Clutter identification method, equipment, device and storage medium
Through density clustering algorithm and DBSCAN recognition radar detection points, combined with shape information in the target scene, the problem of complex shape and small size clutter recognition in synesthesia integrated system is solved, and the accuracy and efficiency of clutter recognition are improved.
Patent Information
- Application Number
- CN202310532728.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2023-05-11
- Publication Date
- 2024-11-12
AI Technical Summary
The prior art is difficult to effectively identify complex shapes and smaller sizes in synesthesia integrated systems, especially when the target is to be perceived to be in a clutter background, which makes it difficult to observe the target.
The density clustering algorithm is used to cluster the radar detection points, and the radar detection points caused by clutter are identified based on the cluster size and shape information of the target to be perceived in the target scene, including the use of the DBSCAN algorithm and the identification of L-shaped, inverse L-shaped and horizontal. Linear cluster shape.
It improves the accuracy of clutter recognition and can identify clutter caused by complex objects, thereby improving the difficulty and recognition efficiency of target observation.
Smart Images

Figure CN118938135A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of radar data processing technology, and in particular to a clutter identification method, equipment, device and storage medium. Background Art
[0002] When clutter and the target echo are displayed simultaneously on the radar chart of the integrated system, it will make it difficult to observe the target. If the target is in the clutter background, it is not easy to distinguish which echoes are the target echoes from the clutter echoes.
[0003] An existing clutter identification scheme is to identify clutter by the geometric features of the objects that cause clutter (hereinafter referred to as the geometric feature method). The geometric feature method fits the two-dimensional line segments between any two perceived radar detection points. When the two-dimensional line segments between certain radar detection points that are close in position can be fitted relatively ideally, it can be considered that the two-dimensional line segments between these radar detection points define possible reflection surfaces. The geometric feature method requires the geometric features of the clutter to be known in advance, and it cannot be used for complex object shapes. It can usually only be used to identify clutter caused by linear objects that are large enough. Summary of the invention
[0004] In view of the problems existing in the prior art, the embodiments of the present application provide a clutter identification method, device, apparatus and storage medium.
[0005] In a first aspect, an embodiment of the present application provides a clutter identification method, comprising:
[0006] Acquire multiple radar detection points obtained by detecting the target range in the target scene;
[0007] Clustering the multiple radar detection points based on a density clustering algorithm;
[0008] According to the clustering result and the cluster size and shape information corresponding to the target to be sensed in the target scene, the radar detection points caused by clutter in the target range are identified.
[0009] Optionally, the cluster size and shape information corresponding to the target to be sensed includes a plurality of cluster size and shape information units, each of which contains information about a cluster size and information about at least one cluster shape.
[0010] Optionally, the cluster shape includes at least one of an L-shape, an inverted L-shape and a horizontal line shape.
[0011] Optionally, identifying radar detection points caused by clutter within the target range according to the clustering result and the cluster size and shape information corresponding to the target to be sensed in the target scene includes:
[0012] The radar detection points in the target cluster that do not match the multiple cluster size and shape information units in the clustering result are identified as the radar detection points caused by clutter.
[0013] Optionally, the method further comprises:
[0014] The algorithm parameters used by the density clustering algorithm are determined according to the perception performance index and / or the communication performance index in the target scene.
[0015] Optionally, the density clustering algorithm includes a density-based clustering algorithm with noise DBSCAN.
[0016] Optionally, the method further comprises:
[0017] The number of the targets to be sensed in the target scene is determined according to the number of clusters corresponding to the targets to be sensed.
[0018] Optionally, the method further comprises:
[0019] The long axis direction of the first target to be sensed is determined according to the long side direction of the cluster corresponding to the first target to be sensed.
[0020] In a second aspect, an embodiment of the present application further provides an electronic device, including a memory, a transceiver, and a processor;
[0021] A memory for storing a computer program; a transceiver for transmitting and receiving data under the control of the processor; and a processor for reading the computer program in the memory and performing the following operations:
[0022] Acquire multiple radar detection points obtained by detecting the target range in the target scene;
[0023] Clustering the multiple radar detection points based on a density clustering algorithm;
[0024] According to the clustering result and the cluster size and shape information corresponding to the target to be sensed in the target scene, the radar detection points caused by clutter in the target range are identified.
[0025] Optionally, the cluster size and shape information corresponding to the target to be sensed includes a plurality of cluster size and shape information units, each of which contains information about a cluster size and information about at least one cluster shape.
[0026] Optionally, the cluster shape includes at least one of an L-shape, an inverted L-shape and a horizontal line shape.
[0027] Optionally, identifying radar detection points caused by clutter within the target range according to the clustering result and the cluster size and shape information corresponding to the target to be sensed in the target scene includes:
[0028] The radar detection points in the target cluster that do not match the multiple cluster size and shape information units in the clustering result are identified as the radar detection points caused by clutter.
[0029] Optionally, the operation further includes:
[0030] The algorithm parameters used by the density clustering algorithm are determined according to the perception performance index and / or the communication performance index in the target scene.
[0031] Optionally, the density clustering algorithm includes a density-based clustering algorithm with noise DBSCAN.
[0032] Optionally, the operation further includes:
[0033] The number of the targets to be sensed in the target scene is determined according to the number of clusters corresponding to the targets to be sensed.
[0034] Optionally, the operation further includes:
[0035] The long axis direction of the first target to be sensed is determined according to the long side direction of the cluster corresponding to the first target to be sensed.
[0036] In a third aspect, an embodiment of the present application further provides a clutter identification device, comprising:
[0037] An acquisition unit, used to acquire multiple radar detection points obtained by detecting a target range in a target scene;
[0038] A clustering unit, used for clustering the plurality of radar detection points based on a density clustering algorithm;
[0039] The identification unit is used to identify the radar detection points caused by clutter within the target range according to the clustering results and the cluster size and shape information corresponding to the target to be sensed in the target scene.
[0040] Optionally, the cluster size and shape information corresponding to the target to be sensed includes a plurality of cluster size and shape information units, each of which contains information about a cluster size and information about at least one cluster shape.
[0041] Optionally, the cluster shape includes at least one of an L-shape, an inverted L-shape and a horizontal line shape.
[0042] Optionally, identifying radar detection points caused by clutter within the target range according to the clustering result and the cluster size and shape information corresponding to the target to be sensed in the target scene includes:
[0043] The radar detection points in the target cluster that do not match the multiple cluster size and shape information units in the clustering result are identified as the radar detection points caused by clutter.
[0044] Optionally, the device further comprises:
[0045] The first determination unit is used to determine the algorithm parameters used by the density clustering algorithm according to the perception performance index and / or the communication performance index in the target scene.
[0046] Optionally, the density clustering algorithm includes a density-based clustering algorithm with noise DBSCAN.
[0047] Optionally, the device further comprises:
[0048] The second determining unit is used to determine the number of the to-be-perceived targets in the target scene according to the number of clusters corresponding to the to-be-perceived targets.
[0049] Optionally, the device further comprises:
[0050] The third determining unit is used to determine the long axis direction of the first target to be sensed according to the long side direction of the cluster corresponding to the first target to be sensed.
[0051] In a fourth aspect, an embodiment of the present application further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and the computer program is used to enable a computer to execute the clutter identification method described in the first aspect as described above.
[0052] In a fifth aspect, an embodiment of the present application further provides a communication device, wherein a computer program is stored in the communication device, and the computer program is used to enable the communication device to execute the clutter identification method described in the first aspect as described above.
[0053] In a sixth aspect, an embodiment of the present application further provides a processor-readable storage medium, wherein the processor-readable storage medium stores a computer program, and the computer program is used to enable a processor to execute the clutter identification method described in the first aspect as described above.
[0054] In a seventh aspect, an embodiment of the present application further provides a chip product, wherein a computer program is stored in the chip product, and the computer program is used to enable the chip product to execute the clutter identification method described in the first aspect as described above.
[0055] The clutter identification method, device, apparatus, and storage medium provided in the embodiments of the present application cluster multiple radar detection points within a target range in a target scene through a density clustering algorithm, and identify radar detection points caused by clutter within the target range in the clustering results according to the cluster size and shape information corresponding to the target to be sensed in the target scene. The present application can identify clutter caused by objects with complex shapes and small sizes, thereby improving the accuracy of clutter identification. BRIEF DESCRIPTION OF THE DRAWINGS
[0056] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the related technologies, the following is a brief introduction to the drawings required for use in the embodiments or the related technical descriptions. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.
[0057] Figure 1 It is a flow chart of the clutter identification method provided in the embodiment of the present application;
[0058] Figure 2 It is a schematic diagram of the clustering process of DBSCAN provided in an embodiment of the present application;
[0059] Figure 3 This is a schematic diagram of an L-shaped intersection autonomous driving simulation scenario provided by an embodiment of the present application;
[0060] Figure 4 This is one of the clustering result schematic diagrams of the simulation scenario provided in the embodiment of the present application;
[0061] Figure 5 This is the second schematic diagram of clustering results of the simulation scenario provided in the embodiment of the present application;
[0062] Figure 6 This is the third schematic diagram of the clustering results of the simulation scenario provided in the embodiment of the present application;
[0063] Figure 7 This is one of the schematic diagrams of clustering results caused by the reflection points of the vehicles to be sensed provided in the embodiments of the present application;
[0064] Figure 8 This is the second schematic diagram of the clustering result caused by the reflection points of the vehicle to be sensed provided in the embodiment of the present application;
[0065] Fig. 9 is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application;
[0066] Fig.10 It is a schematic diagram of the structure of the clutter identification device provided in an embodiment of the present application. DETAILED DESCRIPTION
[0067] In the embodiments of the present application, the term "and / or" describes the association relationship of the associated objects, indicating that three relationships may exist. For example, A and / or B may represent: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the associated objects before and after are in an "or" relationship.
[0068] In the embodiments of the present application, the term "plurality" refers to two or more than two, and other quantifiers are similar.
[0069] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.
[0070] In order to facilitate a clearer understanding of the technical solutions of the embodiments of the present application, some technical contents related to the embodiments of the present application are first introduced.
[0071] The target to be detected by the synaesthesia integration system is called the target to be sensed, such as vehicles, pedestrians, obstacles in the autonomous driving scene, or people walking in the indoor scene and the mobile automatic guided vehicle (AGV), etc. can all be used as the target to be sensed. However, there are often various backgrounds around these targets to be sensed, such as objects, guardrails on both sides of the road, buildings and plants, indoor walls and floors. The echoes generated by these backgrounds are called clutter or passive interference. These clutter interferences will affect the subsequent processing of the perception data. The embodiment of the present application provides a method for clutter recognition.
[0072] Figure 1 A schematic diagram of the flow chart of the clutter identification method provided in the embodiment of the present application, such as Figure 1 As shown, the method comprises the following steps:
[0073] Step 100: Acquire multiple radar detection points obtained by detecting a target range in a target scene.
[0074] Step 101: Cluster multiple radar detection points based on a density clustering algorithm.
[0075] Step 102: According to the clustering result and the cluster size and shape information corresponding to the target to be sensed in the target scene, the radar detection points caused by clutter within the target range are identified.
[0076] Specifically, the target scene may be a scene that the synaesthesia integration system needs to perceive, such as an autonomous driving scene, an indoor scene, etc. The target range may be a range that the synaesthesia integration system detects. When the synaesthesia integration system detects the target range in the target scene, it may transmit a detection beam to the target range through a radar to scan, obtain multiple echo signals, and use the local maximum value of the echo signal intensity as the radar detection point.
[0077] After obtaining multiple radar detection points, the multiple radar detection points can be clustered based on a density clustering algorithm, and the radar detection points can be clustered into multiple clusters as clustering results.
[0078] The density clustering algorithm used when clustering multiple radar detection points may be an existing density clustering algorithm or other density clustering algorithms, which is not specifically limited here.
[0079] Optionally, the density clustering algorithm may include a density-based clustering algorithm with noise (Density Based Spatial Clustering of Application with Noise, DBSCAN).
[0080] The core idea of DBSCAN is to classify a set of points into clusters based on two parameters: radius (epsilon) and the minimum number of points within the radius (minpts). For points within a cluster, the number of points contained in the neighborhood of a given radius must be no less than the minimum number of points within the radius. Points that meet these requirements are called "core points" of the cluster. Points in a cluster that do not meet the minimum number of points within the radius but are within the neighborhood of the core point are called border points. All points that are connected to each other with respect to the given parameter radius and the minimum number of points within the radius are classified into one cluster, and points not included in any cluster are noise.
[0081] Figure 2 The schematic diagram of the clustering process of DBSCAN provided in the embodiment of the present application is as follows: Figure 2 As shown, for a given radius and a minimum number of points within the radius of 4, Figure 2 The number of points within the radius of each solid point A is equal to 4 and these solid points are adjacent. These solid points are clustered into the same cluster, and these solid points are the core points of this cluster. Figure 2The hollow points B and C in are within the radius of the solid points near them, so the hollow points B and C are still clustered into the same cluster as the solid points. However, the number of points contained in the circle with points B and C as the radius is less than 4, so points B and C are edge points of the cluster. The number of points within the given radius of the hollow point N is less than 4, and point N is not within the radius of any solid point, so point N is clustered into another cluster.
[0082] The targets to be perceived in the target scene are the targets to be detected by the synaesthesia integration system, such as vehicles, pedestrians, obstacles, etc. in the autonomous driving scene.
[0083] According to the possible size and shape of the target to be perceived in the target scene, the possible size and shape of the cluster of radar detection points caused by the target to be perceived in the clustering result can be estimated in advance, and then the cluster size and shape information corresponding to the target to be perceived in the target scene is determined according to the possible size and shape of the cluster of radar detection points caused by the target to be perceived, and stored in the synaesthesia integration system.
[0084] The cluster size and shape information corresponding to the target to be perceived in the target scene may include at least one cluster size threshold and multiple cluster shape information, or include multiple cluster size and multiple cluster shape information, or be other forms of cluster size and shape information, and the embodiments of the present application do not make specific limitations on this.
[0085] The synaesthesia integration system can identify the clusters of radar detection points caused by clutter within the target range based on the cluster size and shape information corresponding to the target to be sensed in the target scene and the clustering results, and then can identify the radar detection points caused by clutter within the target range.
[0086] The clutter identification method provided in the embodiment of the present application clusters multiple radar detection points within the target range in the target scene through a density clustering algorithm, and identifies the radar detection points caused by clutter within the target range in the clustering results according to the cluster size and shape information corresponding to the target to be sensed in the target scene. The present application can identify clutter caused by objects with complex shapes and small sizes, thereby improving the accuracy of clutter identification.
[0087] Optionally, the cluster size and shape information corresponding to the target to be sensed includes multiple cluster size and shape information units, and each cluster size and shape information unit includes information of a cluster size and information of at least one cluster shape.
[0088] Specifically, the cluster size and shape information corresponding to the target to be perceived in the target scene determined according to the possible size and shape of the cluster of radar detection points clustered caused by the target to be perceived may include multiple cluster size and shape information units, and each cluster size and shape information unit may correspond to a target to be perceived in the target scene.
[0089] For example, in a highway scene, cars, trucks, motorcycles, and pedestrians waiting for perception targets can correspond to different cluster size and shape information units.
[0090] When the target to be sensed is at different positions, the clusters of radar detection points caused by them may have different shapes. The cluster shape in the shape information unit determined by its shape may also be one or more. Therefore, each cluster size and shape information unit may contain a cluster size information and at least one cluster shape information.
[0091] Optionally, the cluster shape may include at least one of an L-shape, an inverted L-shape, and a horizontal line shape.
[0092] In one embodiment, when the target scene is a road scene and the target to be sensed is a vehicle, the shape of the target to be sensed can be determined to be a rectangle. At this time, according to the position of the vehicle to be sensed within the detection range, the cluster shape of the cluster of radar detection points caused by the vehicle to be sensed can be determined.
[0093] For example, when the vehicle to be sensed is on the right side of the target range and its reflective surface is the left and rear sides of the vehicle, the cluster shape of the cluster of radar detection points caused by it is L-shaped. The L-shape mentioned here is only a rough shape estimate. In real scenes, the line segments connected by the radar detection points in the cluster are not necessarily strictly straight lines, and the connection between the line segments formed by the radar detection points of the two reflective surfaces is not necessarily a right angle, but may be other angles. These approximate L-shaped shapes can all be considered as L-shaped in this application.
[0094] For example, when the vehicle to be sensed is on the left side of the target range and its reflective surface is the right side and rear side of the vehicle, the cluster shape of the cluster of radar detection points caused by it is an inverted L-shape. The inverted L-shape mentioned here is only a rough shape estimate. In real scenes, the line segments connecting the radar detection points in the cluster are not necessarily strictly straight lines, and the connection between the line segments formed by the radar detection points of the two reflective surfaces is not necessarily a right angle, but may be other angles. These approximate inverted L-shapes can all be considered as inverted L-shapes in this application.
[0095] For example, when the vehicle to be sensed is in the middle of the target range and its reflecting surface is the rear side of the vehicle, the cluster shape of the radar detection points caused by it is a horizontal line shape. The horizontal line shape mentioned here is only a rough shape estimate. In the real scene, the line segments connecting the radar detection points in the cluster are not necessarily strictly straight lines. Shapes that are approximate to horizontal lines can also be considered as horizontal lines in this application.
[0096] Optionally, according to the clustering result and the cluster size and shape information corresponding to the target to be sensed in the target scene, the radar detection point caused by the clutter in the target range is identified, including:
[0097] The radar detection points in the target cluster that do not match multiple cluster size and shape information units in the clustering results are identified as radar detection points caused by clutter.
[0098] Specifically, after obtaining the clustering results of radar detection points, multiple clusters in the clustering results can be identified based on the cluster size and shape information corresponding to the target to be perceived in the target scene, and the target cluster that does not match multiple cluster size and shape information units can be identified as a radar detection point caused by clutter.
[0099] In one implementation, when the cluster shape of the target cluster does not match the multiple cluster size and shape information units, the target cluster may be identified as a radar detection point caused by clutter.
[0100] In one implementation, when the cluster size of the target cluster does not match the multiple cluster size and shape information units, the target cluster may be identified as a radar detection point caused by clutter.
[0101] Optionally, the method further comprises:
[0102] According to the perception performance index and / or the communication performance index in the target scene, the algorithm parameters used by the density clustering algorithm are determined.
[0103] Specifically, the perception performance index in the target scene can be used to characterize the accuracy of identifying the target to be perceived in the target scene, such as the ratio between the number of correctly identified targets to be perceived and the number of actual targets to be perceived in the target scene. The communication performance index in the target scene can be used to characterize the efficiency of identifying clutter in the target scene, such as the time required to complete the clutter identification process.
[0104] In one implementation, in a target scenario with high real-time requirements, before clutter identification, multiple groups of algorithm parameters may be tested first, and then at least one group of algorithm parameters that meets the requirements of perception performance indicators and / or communication performance indicators may be selected for clutter identification.
[0105] In one implementation, in a target scenario with low real-time requirements, all possible algorithm parameters may be traversed, and at least one set of algorithm parameters with the highest perception performance index and / or communication performance index may be selected for clutter identification.
[0106] Optionally, the method further comprises:
[0107] According to the number of clusters corresponding to the targets to be perceived, the number of targets to be perceived in the target scene is determined.
[0108] Specifically, after identifying the clusters of radar detection points caused by clutter in the target range according to the cluster size and shape information corresponding to the target to be sensed in the target scene and the clustering results, the clusters of radar detection points caused by clutter can be eliminated, and the remaining clusters can be used as the clusters of radar detection points caused by the target to be sensed. The number of clusters corresponding to the target to be sensed can be used as the number of targets to be sensed in the target scene.
[0109] Optionally, the method further comprises:
[0110] The long axis direction of the first target to be sensed is determined according to the long side direction of the cluster corresponding to the first target to be sensed.
[0111] Specifically, the first target to be sensed may be any target to be sensed in the target scene. After determining the cluster corresponding to the target to be sensed, the long axis direction of the first target to be sensed may be determined according to the long side direction of the cluster corresponding to the first target to be sensed (for example, the long side direction when the cluster shape is L-shaped). For example, the angle between the long side of the cluster corresponding to the first target to be sensed and the central axis of the detection range may be used as the long axis direction of the first target to be sensed.
[0112] The methods provided in the above embodiments of the present application are illustrated below through examples of specific application scenarios.
[0113] The embodiment of the present application proposes a clutter identification method based on a density clustering algorithm, which utilizes the fact that radar detection points caused by clutter in a scene can be clustered into detection point clusters of a specific shape and size to distinguish between targets to be sensed and objects causing clutter.
[0114] The density clustering algorithm can be used to identify clutter within the detection range of the radar or synaesthesia system. The following is a brief description of the steps by simulating the radar detection point processing in a road scene.
[0115] Figure 3 A schematic diagram of an L-shaped intersection autonomous driving simulation scenario provided in an embodiment of the present application, such as Figure 3 As shown, there are two vehicles moving on the road, of which the lower vehicle is equipped with a synesthesia integration system for detecting the target to be sensed in front and communicating. The area above the broken line indicates the sensing range of the vehicle synesthesia integration system; the upper vehicle is a target to be sensed; guardrails are installed on both sides of the road, and the echo caused by the guardrails is regarded as clutter interference and needs to be identified. Figure 3 The middle dot is the radar detection point sensed by the vehicle radar or the integrated interaceptive system through the echo.
[0116] The parameters used in the simulation are shown in Table 1.
[0117] Table 1: Simulation system parameters and parameter values
[0118]
[0119] The local maximum value of the echo signal strength at each beam scanning angle is a radar detection point.
[0120] from Figure 3 It can be seen that the radar or interawareness system of the vehicle below can sense the vehicle above and the guardrails on both sides of the road through the echo. The radar detection point of the echo of the vehicle above needs to be sensed, and the echo caused by the guardrails on both sides of the road can be regarded as clutter. Before the radar or interawareness system further processes the perception data, the radar detection point caused by the clutter needs to be identified.
[0121] right Figure 3 The radar detection points in the image are clustered by density clustering algorithm, which can cluster the radar detection points generated by vehicles above the target to be sensed and the radar detection points generated by the reflection of the guardrails on both sides of the road into different clusters. Different epsilon and minpts parameter selections will generate different cluster clustering results.
[0122] Figure 4 This is one of the clustering result diagrams of the simulation scenario provided in the embodiment of the present application, such as Figure 4 As shown in the figure, the parameters epsilon = 5 and minpts = 2 are used for clustering, and the radar detection points in different rectangular boxes are clustered into different clusters. The radar detection points of the guardrail on the left side of the road are clustered into one cluster; the radar detection points of the guardrail on the right side of the road are clustered into one cluster; the radar detection points generated by the reflection of the vehicle above are clustered into one cluster. Clutter and the target to be sensed can be easily distinguished, which is conducive to the subsequent processing of the radar detection points by the synaesthesia integrated system. However, the radius epsilon of this set of parameters epsilon = 5 and minpts = 2 is large. If the target to be sensed is close to the guardrail, the radar detection points generated by the reflection of the target to be sensed may be clustered into one cluster with the clutter detection points caused by the guardrail.
[0123] Figure 5 The second schematic diagram of the clustering result of the simulation scenario provided in the embodiment of the present application is as follows: Figure 5As shown in the figure, the parameters epsilon = 4, minpts = 2 and epsilon = 4, minpts = 3 are used for clustering, and the radar detection points in different rectangular boxes are clustered into different clusters. The clustering results of the two sets of parameters epsilon = 4, minpts = 2 and epsilon = 4, minpts = 3 are exactly the same. The radar detection points of the guardrail on the left side of the road are clustered into one cluster; the radar detection points of the guardrail on the right side of the road are clustered into two clusters; the radar detection points generated by the reflection of the upper vehicle are clustered into one cluster. Clutter and the target to be sensed can be easily distinguished, which is conducive to the subsequent processing of the radar detection points by the synaesthesia integrated system. The radar detection points of the guardrail on the right side of the road are clustered into two clusters because the distance between point A and point B is large. Minpts of 2 and 3 have no effect on the clustering results because the radar detection points generated by the reflection of the guardrail are basically distributed in a straight line. The number of detection points in each radius is mainly determined by the radius epsilon, and minpts has little effect on the clustering results. This set of parameters can still cluster the clutter detection points well and has a smaller epsilon.
[0124] Figure 6 The third schematic diagram of the clustering results of the simulation scenario provided in the embodiment of the present application is as follows: Figure 6 As shown in the figure, the parameters epsilon = 3 and minpts = 2 are used for clustering, and the radar detection points in different rectangular boxes are clustered into different clusters. The radar detection points of the guardrail on the left side of the road are clustered into two clusters; the radar detection points of the guardrail on the right side of the road are clustered into four clusters; the radar detection points generated by the reflection of the vehicle above are clustered into one cluster. Since the clutter detection points are clustered into too many clusters, it is not conducive to the clutter recognition of the synaesthesia integration system and the subsequent processing of the radar detection points. For example, the cluster generated by the clutter may be mistakenly identified as the target to be sensed. Therefore, the set of parameters epsilon = 3, minpts = 2 is not suitable for this simulation scenario.
[0125] The steps to use the density clustering algorithm to identify and distinguish the clutter caused by the vehicles to be sensed and the roadside guardrails in the highway scene are as follows:
[0126] 1. Determine the perception scene, as well as the possible targets to be perceived and objects that cause clutter interference in the scene. In the highway scene, the targets to be perceived are vehicles on the road. The echoes reflected by the buildings and guardrails on both sides of the road are clutter interference.
[0127] The clustering results in the radar map can be estimated based on the possible shapes and sizes of the vehicle targets and objects causing clutter interference in the scene. The top view of the vehicle target is usually rectangular, so the clustering results caused by the reflection points of the vehicle to be sensed can be divided into three categories: (1) inverted L-shaped; (2) L-shaped; (3) horizontal line shape.
[0128] Figure 7 This is one of the schematic diagrams of clustering results caused by the reflection points of the vehicles to be sensed provided in the embodiment of the present application, such as Figure 7 As shown in FIG, when the direction of travel of the to-be-sensed vehicle is close to that of the sensing vehicle, and the right side of the to-be-sensed vehicle is within the sensing field of view (FOV) of the sensing vehicle, the clustering result caused by the reflection point of the to-be-sensed vehicle presents an inverted L shape, as shown in FIG. Figure 7 As shown in the figure, the vehicle to be sensed on the left in the middle; when the vehicle to be sensed is close to the driving direction of the sensing vehicle, and the left side of the vehicle to be sensed is within the FOV range of the sensing vehicle, the clustering result caused by the reflection point of the vehicle to be sensed is L-shaped, as shown in Figure 7 The vehicle to be sensed is shown on the middle right.
[0129] Figure 8 The second schematic diagram of the clustering result caused by the reflection points of the vehicle to be sensed provided in the embodiment of the present application is as follows: Figure 8 As shown in , when the direction of travel of the vehicle to be sensed is close to that of the sensing vehicle, and the left side and right side of the vehicle to be sensed are not within the FOV range of the sensing vehicle, the clustering result caused by the reflection points of the vehicle to be sensed presents a horizontal line shape, as shown in Figure 8 The vehicle to be sensed is shown in the figure.
[0130] The reflection points of buildings and guardrails in this scene usually do not form reverse L-shaped, L-shaped and horizontal line-shaped clustering results.
[0131] Each time a measurement is made (for example, the intersensory integrated base station transmits a sensing signal in a certain direction), the target to be sensed, the object causing clutter interference, and the noise will form multiple local maximum values of echo signal strength in the echo signal. By setting the echo signal strength threshold or the number of local maximum values of echo signal strength, the local maximum values of echo signal strength caused by some noise can be filtered out to reduce the radar detection points caused by noise.
[0132] The remaining local maximum values of echo signal strength that may be generated by the target to be sensed and the objects that cause clutter interference are used for ranging and angle measurement. For example, the Fast Fourier Transform (FFT) periodogram method, the Multiple Signal Classification (MUSIC) algorithm, the Estimation of Signal Parameters via Rotational Invariance Techniques (ESPRIT) algorithm, etc. can be used. The coordinates of each radar detection point can be obtained through ranging and angle measurement.
[0133] Apply a density clustering algorithm to the set of radar detection points. Compare the clusters generated by the clustering algorithm with the possible shapes and sizes of the targets to be sensed and the objects causing clutter interference, estimate which clusters are caused by clutter, and remove the clutter data in subsequent data processing.
[0134] For application scenarios with high real-time requirements, before applying the clutter interference identification method to a certain perception scenario, appropriate density clustering algorithm parameters epsilon and minpts can be obtained through testing.
[0135] For application scenarios where real-time requirements are not high, possible combinations of parameters epsilon and minpts can be traversed to eliminate clusters that may be clutter and obtain the perception performance index (or communication performance index in the perception-assisted communication scenario). The density clustering algorithm parameters epsilon and minpts with higher perception performance index (or communication performance index) are selected.
[0136] When the size of the target to be sensed is larger than the distance resolution of the synaesthesia integrated system, the target to be sensed will usually reflect the sensing signal at multiple locations, which is shown as multiple radar detection points in the radar map. Radar detection points belonging to the same cluster after clustering by the density clustering algorithm can be considered to belong to the same target to be sensed. Therefore, the present invention can estimate the number of targets to be sensed in a multi-target scenario.
[0137] According to the L-shaped or reverse L-shaped clustering results, the long axis direction of the vehicle to be sensed can be roughly estimated according to the direction of the long side of the L-shaped or reverse L-shaped. For example, the long axis direction of the vehicle to be sensed can be estimated according to the angle between the long side of the L-shaped clustering result of the vehicle to be sensed and the central axis of the FOV of the sensed vehicle.
[0138] The clutter interference identification method based on the density clustering algorithm provided in the embodiment of the present application can be applied to mobile scenes and can identify clutter interference caused by complex objects.
[0139] The methods and devices provided in the various embodiments of the present application are based on the same application concept. Since the methods and devices solve problems based on similar principles, the implementation of the devices and methods can refer to each other, and the repeated parts will not be repeated.
[0140] Fig. 9 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application, such as Fig. 9 As shown, the electronic device includes a memory 920, a transceiver 910 and a processor 900; wherein the processor 900 and the memory 920 may also be arranged physically separately.
[0141] The memory 920 is used to store computer programs; the transceiver 910 is used to send and receive data under the control of the processor 900.
[0142] Specifically, the transceiver 910 is used to receive and send data under the control of the processor 900 .
[0143] Among them, Fig. 9 In the embodiment, the bus architecture may include any number of interconnected buses and bridges, specifically linking together various circuits of one or more processors represented by processor 900 and memory represented by memory 920. The bus architecture may also link together various other circuits such as peripherals, voltage regulators, and power management circuits, which are well known in the art and are therefore not further described in this application. The bus interface provides an interface. The transceiver 910 may be a plurality of components, namely, a transmitter and a receiver, providing a unit for communicating with various other devices on a transmission medium, which may include a wireless channel, a wired channel, an optical cable, and other transmission media.
[0144] The processor 900 is responsible for managing the bus architecture and general processing, and the memory 920 can store data used by the processor 900 when performing operations.
[0145] The processor 900 may be a central processing unit (CPU), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or a complex programmable logic device (CPLD). The processor may also adopt a multi-core architecture.
[0146] The processor 900 is configured to execute any of the methods provided in the embodiments of the present application according to the obtained executable instructions by calling the computer program stored in the memory 920, for example: obtaining multiple radar detection points obtained by detecting the target range in the target scene;
[0147] Based on the density clustering algorithm, multiple radar detection points are clustered;
[0148] According to the clustering results and the cluster size and shape information corresponding to the target to be perceived in the target scene, the radar detection points caused by clutter in the target range are identified.
[0149] Optionally, the cluster size and shape information corresponding to the target to be sensed includes multiple cluster size and shape information units, and each cluster size and shape information unit includes information of a cluster size and information of at least one cluster shape.
[0150] Optionally, the cluster shape includes at least one of an L-shape, an inverted L-shape and a horizontal line shape.
[0151] Optionally, according to the clustering result and the cluster size and shape information corresponding to the target to be sensed in the target scene, the radar detection point caused by the clutter in the target range is identified, including:
[0152] The radar detection points in the target cluster that do not match multiple cluster size and shape information units in the clustering results are identified as radar detection points caused by clutter.
[0153] Optionally, the operation further includes:
[0154] According to the perception performance index and / or the communication performance index in the target scene, the algorithm parameters used by the density clustering algorithm are determined.
[0155] Optionally, the density clustering algorithm includes a density-based clustering algorithm with noise DBSCAN.
[0156] Optionally, the operation further includes:
[0157] According to the number of clusters corresponding to the targets to be perceived, the number of targets to be perceived in the target scene is determined.
[0158] Optionally, the operation further includes:
[0159] The long axis direction of the first target to be sensed is determined according to the long side direction of the cluster corresponding to the first target to be sensed.
[0160] It should be noted here that the above-mentioned electronic device provided in the embodiment of the present application can implement all the method steps implemented in the above-mentioned method embodiment, and can achieve the same technical effect. The parts and beneficial effects of this embodiment that are the same as the method embodiment will not be described in detail here.
[0161] Fig.10 A schematic diagram of the structure of the clutter identification device provided in the embodiment of the present application is shown in FIG. Fig.10 As shown, the device comprises:
[0162] An acquisition unit 1000 is used to acquire a plurality of radar detection points obtained by detecting a target range in a target scene;
[0163] A clustering unit 1010 is used to cluster multiple radar detection points based on a density clustering algorithm;
[0164] The identification unit 1020 is used to identify the radar detection points caused by clutter within the target range according to the clustering results and the cluster size and shape information corresponding to the target to be sensed in the target scene.
[0165] Optionally, the cluster size and shape information corresponding to the target to be sensed includes multiple cluster size and shape information units, and each cluster size and shape information unit includes information of a cluster size and information of at least one cluster shape.
[0166] Optionally, the cluster shape includes at least one of an L-shape, an inverted L-shape and a horizontal line shape.
[0167] Optionally, according to the clustering result and the cluster size and shape information corresponding to the target to be sensed in the target scene, the radar detection point caused by the clutter in the target range is identified, including:
[0168] The radar detection points in the target cluster that do not match multiple cluster size and shape information units in the clustering results are identified as radar detection points caused by clutter.
[0169] Optionally, the device further comprises:
[0170] The first determination unit is used to determine the algorithm parameters used by the density clustering algorithm according to the perception performance index and / or the communication performance index in the target scene.
[0171] Optionally, the density clustering algorithm includes a density-based clustering algorithm with noise DBSCAN.
[0172] Optionally, the device further comprises:
[0173] The second determining unit is used to determine the number of the to-be-perceived targets in the target scene according to the number of clusters corresponding to the to-be-perceived targets.
[0174] Optionally, the device further comprises:
[0175] The third determining unit is used to determine the long axis direction of the first target to be perceived according to the long side direction of the cluster corresponding to the first target to be perceived.
[0176] It should be noted that the division of units in the embodiments of the present application is schematic and is only a logical function division. There may be other division methods in actual implementation. In addition, each functional unit in each embodiment of the present application may be integrated into a processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.
[0177] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a processor-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) or a processor (processor) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), disk or optical disk and other media that can store program codes.
[0178] It should be noted here that the above-mentioned device provided in the embodiment of the present application can implement all the method steps implemented in the above-mentioned method embodiment, and can achieve the same technical effect. The parts and beneficial effects of this embodiment that are the same as those in the method embodiment will not be described in detail here.
[0179] On the other hand, an embodiment of the present application further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and the computer program is used to enable a computer to execute the clutter identification method provided by the above embodiments.
[0180] It should be noted here that the computer-readable storage medium provided in the embodiment of the present application can implement all the method steps implemented in the above-mentioned method embodiment and can achieve the same technical effect. The parts and beneficial effects of this embodiment that are the same as those in the method embodiment will not be described in detail here.
[0181] The computer-readable storage medium can be any available medium or data storage device that can be accessed by a computer, including but not limited to magnetic storage (such as floppy disks, hard disks, magnetic tapes, magneto-optical disks (MO), etc.), optical storage (such as CD, DVD, BD, HVD, etc.), and semiconductor storage (such as ROM, EPROM, EEPROM, non-volatile memory (NANDFLASH), solid-state drive (SSD)), etc.
[0182] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage and optical storage, etc.) that contain computer-usable program code.
[0183] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer executable instructions. These computer executable instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.
[0184] These processor executable instructions may also be stored in a processor readable memory that can direct a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the processor readable memory produce an article of manufacture including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.
[0185] These processor-executable instructions may also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for executing on the computer or other programmable device to implement the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.
[0186] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is also intended to include these modifications and variations.
Claims
1. A clutter identification method, characterized in that: include: Acquire multiple radar detection points obtained by detecting the target range in the target scene; Clustering the multiple radar detection points based on a density clustering algorithm; According to the clustering result and the cluster size and shape information corresponding to the target to be sensed in the target scene, the radar detection points caused by clutter in the target range are identified.
2. The clutter identification method according to claim 1, characterized in that: The cluster size and shape information corresponding to the target to be sensed includes a plurality of cluster size and shape information units, each of which includes information about a cluster size and information about at least one cluster shape.
3. The clutter identification method according to claim 2, characterized in that: The cluster shape includes at least one of an L-shape, an inverted L-shape and a horizontal line shape.
4. The clutter identification method according to claim 2 or 3, characterized in that: The identifying of radar detection points caused by clutter within the target range according to the clustering result and the cluster size and shape information corresponding to the target to be sensed in the target scene comprises: The radar detection points in the target cluster that do not match the multiple cluster size and shape information units in the clustering result are identified as the radar detection points caused by clutter.
5. The clutter identification method according to claim 1, characterized in that: The method further comprises: The algorithm parameters used by the density clustering algorithm are determined according to the perception performance index and / or the communication performance index in the target scene.
6. The clutter identification method according to claim 1 or 5, characterized in that: The density clustering algorithm includes a density-based clustering algorithm with noise, DBSCAN.
7. The clutter identification method according to claim 1, characterized in that: The method further comprises: The number of the targets to be sensed in the target scene is determined according to the number of clusters corresponding to the targets to be sensed.
8. The clutter identification method according to claim 1, characterized in that: The method further comprises: The long axis direction of the first target to be sensed is determined according to the long side direction of the cluster corresponding to the first target to be sensed.
9. An electronic device, characterized in that: Including memory, transceiver, processor; A memory for storing a computer program; a transceiver for transmitting and receiving data under the control of the processor; and a processor for reading the computer program in the memory and performing the following operations: Acquire multiple radar detection points obtained by detecting the target range in the target scene; Clustering the multiple radar detection points based on a density clustering algorithm; According to the clustering result and the cluster size and shape information corresponding to the target to be sensed in the target scene, the radar detection points caused by clutter in the target range are identified.
10. The electronic device according to claim 9, characterized in that: The cluster size and shape information corresponding to the target to be sensed includes a plurality of cluster size and shape information units, each of which includes information about a cluster size and information about at least one cluster shape.
11. The electronic device according to claim 10, characterized in that: The cluster shape includes at least one of an L-shape, an inverted L-shape and a horizontal line shape.
12. The electronic device according to claim 10 or 11, characterized in that: The identifying of radar detection points caused by clutter within the target range according to the clustering result and the cluster size and shape information corresponding to the target to be sensed in the target scene comprises: The radar detection points in the target cluster that do not match the multiple cluster size and shape information units in the clustering result are identified as the radar detection points caused by clutter.
13. The electronic device according to claim 9, characterized in that: The operations also include: The algorithm parameters used by the density clustering algorithm are determined according to the perception performance index and / or the communication performance index in the target scene.
14. The electronic device according to claim 9 or 13, characterized in that: The density clustering algorithm includes a density-based clustering algorithm with noise, DBSCAN.
15. The electronic device according to claim 9, characterized in that: The operations also include: The number of the targets to be sensed in the target scene is determined according to the number of clusters corresponding to the targets to be sensed.
16. The electronic device according to claim 9, characterized in that: The operations also include: The long axis direction of the first target to be sensed is determined according to the long side direction of the cluster corresponding to the first target to be sensed.
17. A clutter identification device, characterized in that: include: An acquisition unit, used to acquire multiple radar detection points obtained by detecting a target range in a target scene; A clustering unit, used for clustering the plurality of radar detection points based on a density clustering algorithm; The identification unit is used to identify the radar detection points caused by clutter within the target range according to the clustering results and the cluster size and shape information corresponding to the target to be sensed in the target scene.
18. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and the computer program is used to enable a computer to execute the method according to any one of claims 1 to 8.
