this application. For example... Figure 8a As shown, the device includes:

[0178] The first determining module 801 is used to determine the first distance and the first angle between the radar and the target based on the point cloud data collected by the radar.

[0179] The second determining module 802 is used to determine the first clustering range of the target object in the preset direction for each preset direction based on the first distance, the first angle, the angular resolution of the radar and the target preset clustering range determination function corresponding to the preset direction.

[0180] The third determining module 803 is used to determine each point within the first target cluster range according to each first cluster range, and to determine the point cloud data corresponding to each point within the first target cluster range as the point cloud data corresponding to the target object.

[0181] In one possible implementation, the second determining module 802 is further configured to determine the first clustering range of the target object in the preset direction based on the first distance, the first angle, the angular resolution of the radar, the preset value of the second parameter corresponding to the preset direction, and the target preset clustering range determining function; or to determine the value range of the second parameter, arbitrarily select a value in the value range as the first value of the second parameter, and determine the first clustering range of the target object in the preset direction based on the first value of the second parameter, the first distance, the first angle, the angular resolution of the radar, the target preset clustering range determining function in the preset direction.

[0182] Figure 8b This is a second schematic diagram of a point cloud data clustering device provided in some embodiments of this application. Figure 8b As shown, in one possible implementation, the device further includes:

[0183] Selection module 804 is used to randomly select a preset number of first numerical groups, wherein the first numerical group includes a second value of a second parameter arbitrarily selected from the value range of the second parameter corresponding to each preset direction; the preset number of first numerical groups are processed based on a genetic algorithm to obtain a preset number of second numerical groups; and a second numerical group is selected from the preset number of second numerical groups as the optimal numerical group.

[0184] The second determining module 802 is further configured to, for each preset direction, determine the first clustering range of the target object in the preset direction based on the third value of the second parameter corresponding to the preset direction in the optimal value group, the first distance, the first angle, the angular resolution of the radar, and the target preset clustering range determination function in the preset direction.

[0185] In one possible implementation, the selection module 804 is specifically used to determine the target optimization values ​​corresponding to the preset number of second value groups according to the preset number of second value groups and the preset target optimization algorithm; determine the optimal target optimization value according to the preset number of target optimization values ​​and the simulated annealing algorithm; and take the second value group corresponding to the optimal target optimization value as the optimal value group.

[0186] In one possible implementation, the selection module 804 is specifically configured to, for each second numerical group, determine the second clustering range of the target object in each preset direction based on the parameter values ​​of the second parameters corresponding to each preset direction included in the second numerical group, the first distance, the first angle, the angular resolution of the radar, and the target preset clustering range determination function for each preset direction; determine each point of the target object located within the second target clustering range based on each second clustering range; determine the weight coefficient corresponding to each point within the second target clustering range based on the first position coordinates of the cluster center of the second target clustering range, the second position coordinates of each point within the second target clustering range, the first distance, and the median value of the weight coefficient; determine the point cloud data density and data similarity corresponding to the second numerical group based on the total number of points contained in the second target clustering range, the weight coefficient, the preset point cloud data density determination function, and the preset data similarity determination function; and determine the sum of the point cloud data density and the data similarity as the target optimization value corresponding to the second numerical group.

[0187] This device can be specifically deployed in a terminal, and other functions of the terminal are described in the other embodiments above.

[0188] Based on the above embodiments, this application also provides an electronic device. Figure 9 This is a schematic diagram of an electronic device structure provided for some embodiments of this application. For example... Figure 9 As shown, it includes: processor 901, communication interface 902, memory 903 and communication bus 904, wherein processor 901, communication interface 902 and memory 903 communicate with each other through communication bus 904.

[0189] The memory 903 stores a computer program, which, when executed by the processor 901, causes the processor 901 to complete the steps of any of the point cloud data clustering methods described above.

[0190] The communication bus mentioned in the above electronic devices can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This communication bus can be divided into address bus, data bus, control bus, etc. For ease of illustration, only one thick line is used to represent it in the diagram, but this does not mean that there is only one bus or one type of bus.

[0191] The communication interface 902 is used for communication between the above-mentioned electronic device and other devices.

[0192] The memory may include random access memory (RAM) or non-volatile memory (NVM), such as at least one disk storage device. Optionally, the memory may also be at least one storage device located remotely from the aforementioned processor.

[0193] The processors mentioned above can be general-purpose processors, including central processing units, network processors (NPs), etc.; they can also be digital signal processors (DSPs), application-specific integrated circuits, field-programmable gate arrays or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc.

[0194] Based on the above embodiments, the present invention provides a computer-readable storage medium storing a computer program executable by an electronic device, wherein computer-executable instructions are used to cause a computer to execute the process performed by any of the aforementioned point cloud data clustering methods.

[0195] The aforementioned computer-readable storage medium can be any available medium or data storage device that can be accessed by the processor in an electronic device, including but not limited to magnetic storage such as floppy disks, hard disks, magnetic tapes, magneto-optical disks (MO), optical storage such as CDs, DVDs, BDs, HVDs, etc., and semiconductor storage such as ROMs, EPROMs, EEPROMs, non-volatile memory (NAND flash), solid-state drives (SSDs), etc.

[0196] In this application, based on point cloud data acquired by radar, a first distance and a first angle between the radar and the target are determined. For each preset direction, a first clustering range of the target in that preset direction is determined according to the first distance, the first angle, the radar's angular resolution, and a target preset clustering range determination function corresponding to that preset direction. Furthermore, based on each first clustering range, a first target clustering range of the target is determined, and the point cloud data corresponding to each point within the first target clustering range is identified as the point cloud data corresponding to the target. Since in this embodiment, the first distance and the first angle between the radar and the target are determined based on the point cloud data acquired by radar, and the determined first distance, first angle, and radar's angular resolution are input into the target preset clustering range determination function corresponding to each preset direction to determine the first clustering range of the target in each preset direction, the first target clustering range of the target can be determined based on each determined first clustering range. This achieves the goal of dynamically changing the clustering range with distance during point cloud data clustering.

[0197] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0198] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to this application. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0199] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0200] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0201] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A point cloud data clustering method, characterized in that, The method includes: Based on the point cloud data collected by the radar, a first distance and a first angle between the radar and the target are determined; For each preset direction, the first clustering range of the target object in the preset direction is determined based on the first distance, the first angle, the angular resolution of the radar, and the target preset clustering range determination function corresponding to the preset direction. The preset direction includes the X-axis direction, the Y-axis direction, and the Z-axis direction. Based on each of the first cluster ranges, determine each point located within the first target cluster range, and determine the point cloud data corresponding to each point within the first target cluster range as the point cloud data corresponding to the target object; The function for determining the target pre-defined clustering range corresponding to the X-axis direction is: ； The function for determining the target preset clustering range corresponding to the Y-axis direction is: ； The function for determining the target preset clustering range corresponding to the Z-axis direction is: ; in, The first cluster range along the X-axis. This is the preset second parameter corresponding to the X-axis direction. The first distance, For the radar's elevation angle resolution, For the horizontal angular resolution of the radar, The first horizontal angle included in the first angle; This represents the first clustering range along the Y-axis. This is the preset second parameter corresponding to the Y-axis direction; The first clustering range along the Z-axis. This is the preset second parameter corresponding to the Z-axis direction.

2. The method as described in claim 1, characterized in that, The step of determining the first clustering range of the target object in the preset direction based on the first distance, the first angle, the angular resolution of the radar, and the target preset clustering range determination function corresponding to the preset direction includes: Based on the first distance, the first angle, the angular resolution of the radar, the preset value of the second parameter corresponding to the preset direction, and the target preset clustering range determination function, the first clustering range of the target object in the preset direction is determined; or The range of values ​​for the second parameter is determined, and any value within the range is taken as the first value of the second parameter. Based on the first value of the second parameter, the first distance, the first angle, the angular resolution of the radar, and the target preset clustering range determination function in the preset direction, the first clustering range of the target is determined.

3. The method as described in claim 1, characterized in that, The step of determining the first clustering range of the target object in each preset direction, based on the first distance, the first angle, the angular resolution of the radar, and the target preset clustering range determination function corresponding to the preset direction, includes: A first set of values ​​of a predetermined number is randomly selected. The first set of values ​​includes a second value of a second parameter that is arbitrarily selected from the range of values ​​of the second parameter corresponding to each predetermined direction. The first set of values ​​of a preset number is processed using a genetic algorithm to obtain a second set of values ​​of a preset number. Select one of the preset number of second value groups as the optimal value group; For each preset direction, the first clustering range of the target object in that preset direction is determined based on the third value of the second parameter corresponding to that preset direction in the optimal value group, the first distance, the first angle, the angular resolution of the radar, and the target preset clustering range determination function for that preset direction.

4. The method as described in claim 3, characterized in that, The step is to select one of the second numerical groups from the preset number of second numerical groups as the optimal numerical group: Based on the preset number of second numerical groups and the preset target optimization algorithm, determine the target optimization values ​​corresponding to the preset number of second numerical groups respectively; Based on the preset number of target optimization values ​​and the simulated annealing algorithm, the optimal target optimization value is determined, and the second value group corresponding to the optimal target optimization value is taken as the optimal value group.

5. The method as described in claim 4, characterized in that, Based on the preset number of second numerical groups and the preset target optimization algorithm, the target optimization values ​​corresponding to the preset number of second numerical groups are determined as follows: For each second numerical group, based on the parameter values ​​of the second parameters corresponding to each preset direction included in the second numerical group, the first distance, the first angle, the angular resolution of the radar, and the target preset clustering range determination function for each preset direction, the second clustering range of the target object in each preset direction is determined; based on each second clustering range, each point of the target object located within the second target clustering range is determined; based on the first position coordinates of the cluster center of the second target clustering range, the second position coordinates of each point within the second target clustering range, the first distance, and the median value of the weight coefficients, the weight coefficients corresponding to each point within the second target clustering range are determined; based on the total number of points contained within the second target clustering range, the weight coefficients, the preset point cloud data density determination function, and the preset data similarity determination function, the point cloud data density and data similarity corresponding to the second numerical group are determined; the sum of the point cloud data density and the data similarity is determined as the target optimization value corresponding to the second numerical group.

6. A point cloud data clustering device, characterized in that, The device includes: The first determining module is used to determine the first distance and the first angle between the radar and the target based on the point cloud data collected by the radar. The second determining module is used to determine the first clustering range of the target object in the preset direction for each preset direction based on the first distance, the first angle, the angular resolution of the radar and the target preset clustering range determination function corresponding to the preset direction, wherein the preset direction includes the X-axis direction, the Y-axis direction and the Z-axis direction; The third determining module is used to determine each point located within the first target cluster range based on each first cluster range, and to determine the point cloud data corresponding to each point within the first target cluster range as the point cloud data corresponding to the target object; The function for determining the target pre-defined clustering range corresponding to the X-axis direction is: ； The function for determining the target preset clustering range corresponding to the Y-axis direction is: ； The function for determining the target preset clustering range corresponding to the Z-axis direction is: ; in, The first cluster range along the X-axis. This is the preset second parameter corresponding to the X-axis direction. The first distance, For the radar's elevation angle resolution, For the horizontal angular resolution of the radar, The first horizontal angle included in the first angle; This represents the first clustering range along the Y-axis. This is the preset second parameter corresponding to the Y-axis direction; The first clustering range along the Z-axis. This is the preset second parameter corresponding to the Z-axis direction.

7. The apparatus according to claim 6, characterized in that, The second determining module is further configured to determine the first clustering range of the target object in the preset direction based on the first distance, the first angle, the angular resolution of the radar, the preset value of the second parameter corresponding to the preset direction, and the target preset clustering range determining function; or to determine the value range of the second parameter, arbitrarily select a value in the value range as the first value of the second parameter, and determine the first clustering range of the target object in the preset direction based on the first value of the second parameter, the first distance, the first angle, the angular resolution of the radar, the target preset clustering range determining function in the preset direction.

8. The apparatus according to claim 6, characterized in that, The device further includes: The selection module is used to randomly select a preset number of first numerical groups, wherein the first numerical group includes a second value of a second parameter arbitrarily selected from the value range of the second parameter corresponding to each preset direction; the preset number of first numerical groups are processed based on a genetic algorithm to obtain a preset number of second numerical groups; and a second numerical group is selected from the preset number of second numerical groups as the optimal numerical group. The second determining module is further configured to, for each preset direction, determine the first clustering range of the target object in the preset direction based on the third value of the second parameter corresponding to the preset direction in the optimal value group, the first distance, the first angle, the angular resolution of the radar, and the target preset clustering range determination function in the preset direction.

9. An electronic device, characterized in that, The electronic device includes at least a processor and a memory, wherein the processor is used to execute a computer program stored in the memory to implement the steps of the point cloud data clustering method as described in any one of claims 1-5.

10. A computer-readable storage medium, characterized in that, It stores a computer program that, when executed by a processor, implements the steps of the point cloud data clustering method as described in any one of claims 1-5.

Images