Point cloud clustering method, apparatus, device, and medium

Abstract

Embodiments of the present application disclose a point cloud clustering method, device, equipment and medium. The method comprises: determining a target point cloud cluster according to point cloud quantity information and cluster minimum enclosing circle information in a candidate point cloud cluster; the candidate point cloud cluster is obtained by preliminary clustering of to-be-clustered point cloud data; determining a candidate single-target minimum enclosing circle in the target point cloud cluster according to a single-target minimum enclosing circle radius of a detection target; determining a candidate single-target minimum enclosing circle matching position from a traversal process of the candidate single-target minimum enclosing circle in a cluster minimum enclosing circle of the target point cloud cluster according to a single-target point cloud quantity threshold of the detection target; determining a final single-target minimum enclosing circle matching position according to point cloud density in the candidate single-target minimum enclosing circle matching position, and determining a final clustering result of the target point cloud cluster according to the final single-target minimum enclosing circle matching position. The technical solution can effectively solve the problem of unreasonable high-density point cloud clustering.

Description

Technical Field

[0001] This invention relates to the field of radar detection technology, and in particular to a point cloud clustering method, apparatus, equipment, and medium. Background Technology

[0002] With the development of radar detection technology, radar has received widespread attention in target detection. Existing radar point cloud clustering mainly uses the DBSCAN algorithm, which groups eligible point clouds in the neighborhood into a cluster based on the Euclidean distance of the point cloud.

[0003] After radar point cloud data is processed by the DBSCAN algorithm, it will result in several clusters. When targets are sparse, the clusters can intuitively represent a single target. However, when targets are dense, for example, if the distance between two targets is less than the threshold set by the algorithm, the point cloud data will agglomerate into a single cluster after DBSCAN clustering. See [link to relevant documentation]. Figure 1A .in, Figure 1A The actual clustering situation of point cloud data in the data is as follows: Figure 1B As shown, because the Euclidean distance between adjacent point clouds is relatively short, two targets are clustered into one target, making it impossible to obtain the true target motion trajectory.

[0004] In related technologies, data region grids are divided based on the velocity and distance of radar point cloud data, and DBSCAN clustering is performed based on these grids to address the multi-density clustering problem of millimeter-wave radar data. However, this approach requires reference to the velocity of the point cloud for grid region division. Since targets are typically not rigid bodies, the velocity deviations between point clouds are significant, which can easily lead to unreasonable grid region division. Summary of the Invention

[0005] This invention provides a point cloud clustering method, apparatus, device, and medium that can optimize the clustering of high-density point clouds based on the minimum bounding circle of radar point cloud data, effectively solving the problem of unreasonable clustering of high-density point clouds.

[0006] According to one aspect of the present invention, a point cloud clustering method is provided, the method comprising:

[0007] The target point cloud cluster is determined based on the number of points in the candidate point cloud cluster and the minimum bounding circle of the cluster; wherein, the candidate point cloud cluster is obtained by performing preliminary clustering on the point cloud data to be clustered;

[0008] The candidate minimum bounding circle of a single target in the target point cloud cluster is determined based on the radius of the minimum bounding circle of the single target detected;

[0009] Based on the single target point cloud number threshold of the detected target, the matching position of the minimum bounding circle of the candidate single target is determined during the traversal of the minimum bounding circle of the target point cloud cluster.

[0010] The final minimum bounding circle matching position of a single target is determined based on the point cloud density in the candidate single target minimum bounding circle matching position, and the final clustering result of the target point cloud cluster is determined based on the final single target minimum bounding circle matching position.

[0011] According to another aspect of the present invention, a point cloud clustering apparatus is provided, comprising:

[0012] The target point cloud cluster determination module is used to determine the target point cloud cluster based on the point cloud quantity information and the minimum bounding circle information of the candidate point cloud clusters; wherein, the candidate point cloud clusters are obtained by performing preliminary clustering on the point cloud data to be clustered;

[0013] The candidate single-target minimum bounding circle determination module is used to determine the candidate single-target minimum bounding circle in the target point cloud cluster based on the radius of the single-target minimum bounding circle of the detected target;

[0014] The matching position determination module is used to determine the matching position of the minimum bounding circle of the candidate single target during the traversal of the minimum bounding circle of the cluster of the target point cloud based on the single target point cloud number threshold of the detected target.

[0015] The final clustering result determination module is used to determine the final minimum bounding circle matching position of a single target based on the point cloud density in the candidate single target minimum bounding circle matching position, and to determine the final clustering result of the target point cloud cluster based on the final minimum bounding circle matching position of the single target.

[0016] According to another aspect of the present invention, a point cloud clustering electronic device is provided, the electronic device comprising:

[0017] At least one processor; and

[0018] A memory communicatively connected to the at least one processor; wherein,

[0019] The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the point cloud clustering method according to any embodiment of the present invention.

[0020] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions for causing a processor to execute and implement the point cloud clustering method according to any embodiment of the present invention.

[0021] The technical solution of this invention can optimize the clustering of high-density point clouds based on the minimum bounding circle of radar point cloud data, thereby effectively solving the problem of unreasonable clustering of high-density point clouds.

[0022] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of the present invention, 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 the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1A This is a schematic diagram of a candidate point cloud cluster in the prior art;

[0025] Figure 1B This is a schematic diagram of the actual clustering of a candidate point cloud cluster in the existing technology;

[0026] Figure 2 This is a flowchart of a point cloud clustering method provided in Embodiment 1 of the present invention;

[0027] Figure 3A This is a schematic diagram of a single-target minimum enclosing circle provided according to Embodiment 1 of the present invention;

[0028] Figure 3B This is a schematic diagram of the minimum enclosing circle of a candidate point cloud cluster according to Embodiment 1 of the present invention;

[0029] Figure 3C This is a schematic diagram of a candidate single-target minimum enclosing circle provided in Embodiment 1 of the present invention;

[0030] Figure 4 This is a schematic diagram of a final clustering result provided in Embodiment 1 of the present invention;

[0031] Figure 5 This is a flowchart of a point cloud clustering method provided in Embodiment 2 of the present invention;

[0032] Figure 6A This is a schematic diagram of a target being detected being occluded, according to Embodiment 2 of the present invention;

[0033] Figure 6B This is a schematic diagram of the position of an obscured detection target according to Embodiment 2 of the present invention;

[0034] Figure 7A This is a schematic diagram of an unobstructed view provided according to Embodiment 2 of the present invention;

[0035] Figure 7B This is a schematic diagram of partial occlusion provided according to Embodiment 2 of the present invention;

[0036] Figure 7C This is a schematic diagram of a full occlusion relationship provided in Embodiment 2 of the present invention;

[0037] Figure 8 This is a schematic diagram of another partial occlusion provided according to Embodiment 2 of the present invention;

[0038] Figure 9 This is a schematic diagram of the structure of a point cloud clustering device according to Embodiment 3 of the present invention;

[0039] Figure 10 This is a schematic diagram of the structure of an electronic device that implements a point cloud clustering method according to an embodiment of the present invention. Detailed Implementation

[0040] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0041] It should be noted that the terms "first," "second," "target," etc., used in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0042] Example 1

[0043] Figure 2This is a flowchart of a point cloud clustering method provided in Embodiment 1 of the present invention. This embodiment is applicable to the optimization and clustering of high-density point cloud data. The method can be executed by a point cloud clustering device, which can be implemented in hardware and / or software and can be configured in an electronic device with data processing capabilities. Figure 2 As shown, the method includes:

[0044] S110, determine the target point cloud cluster based on the point cloud quantity information and the minimum enclosing circle information in the candidate point cloud cluster; wherein, the candidate point cloud cluster is obtained by performing preliminary clustering on the point cloud data to be clustered.

[0045] Here, the point cloud data to be clustered can refer to radar point cloud data awaiting clustering processing. Candidate point cloud clusters can refer to point cloud clusters obtained through preliminary clustering of the point cloud data to be clustered. For example, the DBSCAN clustering algorithm can be used to perform preliminary clustering of the point cloud data to be clustered based on the Euclidean distance between each point cloud data to be clustered, thus obtaining candidate point cloud clusters. It should be noted that when the detected targets are densely packed, multiple target point clouds may easily be clustered into one cluster. Point cloud quantity information can be used to characterize the number of points in a candidate point cloud cluster.

[0046] The minimum enclosing circle of a cluster can refer to the minimum enclosing circle constructed from candidate point cloud clusters. It should be noted that for a set P consisting of n points on a plane, if there exists a circle containing all points in P with the smallest radius, then that circle is the minimum enclosing circle of set P. The minimum enclosing circle of a finite point set P is unique. For example, the minimum enclosing circle of set P can be determined using the Wang Wei farthest point priority asymptotic method. The specific process is as follows: (1) Select three points a, b, and c arbitrarily in the point set P; (2) Construct a minimum circle containing points a, b, and c. The circumference may pass through these three points, or it may pass through only two of the points but contain the third point (in this case, the two points on the circumference must be located at the two ends of a diameter of the circle); (3) Find the point d in the point set that is farthest from the center of the circle constructed in step (2). If point d is already inside the circle or on the circumference, then the circle is the desired circle; otherwise, proceed to step (4); (4) Select three points from a, b, c, and d such that a circle containing these four points generated by them is the smallest. Then these three points become the new a, b, and c, and return to step (2). For candidate point cloud clusters obtained by performing preliminary clustering on the point cloud data to be clustered, each candidate point cloud cluster is a point set P. The minimum enclosing circle information of the cluster can be used to characterize the feature information of the minimum enclosing circle of the cluster. For example, the minimum enclosing circle information of the cluster can include the center and radius of the minimum enclosing circle of the cluster. The target point cloud cluster can refer to the point cloud cluster selected from the candidate point cloud clusters based on the information of the number of points in the candidate point cloud clusters and the information of the minimum bounding circle of the cluster.

[0047] In this embodiment, indoor personnel can be detected using radar. The target is the human body. In this radar scenario for indoor personnel detection, the radar point cloud data generated by a single target can be clustered using the DBSCAN algorithm to generate clusters. Then, a minimum bounding circle for each single target is constructed, with a threshold value for its radius. Figure 3A As shown. Here, r represents the radius of the minimum bounding circle of a single target. The radius of the minimum bounding circle of a single target can be predetermined according to the type of the target being detected; in this embodiment of the invention, this size is not limited. When there are many people indoors, it is easy for multiple target point clouds to be clustered into a candidate point cloud cluster. In this case, the radius of the minimum bounding circle of the cluster constructed from this candidate point cloud cluster will exceed the set radius threshold, and the number of point clouds inside the candidate point cloud cluster will also exceed the set number threshold, such as... Figure 3B As shown in the diagram. Here, R represents the radius of the minimum bounding circle of the candidate point cloud cluster. It should be noted that multiple single targets may exist within the candidate point cloud clusters obtained through DBSCAN clustering. This technical solution can address the target overlap problem when there are two or three single targets in a candidate point cloud cluster. Therefore, it is first necessary to determine the number of single targets in the candidate point cloud cluster based on the number of points in the cluster and the minimum bounding circle information. If there are two or three single targets, the candidate point cloud cluster can be identified as the target point cloud cluster; otherwise, it is discarded without further processing.

[0048] In this embodiment, optionally, determining the target point cloud cluster based on the point cloud quantity information and the minimum enclosing circle information of the candidate point cloud cluster includes: determining the number of single targets in the candidate point cloud cluster based on the point cloud quantity information and the radius information of the minimum enclosing circle of the cluster; and determining the target point cloud cluster as the cluster whose single target quantity satisfies the single target quantity condition.

[0049] The single-target quantity condition can be set to two or three single targets in the candidate point cloud cluster. In this embodiment, the number of point clouds in the candidate point cloud cluster and the radius of the minimum enclosing circle of the cluster are compared with the number of point clouds in the minimum enclosing circle of the single target and the radius, respectively, to obtain the possible number of single targets in the candidate point cloud cluster. For example, when the number of point clouds in the candidate point cloud cluster meets a preset multiple range of the number of point clouds in the minimum enclosing circle of the single target, and the radius of the minimum enclosing circle of the cluster meets a preset multiple range of the radius of the minimum enclosing circle of the single target, it can be considered that there may be two or three single targets in the candidate point cloud cluster. For example, assuming that the number of point clouds in the candidate point cloud cluster is greater than the number of point clouds in the minimum enclosing circle of the single target but less than twice the number of point clouds, and the radius of the minimum enclosing circle of the cluster is greater than the radius of the minimum enclosing circle of the single target but less than twice the radius, it can be considered that there may be two single targets in the candidate point cloud cluster. It should be noted that this embodiment does not impose any limitations on the preset multiple range of the number of point clouds in the minimum enclosing circle of the single target and the preset multiple range of the radius of the minimum enclosing circle of the single target, and can be set according to actual needs. If the number of single targets meets the single target quantity condition, that is, the number of single targets is two or three, then the candidate point cloud cluster can be determined as the target point cloud cluster.

[0050] This scheme, through this setting, can quickly and accurately determine the number of individual targets in a candidate point cloud cluster based on the number of points in the cluster and the radius of the minimum enclosing circle of the cluster, so as to subsequently determine the target point cloud cluster based on the number of individual targets and the condition of the number of individual targets.

[0051] S120, determine the candidate single target minimum bounding circle in the target point cloud cluster based on the radius of the single target minimum bounding circle of the detected target.

[0052] The minimum bounding circle of a candidate single target can refer to the minimum bounding circle of a single target within the target point cloud cluster. It is understood that the number of single targets contained in the target point cloud cluster indicates the number of minimum bounding circles of candidate single targets within the cluster; that is, the number of minimum bounding circles of candidate single targets is equal to the number of single targets. In this embodiment, the radius of the minimum bounding circle of the detected target can be directly determined as the minimum bounding circle of a candidate single target within the target point cloud cluster; that is, the radius of the minimum bounding circle of the detected target is equal to the radius of the minimum bounding circle of a candidate single target within the target point cloud cluster.

[0053] It should be noted that when there are two single targets in the target point cloud cluster (i.e., there are two candidate single target minimum enclosing circles in the target point cloud cluster), since the boundary point cloud of the minimum enclosing circle of the target point cloud cluster is also the boundary point cloud of the candidate single target minimum enclosing circle, the candidate single target minimum enclosing circle is inscribed in the target point cloud cluster's minimum enclosing circle. Assuming the center of the target point cloud cluster's minimum enclosing circle is O and its radius is R, and the radius of the candidate single target's minimum enclosing circle is r, then the difference between the radius of the center of the target point cloud cluster's minimum enclosing circle and the radius of the candidate single target's minimum enclosing circle can be expressed as Rr. In this case, the trajectory of the candidate single target's minimum enclosing circle can be represented as a circle with center O and radius Rr, as shown below. Figure 3C The dotted lines in the diagram illustrate this. Furthermore, since the point clouds of actual detected targets are independent, and multiple targets do not share a single point cloud, it is necessary to ensure that the positions of the two inscribed circles do not intersect. When there are three single targets in the target point cloud cluster (i.e., there are three candidate single-target minimum bounding circles in the target point cloud cluster), assuming the three candidate single-target minimum bounding circles are circle c1, circle c2, and circle c3, at least two of the candidate single-target minimum bounding circles (e.g., circle c1 and circle c2) are inscribed in the minimum bounding circle of the target point cloud cluster, and circle c3 is circumscribed in either circle c1 or circle c2. Here, c1, c2, and c3 are the centers of the different candidate single-target minimum bounding circles.

[0054] S130, based on the threshold of the number of single target point clouds of the detected target, determine the matching position of the minimum bounding circle of the candidate single target from the process of traversing the minimum bounding circle of the candidate single target in the cluster minimum bounding circle of the target point cloud cluster.

[0055] The single-target point cloud quantity threshold can refer to a pre-set reference value for the number of single-target point clouds, which can be set according to actual needs. This embodiment does not limit this. In this embodiment, after determining the minimum bounding circle of candidate single targets in the target point cloud cluster, the center of the minimum bounding circle of the candidate single target can be traversed along a preset trajectory (for example, the preset trajectory can be a circle with center O and radius Rr) in the minimum bounding circle of the target point cloud cluster. Based on the single-target point cloud quantity threshold of the detected target, the matching position of the minimum bounding circle of the candidate single target is determined from the traversal process of the minimum bounding circle of the candidate single target in the minimum bounding circle of the target point cloud cluster.

[0056] In this embodiment, optionally, the single-target quantity condition is that the number of single targets is two; the candidate single-target minimum bounding circle includes a first single-target minimum bounding circle and a second single-target minimum bounding circle. The first single-target minimum bounding circle and the second single-target minimum bounding circle represent two different candidate single-target minimum bounding circles. The fact that the number of single targets is two indicates that the minimum bounding circle of the target point cloud cluster contains two candidate single-target minimum bounding circles.

[0057] Accordingly, based on the single-target point cloud quantity threshold of the detected target, the matching position of the candidate single-target minimum bounding circle is determined during the traversal of the candidate single-target minimum bounding circle in the cluster minimum bounding circle of the target point cloud cluster. This includes: using the first single-target minimum bounding circle to traverse the cluster minimum bounding circle according to a preset trajectory, determining each position where the number of point clouds in the first single-target minimum bounding circle is greater than the single-target point cloud quantity threshold as the candidate first single-target minimum bounding circle matching position; wherein, the preset trajectory is the trajectory inscribed in the cluster minimum bounding circle; after determining each candidate first single-target minimum bounding circle matching position, the second single-target minimum bounding circle traverses the cluster minimum bounding circle according to the preset trajectory, starting from the candidate first single-target minimum bounding circle matching position, and determining each position where the second single-target minimum bounding circle satisfies the preset position condition as the candidate second single-target minimum bounding circle matching position corresponding to the candidate first single-target minimum bounding circle matching position; wherein, the preset position condition is that the number of point clouds in the second single-target minimum bounding circle is greater than the single-target point cloud quantity threshold and the second single-target minimum bounding circle and the candidate first single-target minimum bounding circle do not intersect.

[0058] In this embodiment, when the rotation angle of the center of the first single target minimum enclosing circle does not exceed 360 degrees, the first single target minimum enclosing circle is first used to traverse the cluster minimum enclosing circle according to a preset trajectory inscribed in the cluster minimum enclosing circle, and each position in the traversal process that makes the number of point clouds in the first single target minimum enclosing circle greater than the single target point cloud number threshold is determined as a candidate first single target minimum enclosing circle matching position. During the process of traversing the first single-target minimum bounding circle within the cluster minimum bounding circle according to a preset trajectory, after determining a candidate first single-target minimum bounding circle matching position, the second single-target minimum bounding circle, starting from that candidate first single-target minimum bounding circle matching position, also traverses within the cluster minimum bounding circle according to a preset trajectory inscribed in the cluster minimum bounding circle. The positions where the second single-target minimum bounding circle satisfies a preset position condition (i.e., the number of point clouds in the second single-target minimum bounding circle is greater than a single-target point cloud number threshold and the second single-target minimum bounding circle and the candidate first single-target minimum bounding circle do not intersect) are determined as the candidate second single-target minimum bounding circle matching positions corresponding to that candidate first single-target minimum bounding circle matching position. Following this process, when the first single-target minimum bounding circle completes its traversal within the cluster minimum bounding circle according to the preset trajectory, the candidate second single-target minimum bounding circle matching positions corresponding to each candidate first single-target minimum bounding circle matching position can be obtained.

[0059] It should be noted that a candidate first single-target minimum enclosing circle matching position may correspond to one or more candidate second single-target minimum enclosing circle matching positions, which can be determined according to the actual situation. Each candidate first single-target minimum enclosing circle matching position and its corresponding candidate second single-target minimum enclosing circle matching position form a matching position pair. For example, assuming the determined candidate first single-target minimum enclosing circle matching position is a1, the corresponding candidate second single-target minimum enclosing circle matching positions are b1 and b2; and if the determined candidate first single-target minimum enclosing circle matching position is a2, the corresponding candidate second single-target minimum enclosing circle matching position is b3, then a total of three matching position pairs can be formed: a1-b1, a1-b2, and a2-b3.

[0060] This solution, through this configuration, can quickly and accurately determine the minimum bounding circle matching position of the candidate single targets when the target point cloud cluster contains two single targets.

[0061] S140, determine the final minimum bounding circle matching position of the single target based on the point cloud density in the candidate single target minimum bounding circle matching position, and determine the final clustering result of the target point cloud cluster based on the final minimum bounding circle matching position of the single target.

[0062] Point cloud density can be used to characterize the density of point cloud distribution within the minimum enclosing circle of a candidate single target. Specifically, point cloud density can be determined by the ratio of the number of points within the minimum enclosing circle to the area of ​​the minimum enclosing circle. It can be understood that a higher point cloud density indicates a denser point cloud distribution within the minimum enclosing circle, meaning a greater number of points; conversely, a lower point cloud density indicates a sparser point cloud distribution, meaning a smaller number of points.

[0063] In this embodiment, firstly, for each matching position pair, the point cloud density of the minimum bounding circle of each candidate single target is determined. Then, the point cloud densities of the minimum bounding circles of each candidate single target are weighted and summed. Based on the weighted summation result, the average point cloud density of the minimum bounding circles of each candidate single target corresponding to each matching position pair is determined. Then, from the average point cloud densities corresponding to each matching position pair, the matching position pair with the highest average point cloud density is selected. This pair is determined as the final single-target minimum bounding circle matching position, and the final clustering result of the target point cloud cluster is determined based on the final single-target minimum bounding circle matching position. Figure 4 This is a schematic diagram of a final clustering result provided in Embodiment 1 of the present invention. Figure 4 As shown, the target point cloud cluster contains two final single-target minimum bounding circles, that is, it contains two single targets.

[0064] The technical solution of this invention determines the target point cloud cluster based on the point cloud quantity information and the minimum bounding circle information of the clusters in the candidate point cloud clusters. The candidate point cloud clusters are obtained by preliminary clustering of the point cloud data to be clustered. The minimum bounding circle radius of the single target in the target point cloud cluster is used to determine the candidate single target minimum bounding circle. Based on the single target point cloud quantity threshold of the detected target, the matching position of the candidate single target minimum bounding circle is determined during the traversal of the candidate single target minimum bounding circle within the minimum bounding circle of the target point cloud cluster. The final single target minimum bounding circle matching position is determined based on the point cloud density at the matching position, and the final clustering result of the target point cloud cluster is determined based on the final single target minimum bounding circle matching position. This technical solution can optimize high-density point cloud clustering based on the minimum bounding circle of radar point cloud data, thereby effectively solving the problem of unreasonable high-density point cloud clustering.

[0065] In this embodiment, optionally, the single-target quantity condition is that the number of single targets is three; the candidate single-target minimum bounding circles include a first single-target minimum bounding circle, a second single-target minimum bounding circle, and a third single-target minimum bounding circle. The first, second, and third single-target minimum bounding circles each represent three different candidate single-target minimum bounding circles. The fact that the number of single targets is three indicates that the minimum bounding circle of the target point cloud cluster contains three candidate single-target minimum bounding circles.

[0066] Accordingly, based on the single-target point cloud quantity threshold of the detected target, the matching position of the candidate single-target minimum bounding circle is determined during the traversal of the candidate single-target minimum bounding circle within the cluster minimum bounding circle of the target point cloud cluster. This includes: traversing the first single-target minimum bounding circle within the cluster minimum bounding circle according to a preset trajectory, and determining each position where the number of point clouds in the first single-target minimum bounding circle is greater than the single-target point cloud quantity threshold as the candidate first single-target minimum bounding circle matching position; wherein, the preset trajectory is the trajectory inscribed within the cluster minimum bounding circle; after determining each candidate first single-target minimum bounding circle matching position, the second single-target minimum bounding circle traverses the cluster minimum bounding circle according to the preset trajectory, starting from the candidate first single-target minimum bounding circle matching position, and determining each position where the second single-target minimum bounding circle satisfies a preset position condition as the candidate second single-target minimum bounding circle matching position corresponding to the candidate first single-target minimum bounding circle matching position; wherein, the preset position condition is... The number of point clouds in the second single-target minimum bounding circle is greater than the single-target point cloud number threshold, and the second single-target minimum bounding circle and the candidate first single-target minimum bounding circle do not intersect. After determining the matching position of each candidate second single-target minimum bounding circle, the third single-target minimum bounding circle traverses the cluster minimum bounding circle according to the second preset trajectory, and determines each position that makes the third single-target minimum bounding circle satisfy the second preset position condition as the candidate third single-target minimum bounding circle matching position corresponding to the matching position of the candidate first bounding circle and the candidate second bounding circle. The second preset trajectory is a trajectory that is tangent to the candidate first single-target bounding circle or the candidate second single-target bounding circle. The second preset position condition is that the number of point clouds in the third single-target minimum bounding circle is greater than the single-target point cloud number threshold, and the third single-target minimum bounding circle and the candidate first single-target minimum bounding circle and the candidate second single-target minimum bounding circle do not intersect, while the third single-target minimum bounding circle is within the cluster minimum bounding circle of the target point cloud cluster.

[0067] In this embodiment, when the rotation angle of the center of the first single target minimum enclosing circle does not exceed 360 degrees, the first single target minimum enclosing circle is first used to traverse the cluster minimum enclosing circle according to a preset trajectory inscribed in the cluster minimum enclosing circle, and each position in the traversal process that makes the number of point clouds in the first single target minimum enclosing circle greater than the single target point cloud number threshold is determined as a candidate first single target minimum enclosing circle matching position. During the process of traversing the minimum bounding circle of the first single target within the minimum bounding circle of the cluster according to a preset trajectory, after determining a matching position of the first single target minimum bounding circle, the second single target minimum bounding circle, starting from the matching position of the first single target minimum bounding circle, also traverses the minimum bounding circle of the cluster according to a preset trajectory inscribed in the minimum bounding circle of the cluster. The position that makes the second single target minimum bounding circle satisfy the preset position condition (i.e., the number of point clouds in the second single target minimum bounding circle is greater than the threshold of the number of single target point clouds and the second single target minimum bounding circle and the candidate first single target minimum bounding circle do not intersect) during the traversal of the second single target minimum bounding circle is determined as the matching position of the candidate second single target minimum bounding circle corresponding to the matching position of the first single target minimum bounding circle. During the process of the second single-target minimum enclosing circle traversing the cluster minimum enclosing circle according to the preset trajectory, after each candidate second single-target minimum enclosing circle matching position is determined, the third single-target minimum enclosing circle traverses the cluster minimum enclosing circle according to the second preset trajectory that is tangent to the candidate first single-target enclosing circle or the candidate second single-target enclosing circle. The position that makes the third single-target minimum enclosing circle satisfy the second preset position condition (the number of point clouds in the third single-target minimum enclosing circle is greater than the single target point cloud number threshold and the third single-target minimum enclosing circle does not intersect with the candidate first single-target minimum enclosing circle and the candidate second single-target minimum enclosing circle, and the third single-target minimum enclosing circle is within the cluster minimum enclosing circle of the target point cloud cluster) is determined as the candidate third single-target minimum enclosing circle matching position corresponding to the candidate first enclosing circle matching position and the candidate second enclosing circle matching position.

[0068] Optionally, before determining each position that makes the minimum bounding circle of the second single target satisfy a preset position condition as the matching position of the candidate minimum bounding circle of the first single target, the method further includes:

[0069] During the traversal of the second single-target minimum bounding circle, the occlusion relationship with the matching position of the corresponding candidate first single-target minimum bounding circle is determined; wherein, the occlusion relationship includes no occlusion, partial occlusion and full occlusion;

[0070] The point cloud compensation parameters for the minimum bounding circle matching position of the second single target are determined based on the occlusion relationship.

[0071] After compensating the number of point clouds in the matching position of the second single target minimum enclosing circle according to the point cloud compensation parameters, it is determined whether the preset position conditions are met.

[0072] Optionally, before determining each position that makes the minimum bounding circle of the third single target satisfy the second preset position condition as the matching position of the candidate third single target minimum bounding circle corresponding to the matching position of the candidate first bounding circle and the candidate second bounding circle, the method further includes:

[0073] During the traversal of the third single-target minimum bounding circle, the occlusion relationship between the matching position of the corresponding candidate first single-target minimum bounding circle and the matching position of the candidate first single-target minimum bounding circle is determined; wherein, the occlusion relationship includes no occlusion, partial occlusion and full occlusion;

[0074] The point cloud compensation parameters for the minimum bounding circle matching position of the third single target are determined based on the occlusion relationship.

[0075] After compensating the number of points in the minimum enclosing circle matching position of the third single target according to the point cloud compensation parameters, it is determined whether the preset position conditions are met.

[0076] The occlusion relationship can be determined based on the positional relationship between the third center of the minimum enclosing circle of the third single target, the first center of the corresponding candidate first single target minimum enclosing circle matching position, and the second center of the candidate first single target minimum enclosing circle matching position. For example, it can be determined based on the shape of the triangle formed by the centers or based on the distance between the centers and the signal transmitting device; this is not limited here. Furthermore, determining the point cloud compensation parameters based on the occlusion relationship, and compensating for the number of points based on the compensation parameters, can refer to the case when there are two targets; this embodiment does not impose any limitations here.

[0077] This solution, through this configuration, can quickly and accurately determine the minimum bounding circle matching position of the candidate single targets when the target point cloud cluster contains three single targets.

[0078] Example 2

[0079] Figure 5 This is a flowchart of a point cloud clustering method provided in Embodiment 2 of the present invention. This embodiment is an optimization based on the above embodiment. Figure 5 As shown, the method in this embodiment specifically includes the following steps:

[0080] S210, determine the target point cloud cluster based on the point cloud quantity information and the minimum enclosing circle information in the candidate point cloud cluster; wherein, the candidate point cloud cluster is obtained by preliminary clustering of the point cloud data to be clustered, and the target point cloud cluster includes two single targets.

[0081] S220, determine the candidate single target minimum bounding circle in the target point cloud cluster based on the radius of the single target minimum bounding circle of the detected target; wherein, the candidate single target minimum bounding circle includes the first single target minimum bounding circle and the second single target minimum bounding circle.

[0082] S230, the first single-target minimum bounding circle is used to traverse the cluster minimum bounding circle according to a preset trajectory, and each position that makes the number of point clouds in the first single-target minimum bounding circle greater than the single-target point cloud number threshold is determined as the candidate first single-target minimum bounding circle matching position; wherein, the preset trajectory is the trajectory inscribed in the cluster minimum bounding circle.

[0083] The specific implementation methods of S210-S230 can be found in the detailed description of S110-S130, and will not be repeated here. It should be noted that due to the complexity of the indoor detection environment and the tendency for occlusion, the number of point clouds of the detected target may be insufficient, thus affecting the accuracy of point cloud clustering. Figure 6A This is a schematic diagram illustrating the detection of an occluded target according to Embodiment 2 of the present invention. Figure 6A As shown, the obstruction area is a fan-shaped region formed by the two tangents of radar A and the minimum enclosing circle (circle c1) of the first single target, where m and n are the two tangent points. Because the actual detected target is obstructed, the point cloud is incomplete, making it impossible to cluster the two targets. However, since the radar's detection signal is an electromagnetic wave, its obstruction area has a certain range. Therefore, the obstruction area can be divided based on the determined single target, and point cloud compensation can be performed on the obstruction area.

[0084] In this embodiment, a point cloud compensation method is proposed to address the issue of occluded detection targets. The final clustering result of the target point cloud cluster is determined based on the compensated point cloud data. This effectively avoids the problem of inaccurate point cloud clustering caused by target occlusion, thus helping to further improve the accuracy of point cloud clustering. In this embodiment, the optimization method for point cloud clustering is illustrated using an example where the target point cloud cluster includes two single targets, and the minimum bounding circle of the candidate single targets includes the minimum bounding circle of the first single target and the minimum bounding circle of the second single target.

[0085] S240, after determining the matching position of each candidate first single target minimum bounding circle, the second single target minimum bounding circle traverses the cluster minimum bounding circle according to a preset trajectory, starting from the matching position of the candidate first single target minimum bounding circle. During the traversal of the second single target minimum bounding circle, the occlusion relationship with the corresponding matching position of the candidate first single target minimum bounding circle is determined; wherein, the occlusion relationship includes no occlusion, partial occlusion and full occlusion.

[0086] In this embodiment, after determining the matching position of the minimum bounding circle of the candidate first single target, the minimum bounding circle of the second single target can be traversed in the minimum bounding circle of the cluster according to a preset trajectory, starting from each matching position of the minimum bounding circle of the candidate first single target. For the specific traversal process, please refer to Embodiment 1. Figure 6B This is a schematic diagram illustrating the location of an obstructed detection target according to Embodiment 2 of the present invention. Figure 6B As shown, when the second single-target minimum bounding circle c2 is traversed, if the triangle formed by points A, c1, and c2 is located within the fan-shaped region, it indicates that the point cloud within the second single-target minimum bounding circle (circle c2) may be occluded. In this case, it is necessary to determine the occlusion relationship with the corresponding candidate first single-target minimum bounding circle during the traversal of the second single-target minimum bounding circle. The occlusion relationship can include no occlusion (e.g., ...). Figure 7A As shown), partial occlusion (such as) Figure 7B (as shown) or full occlusion (such as) Figure 7C (As shown). Among them, Figure 7A , Figure 7B and Figure 7C These are schematic diagrams illustrating unobstructed, partially obstructed, and fully obstructed conditions according to Embodiment 2 of the present invention. Figure 7B and Figure 7C As shown, when the occlusion relationship between the position matching the minimum bounding circle of the corresponding candidate first single target is determined to be partial occlusion or full occlusion, point cloud compensation is required.

[0087] In this embodiment, optionally, during the traversal of the second single-target minimum enclosing circle, determining the occlusion relationship with the corresponding candidate first single-target minimum enclosing circle includes: determining the first center position of the candidate first single-target minimum enclosing circle, the angle formed by the tangent point between the transmitted signal of the detection signal transmitter and the candidate first single-target minimum enclosing circle and the position of the detection signal transmitter as a first reference angle; determining the angle formed by the first center position, the second center position of the second single-target minimum enclosing circle, and the position of the detection signal transmitter as a second reference angle; determining an occlusion reference angle based on the difference between the second reference angle and the first reference angle; wherein, the occlusion reference angle is determined based on the second center position, the target transmitted signal beam, and the detection signal transmitter, and the target transmitted signal beam is the transmitted signal beam that is tangent to the candidate first single-target minimum enclosing circle; determining an occlusion reference distance based on the second distance between the second center position and the position of the detection signal transmitter, and the occlusion reference angle; wherein, the occlusion reference distance represents the distance from the second center position to the target transmitted signal beam; and determining the occlusion relationship based on the occlusion reference angle and the occlusion reference distance.

[0088] The first center position can refer to the center position of the minimum enclosing circle of the candidate first single target. The first reference angle can refer to the angle formed by the first center position, the tangent point between the transmitted signal from the detection signal transmitter and the minimum enclosing circle of the candidate first single target, and the position of the detection signal transmitter. The second center position can refer to the center position of the minimum enclosing circle of the second single target. The second reference angle can refer to the angle formed by the first center position, the second center position, and the position of the detection signal transmitter. The obstruction reference angle can be determined based on the second center position, the target transmitted signal beam, and the detection signal transmitter. The target transmitted signal beam can refer to the transmitted signal beam that is tangent to the minimum enclosing circle of the candidate first single target. Specifically, the magnitude of the obstruction reference angle can be determined based on the difference between the second reference angle and the first reference angle. For example, the magnitude of the obstruction reference angle is equal to the difference between the second reference angle and the first reference angle. The second distance can refer to the distance between the second center position and the position of the detection signal transmitter. The obstruction reference distance can be used to represent the distance from the second center position to the target transmitted signal beam.

[0089] Figure 8 This is a schematic diagram of another partial occlusion provided in Embodiment 2 of the present invention. (See diagram below.) Figure 8 As shown, q1 is the tangent point between the transmitted signal of the detection signal transmitter (radar A) and the minimum enclosing circle (circle c1) of the first candidate single target, where c1 is the center position of the first circle. q is the intersection point of the target transmitted signal beam (ray Aq1) and circle c2. q2 and q3 are both on the target transmitted signal beam (ray Aq1), and q2 is the perpendicular point from the second circle center position c2 to ray Aq1, while q3 is an intersection point of ray Aq1 and circle c2. ∠c1Aq1 is the first reference angle, ∠c1Ac2 is the second reference angle, and ∠q2Ac2 is the blocking reference angle, where ∠q2Ac2 = ∠c1Ac2 - ∠c1Aq1. The radii of circles c1 and c2 are both r. The distance of line segment Ac2 represents the second distance, and the distance of line segment c2q2 represents the blocking reference distance.

[0090] In this embodiment, optionally, determining the occlusion relationship based on the occlusion reference angle and the occlusion reference distance includes: if And if θ3 > 0, then the occlusion relationship is determined to be unoccluded; if And if θ3 < 0, then the occlusion relationship is determined to be total occlusion; if This determines the occlusion relationship to be partial occlusion. Among them, The occlusion reference distance is θ3, the occlusion reference angle is r, and the minimum enclosing circle radius of the single target is r. For the second distance, θ3 = θ2 - θ1, where θ2 is the second reference angle and θ1 is the first reference angle. The first distance is the distance between the first center position and the detection signal transmitting device. The second distance is the distance between the second center position and the detection signal transmitting device. The distance between the centers of the first and second circles is denoted as .

[0091] In this embodiment, it is assumed that the first reference angle ∠c1Aq1 is denoted as θ1, the second reference angle ∠c1Ac2 is denoted as θ2, and the occlusion reference angle ∠q2Ac2 is denoted as θ3, and the following conditions are met: θ3 = θ2 - θ1. Where, The first distance between the first center position and the detection signal transmitting device (i.e. Figure 8 (distance from the median segment Ac1) The second distance between the second center position and the detection signal transmitting device (i.e. Figure 8 (distance from the midline segment Ac2) The distance between the centers of the first and second circles (i.e.) Figure 8 (The distance between the median segments c1 and c2). The occlusion relationship can then be determined based on the occlusion reference angle and the occlusion reference distance.

[0092] Specifically, assuming the minimum enclosing circle radius of a single target is r, if And if θ3 > 0, then the occlusion relationship can be determined to be unoccluded; if And if θ3 < 0, then the occlusion relationship can be determined to be total occlusion; if Therefore, the occlusion relationship can be determined to be partial occlusion. Figure 8 For example, the area where circle c2 is occluded is the region enclosed by line qq3 and arc qq3, that is... Figure 8 The shaded area in the image.

[0093] This solution, through such settings, can quickly and accurately determine the occlusion relationship based on the occlusion reference angle and occlusion reference distance.

[0094] S250 determines the point cloud compensation parameters for the minimum bounding circle matching position of the second single target based on the occlusion relationship.

[0095] Here, the point cloud compensation parameters can refer to the compensated point cloud data. In this embodiment, after determining the occlusion relationship, point cloud compensation parameters for the matching position of the second single-target minimum bounding circle can be determined for different occlusion relationships. Optionally, determining the point cloud compensation parameters for the matching position of the second single-target minimum bounding circle based on the occlusion relationship includes: determining the occlusion area of ​​the second single-target minimum bounding circle occluded by the candidate first single-target minimum bounding circle based on the second distance, the occlusion reference angle, and the radius of the single-target minimum bounding circle; and determining the point cloud compensation parameters based on the occlusion area and the occlusion relationship.

[0096] The occlusion area can refer to the area occupied by the occluded region. In this embodiment, the occlusion area where the second single-target minimum bounding circle is occluded by the candidate first single-target minimum bounding circle is first determined based on the second distance, the occlusion reference angle, and the radius of the single-target minimum bounding circle. Figure 8 For example, and in, Let represent the area of ​​the triangle formed by points q, c2, and q3. Therefore, we can... Get the area of ​​the obstructed region ( Figure 8 (The shaded area in the image). After determining the area of ​​the occluded region, point cloud compensation parameters can be further determined based on the area of ​​the occluded region and the occlusion relationship. For example, point cloud compensation parameters can be expressed as follows:

[0097]

[0098] Where C represents the point cloud compensation parameter, and k represents the preset compensation coefficient, which can be set according to actual needs. In this embodiment, the preset compensation coefficient k is not limited.

[0099] This solution, through such settings, can quickly and accurately determine the point cloud compensation parameters corresponding to different occlusion relationships based on the area of ​​the occlusion region and the occlusion relationship.

[0100] S260, after compensating the number of points in the minimum enclosing circle matching position of the second single target according to the point cloud compensation parameters, determine whether the preset position conditions are met.

[0101] In this embodiment, after determining the point cloud compensation parameters, the number of point clouds in the matching position of the second single-target minimum bounding circle can be further compensated according to the point cloud compensation parameters to determine whether the preset position condition is met. Optionally, the preset position condition is that the compensated point cloud number in the second single-target minimum bounding circle is greater than the single-target point cloud number threshold and the second single-target minimum bounding circle and the candidate first single-target minimum bounding circle do not intersect; wherein, the compensated point cloud number is the number of point clouds after compensation based on the point cloud compensation parameters determined according to the occlusion relationship between the matching position of the second single-target minimum bounding circle and the corresponding candidate first single-target minimum bounding circle.

[0102] In this embodiment, during the traversal of the second single-target minimum bounding circle, if the occlusion relationship with the corresponding candidate first single-target minimum bounding circle is determined to be partial or full occlusion, then the point cloud in the second single-target minimum bounding circle needs to be compensated. Specifically, the number of point clouds in the second single-target minimum bounding circle can be added to the point cloud compensation parameter, and the sum can be determined as the number of compensated point clouds. Then, the number of compensated point clouds is used to determine whether the preset position condition is met. If it is met, the corresponding position of the second single-target minimum bounding circle is determined as the candidate second single-target minimum bounding circle matching position; if it is not met, the traversal continues.

[0103] S270, determine each position that makes the minimum bounding circle of the second single target satisfy the preset position condition as the candidate minimum bounding circle matching position corresponding to the candidate first minimum bounding circle of the single target; wherein, the preset position condition is that the number of point clouds in the minimum bounding circle of the second single target is greater than the single target point cloud number threshold and the minimum bounding circle of the second single target and the candidate first minimum bounding circle of the single target do not intersect.

[0104] S280, determine the final minimum bounding circle matching position of the single target based on the point cloud density in the minimum bounding circle matching position of the candidate single target, and determine the final clustering result of the target point cloud cluster based on the final minimum bounding circle matching position of the single target.

[0105] The technical solution of this invention, before determining each position that makes the minimum bounding circle of the second single target satisfy a preset position condition as the candidate second single target minimum bounding circle matching position corresponding to the candidate first single target minimum bounding circle matching position, determines the occlusion relationship with the corresponding candidate first single target minimum bounding circle matching position during the traversal of the second single target minimum bounding circle; wherein, the occlusion relationship includes no occlusion, partial occlusion, and full occlusion; the point cloud compensation parameters for the matching position of the second single target minimum bounding circle are determined according to the occlusion relationship; after compensating the number of points in the matching position of the second single target minimum bounding circle according to the point cloud compensation parameters, it is determined whether the preset position condition is met. This technical solution can perform point cloud compensation for occluded detection targets and determine the final clustering result of the target point cloud cluster based on the point cloud data after point cloud compensation. It can effectively avoid the problem of inaccurate point cloud clustering due to the occlusion of the detection target, and can optimize the clustering of high-density point clouds based on the minimum bounding circle of radar point cloud data, effectively solving the problem of unreasonable high-density point cloud clustering, and further improving the accuracy of point cloud clustering.

[0106] Example 3

[0107] Figure 9This is a schematic diagram of a point cloud clustering device provided in Embodiment 3 of the present invention. This device can execute the point cloud clustering method provided in any embodiment of the present invention, and possesses the corresponding functional modules and beneficial effects for executing the method. For example... Figure 9 As shown, the device includes:

[0108] The target point cloud cluster determination module 310 is used to determine the target point cloud cluster based on the point cloud quantity information and the minimum enclosing circle information in the candidate point cloud clusters; wherein, the candidate point cloud clusters are obtained by performing preliminary clustering on the point cloud data to be clustered;

[0109] The candidate single-target minimum bounding circle determination module 320 is used to determine the candidate single-target minimum bounding circle in the target point cloud cluster based on the radius of the single-target minimum bounding circle of the detected target;

[0110] The matching position determination module 330 is used to determine the matching position of the minimum bounding circle of the candidate single target during the traversal of the minimum bounding circle of the cluster of the target point cloud cluster, based on the single target point cloud number threshold of the detected target.

[0111] The final clustering result determination module 340 is used to determine the final single-target minimum bounding circle matching position based on the point cloud density in the candidate single-target minimum bounding circle matching position, and to determine the final clustering result of the target point cloud cluster based on the final single-target minimum bounding circle matching position.

[0112] Optionally, the target point cloud cluster determination module 310 is specifically used for:

[0113] The number of single targets in the candidate point cloud cluster is determined based on the point cloud quantity information and the radius information of the minimum enclosing circle of the cluster.

[0114] The clusters of target points whose single target quantity satisfies the single target quantity condition are identified as target point cloud clusters.

[0115] Optionally, the single-target quantity condition is that the number of single targets is two; the candidate single-target minimum enclosing circle includes a first single-target minimum enclosing circle and a second single-target minimum enclosing circle;

[0116] Accordingly, the matching position determination module 330 includes:

[0117] The first matching position determination unit is used to traverse the cluster minimum bounding circle according to a preset trajectory using the first single target minimum bounding circle, and determine each position that makes the number of point clouds in the first single target minimum bounding circle greater than the single target point cloud number threshold as a candidate first single target minimum bounding circle matching position; wherein, the preset trajectory is a trajectory inscribed in the cluster minimum bounding circle;

[0118] The second matching position determination unit is used to determine the matching position of each candidate first single target minimum bounding circle after determining the matching position of each candidate first single target minimum bounding circle. The second single target minimum bounding circle is then traversed in the cluster minimum bounding circle according to the preset trajectory, starting from the matching position of the candidate first single target minimum bounding circle. Each position that makes the second single target minimum bounding circle satisfy a preset position condition is determined as the matching position of the candidate second single target minimum bounding circle corresponding to the matching position of the candidate first single target minimum bounding circle. The preset position condition is that the number of point clouds in the second single target minimum bounding circle is greater than the single target point cloud number threshold and the second single target minimum bounding circle and the candidate first single target minimum bounding circle do not intersect.

[0119] Optionally, the matching position determination module 330 further includes:

[0120] An occlusion relationship determination unit is used to determine the occlusion relationship with the corresponding candidate first single-target minimum bounding circle during the traversal of the second single-target minimum bounding circle before determining each position that makes the second single-target minimum bounding circle satisfy a preset position condition as the candidate second single-target minimum bounding circle matching position corresponding to the candidate first single-target minimum bounding circle matching position; wherein, the occlusion relationship includes no occlusion, partial occlusion, and full occlusion;

[0121] The point cloud compensation parameter determination unit is used to determine the point cloud compensation parameters for the matching position of the minimum bounding circle of the second single target based on the occlusion relationship.

[0122] The preset position condition judgment unit is used to determine whether the preset position condition is met after compensating the number of point clouds in the matching position of the minimum enclosing circle of the second single target according to the point cloud compensation parameters.

[0123] Optionally, the preset position condition is that the number of compensated point clouds in the second single-target minimum bounding circle is greater than the single-target point cloud number threshold and the second single-target minimum bounding circle and the candidate first single-target minimum bounding circle do not intersect; wherein, the number of compensated point clouds is the number of point clouds after compensating the number of point clouds based on the point cloud compensation parameters determined according to the occlusion relationship between the matching positions of the second single-target minimum bounding circle and the corresponding candidate first single-target minimum bounding circle.

[0124] Optionally, the occlusion relationship determination unit is used for:

[0125] The first reference angle determination subunit is used to determine the first center position of the candidate first single target minimum enclosing circle, and the angle formed by the transmission signal of the detection signal transmitter, the tangent point of the candidate first single target minimum enclosing circle, and the position of the detection signal transmitter as the first reference angle;

[0126] The second reference angle determination subunit is used to determine the angle formed by the first circle center position, the second circle center position of the second single target minimum enclosing circle, and the position of the detection signal transmitting device as the second reference angle.

[0127] An occlusion reference angle determination subunit is used to determine an occlusion reference angle based on the difference between the second reference angle and the first reference angle; wherein, the occlusion reference angle is determined based on the second circle center position and the target transmitted signal beam, and the target transmitted signal beam is the signal beam in the transmitted signal that is tangent to the minimum enclosing circle of the candidate first single target;

[0128] The occlusion reference distance determination subunit is used to determine the occlusion reference distance based on the second distance between the second center position and the position of the detection signal transmitting device, and the occlusion reference angle; wherein, the occlusion reference distance represents the distance from the second center position to the target transmitted signal beam;

[0129] The occlusion relationship determination subunit determines the occlusion relationship based on the occlusion reference angle and the occlusion reference distance.

[0130] Optionally, the point cloud compensation parameter determination unit is used for:

[0131] The area of ​​the occlusion region where the second single-target minimum bounding circle is occluded by the candidate first single-target minimum bounding circle is determined based on the second distance, the occlusion reference angle, and the radius of the single-target minimum bounding circle;

[0132] The point cloud compensation parameters are determined based on the area of ​​the occlusion region and the occlusion relationship.

[0133] Optionally, the single target quantity condition is that the number of single targets is three; the candidate single target minimum enclosing circle includes a first single target minimum enclosing circle, a second single target minimum enclosing circle, and a third single target minimum enclosing circle;

[0134] Accordingly, the matching position determination module 330 includes:

[0135] The third matching position determination unit is used to traverse the first single-target minimum enclosing circle in the cluster minimum enclosing circle according to a preset trajectory, and determine each position that makes the number of point clouds in the first single-target minimum enclosing circle greater than the single-target point cloud number threshold as a candidate first single-target minimum enclosing circle matching position; wherein, the preset trajectory is the trajectory inscribed in the cluster minimum enclosing circle;

[0136] The fourth matching position determination unit is used to determine the matching position of each candidate first single target minimum bounding circle after determining the matching position of each candidate first single target minimum bounding circle. Then, the second single target minimum bounding circle is traversed within the cluster minimum bounding circle according to a preset trajectory, starting from the matching position of the candidate first single target minimum bounding circle. Each position that satisfies a preset position condition is determined as the matching position of the candidate second single target minimum bounding circle corresponding to the matching position of the candidate first single target minimum bounding circle. The preset position condition is that the number of point clouds in the second single target minimum bounding circle is greater than the single target point cloud number threshold and the second single target minimum bounding circle and the candidate first single target minimum bounding circle do not intersect.

[0137] The fifth matching position determination unit is used to determine the matching position of the minimum bounding circle of the third single target after determining the matching position of each candidate second single target minimum bounding circle. The minimum bounding circle of the third single target is traversed in the minimum bounding circle of the cluster according to the second preset trajectory, and each position that makes the minimum bounding circle of the third single target satisfy the second preset position condition is determined as the matching position of the candidate first bounding circle and the candidate second bounding circle. The second preset trajectory is a trajectory that is tangent to the candidate first single target bounding circle or the candidate second single target bounding circle. The second preset position condition is that the number of point clouds in the minimum bounding circle of the third single target is greater than the threshold of the number of single target point clouds and the minimum bounding circle of the third single target does not intersect with the minimum bounding circles of the candidate first single target and the candidate second single target. At the same time, the minimum bounding circle of the third single target is within the minimum bounding circle of the cluster of target point clouds.

[0138] The point cloud clustering device provided in this embodiment of the invention can execute a point cloud clustering method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the method.

[0139] Example 4

[0140] Figure 10 A schematic diagram of an electronic device 10 that can be used to implement embodiments of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0141] like Figure 10As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded from storage unit 18 into the RAM 13. The RAM 13 may also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0142] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0143] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as point cloud clustering methods.

[0144] In some embodiments, the point cloud clustering method may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or mounted on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the point cloud clustering method described above may be performed. Alternatively, in other embodiments, processor 11 may be configured to perform the point cloud clustering method by any other suitable means (e.g., by means of firmware).

[0145] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0146] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0147] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0148] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0149] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0150] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0151] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0152] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A point cloud clustering method, characterized in that, The method includes: The target point cloud cluster is determined based on the number of points in the candidate point cloud cluster and the minimum bounding circle of the cluster; wherein, the candidate point cloud cluster is obtained by performing preliminary clustering on the point cloud data to be clustered; The candidate minimum bounding circles of a single target in the target point cloud cluster are determined based on the radius of the minimum bounding circle of the single target; wherein, the candidate minimum bounding circle of a single target refers to the minimum bounding circle of a single target in the target point cloud cluster, and the number of candidate minimum bounding circles of a single target is equal to the number of single targets; Based on the single-target point cloud quantity threshold of the detected target, the matching position of the candidate single-target minimum bounding circle is determined during the traversal of the candidate single-target minimum bounding circle in the cluster minimum bounding circle of the target point cloud cluster; wherein, the center of the candidate single-target minimum bounding circle is traversed along a preset trajectory, and the candidate single-target minimum bounding circles do not intersect. The final minimum bounding circle matching position of a single target is determined based on the point cloud density in the candidate single target minimum bounding circle matching position, and the final clustering result of the target point cloud cluster is determined based on the final single target minimum bounding circle matching position.

2. The method according to claim 1, characterized in that, The target point cloud cluster is determined based on the number of points in the candidate point cloud cluster and the minimum bounding circle of the cluster, including: The number of single targets in the candidate point cloud cluster is determined based on the point cloud quantity information and the radius information of the minimum enclosing circle of the cluster. The clusters of target points whose single target quantity satisfies the single target quantity condition are identified as target point cloud clusters.

3. The method according to claim 2, characterized in that, The single-target quantity condition is that the number of single targets is two; the candidate single-target minimum enclosing circle includes a first single-target minimum enclosing circle and a second single-target minimum enclosing circle. Accordingly, based on the single-target point cloud quantity threshold of the detected target, the matching position of the candidate single-target minimum bounding circle is determined during the traversal of the candidate single-target minimum bounding circle in the cluster minimum bounding circle of the target point cloud cluster, including: The first single-target minimum bounding circle is used to traverse the cluster minimum bounding circle according to a preset trajectory, and each position that makes the number of point clouds in the first single-target minimum bounding circle greater than the single-target point cloud number threshold is determined as a candidate first single-target minimum bounding circle matching position; wherein, the preset trajectory is the trajectory inscribed in the cluster minimum bounding circle; After determining a candidate first single-target minimum bounding circle matching position, the second single-target minimum bounding circle traverses the cluster of minimum bounding circles according to the preset trajectory, starting from the candidate first single-target minimum bounding circle matching position. Each position that makes the second single-target minimum bounding circle satisfy a preset position condition is determined as the candidate second single-target minimum bounding circle matching position corresponding to the candidate first single-target minimum bounding circle matching position. The preset position condition is that the number of point clouds in the second single-target minimum bounding circle is greater than the single-target point cloud number threshold and the second single-target minimum bounding circle and the candidate first single-target minimum bounding circle do not intersect.

4. The method according to claim 3, characterized in that, Before determining each position that makes the minimum bounding circle of the second single target satisfy a preset position condition as the matching position of the candidate minimum bounding circle of the first single target, the method further includes: During the traversal of the second single-target minimum bounding circle, the occlusion relationship with the matching position of the corresponding candidate first single-target minimum bounding circle is determined; wherein, the occlusion relationship includes no occlusion, partial occlusion and full occlusion; The point cloud compensation parameters for the minimum bounding circle matching position of the second single target are determined based on the occlusion relationship. After compensating the number of point clouds in the matching position of the second single target minimum enclosing circle according to the point cloud compensation parameters, it is determined whether the preset position conditions are met.

5. The method according to claim 4, characterized in that, During the traversal of the second single-target minimum bounding circle, the occlusion relationship with the corresponding candidate first single-target minimum bounding circle is determined, including: The first reference angle is determined by the position of the first center of the candidate first single target minimum enclosing circle, the angle formed by the tangent point between the transmitted signal of the detection signal transmitting device and the minimum enclosing circle of the candidate first single target and the position of the detection signal transmitting device; The angle formed by the position of the first circle center, the position of the second circle center of the minimum enclosing circle of the second single target, and the position of the detection signal transmitting device is determined as the second reference angle; The occlusion reference angle is determined based on the difference between the second reference angle and the first reference angle; wherein, the occlusion reference angle is determined based on the position of the second circle center and the target transmitted signal beam, and the target transmitted signal beam is the signal beam in the transmitted signal that is tangent to the minimum enclosing circle of the candidate first single target; The blocking reference distance is determined based on the second distance between the second center position and the position of the detection signal transmitting device, and the blocking reference angle; wherein, the blocking reference distance represents the distance from the second center position to the target transmitted signal beam; The occlusion relationship is determined based on the occlusion reference angle and the occlusion reference distance.

6. The method according to claim 5, characterized in that, The point cloud compensation parameters for the minimum bounding circle matching position of the second single target are determined based on the occlusion relationship, including: The area of ​​the occlusion region where the second single-target minimum bounding circle is occluded by the candidate first single-target minimum bounding circle is determined based on the second distance, the occlusion reference angle, and the radius of the single-target minimum bounding circle; The point cloud compensation parameters are determined based on the area of ​​the occlusion region and the occlusion relationship.

7. The method according to claim 2, characterized in that, The single-target quantity condition is that the number of single targets is three; the candidate single-target minimum enclosing circle includes a first single-target minimum enclosing circle, a second single-target minimum enclosing circle, and a third single-target minimum enclosing circle. Accordingly, based on the single-target point cloud quantity threshold of the detected target, the matching position of the candidate single-target minimum bounding circle is determined during the traversal of the candidate single-target minimum bounding circle in the cluster minimum bounding circle of the target point cloud cluster, including: The first single-target minimum bounding circle is traversed in the cluster minimum bounding circle according to a preset trajectory. Each position that makes the number of point clouds in the first single-target minimum bounding circle greater than the single-target point cloud number threshold is determined as a candidate first single-target minimum bounding circle matching position; wherein, the preset trajectory is the trajectory inscribed in the cluster minimum bounding circle. After determining a candidate first single-target minimum bounding circle matching position, the second single-target minimum bounding circle traverses the cluster minimum bounding circles according to a preset trajectory, starting from the candidate first single-target minimum bounding circle matching position. Each position that satisfies a preset position condition is determined as the candidate second single-target minimum bounding circle matching position corresponding to the candidate first single-target minimum bounding circle matching position. The preset position condition is that the number of point clouds in the second single-target minimum bounding circle is greater than the single-target point cloud number threshold, and the second single-target minimum bounding circle and the candidate first single-target minimum bounding circle do not intersect. After determining the matching position of each candidate second single target minimum enclosing circle, the third single target minimum enclosing circle is traversed in the cluster minimum enclosing circle according to the second preset trajectory. Each position that makes the third single target minimum enclosing circle satisfy the second preset position condition is determined as the matching position of the candidate first enclosing circle and the candidate second enclosing circle. The second preset trajectory is a trajectory that is tangent to the candidate first single target enclosing circle or the candidate second single target enclosing circle. The second preset position condition is that the number of point clouds in the third single target minimum enclosing circle is greater than the threshold of the number of single target point clouds and the third single target minimum enclosing circle does not intersect with the candidate first single target minimum enclosing circle and the candidate second single target minimum enclosing circle. At the same time, the third single target minimum enclosing circle is within the cluster minimum enclosing circle of the target point cloud cluster.

8. A point cloud clustering device, characterized in that, The device includes: The target point cloud cluster determination module is used to determine the target point cloud cluster based on the point cloud quantity information and the minimum bounding circle information of the candidate point cloud clusters; wherein, the candidate point cloud clusters are obtained by performing preliminary clustering on the point cloud data to be clustered; The candidate single-target minimum bounding circle determination module is used to determine the candidate single-target minimum bounding circle in the target point cloud cluster based on the radius of the single-target minimum bounding circle of the detected target; wherein, the candidate single-target minimum bounding circle refers to the single-target minimum bounding circle in the target point cloud cluster, and the number of candidate single-target minimum bounding circles is equal to the number of single targets; The matching position determination module is used to determine the matching position of the minimum bounding circle of the candidate single target during the traversal of the minimum bounding circle of the cluster of the target point cloud based on the single target point cloud number threshold of the detected target; wherein, the center of the minimum bounding circle of the candidate single target is traversed along a preset trajectory, and the minimum bounding circles of the candidate single target do not intersect. The final clustering result determination module is used to determine the final minimum bounding circle matching position of a single target based on the point cloud density in the candidate single target minimum bounding circle matching position, and to determine the final clustering result of the target point cloud cluster based on the final minimum bounding circle matching position of the single target.

9. A point cloud clustering electronic device, characterized in that, The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the point cloud clustering method according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that cause a processor to execute the point cloud clustering method according to any one of claims 1-7.

Images