Radar point data automatic labeling method, system and storage medium

Abstract

The present invention discloses a radar point trace data automatic labeling method, system and storage medium, including the steps of: acquiring radar point trace data; performing feature sensitivity analysis on the radar point trace data, establishing a point trace classification sensitive feature space, and screening out a number of sensitive features with separability above a preset value; using the sensitive features as the input of the CFDP algorithm, clustering the radar point trace data to obtain multiple clusters; merging clusters that meet the intra-cluster distance condition and the inter-cluster distance condition to obtain a merged result, and judging whether the merged result meets the overall DBI condition; when the merged result meets the overall DBI condition, outputting the merged result as the final class. The improved CFDP algorithm divides the point trace data, covers the original categories of the radar point trace data well, can meet the requirements for sample accuracy in the automatic labeling of radar point trace data, and can stably and accurately perform automatic labeling of radar point trace data.

Description

Technical Field

[0001] The present invention relates to the technical field of radar point trace data annotation, and in particular to a radar point trace data automatic annotation method, system and storage medium. Background Art

[0002] In radar data processing, as the data basis for point trace quality assessment, the objectivity of point trace data samples in describing different types of points directly determines the accuracy of point trace quality assessment. The existing point trace data annotation methods are basically based on manual work. By comparing radar track data with transponder data such as automatic dependent surveillance-broadcast, secondary radar or identification friend or foe, radar tracks that are close to the true values ​​of the aforementioned transponder in both space and time dimensions are selected. Those that meet the conditions are considered to be true tracks and are manually annotated. This method can obtain more accurate annotated data, but it requires a high level of experience for the annotators. In radar point trace quality assessment, whether using likelihood estimation or machine learning methods, it is necessary to use statistical methods to extract prior information of historical data. The more effective historical data, the more accurate the classification model. However, this is undoubtedly extremely unfriendly to manual annotation. The accuracy of manual comparison will significantly decrease with the increase of data volume, and the time cost will increase. Therefore, a stable and accurate radar point trace data automatic annotation method is needed for the annotation of massive radar point trace data. Summary of the invention

[0003] The purpose of the present invention is to overcome the above technical deficiencies, and to propose a method, system and storage medium for automatic labeling of radar point trace data, so as to solve the technical problem of difficulty in labeling point trace data in radar point trace quality assessment in the prior art.

[0004] In order to achieve the above technical objectives, in a first aspect, the technical solution of the present invention provides a method for automatically labeling radar point trace data, comprising the following steps:

[0005] Get radar trace data;

[0006] Performing feature sensitivity analysis on the radar point trace data, establishing a point trace classification sensitive feature space, and screening out a number of sensitive features with separability above a preset value;

[0007] The sensitive features are used as input of a CFDP (clustering by fast search and find of density peaks) algorithm to cluster the radar point trace data to obtain a plurality of clusters;

[0008] Merging clusters that meet the intra-cluster distance condition and the inter-cluster distance condition to obtain a merged result, and determining whether the merged result meets the overall DBI (Davies-Bouldin) condition;

[0009] When the merged result meets the overall DBI condition, the merged result is output as the final class.

[0010] Compared with the prior art, the radar point data automatic labeling method provided by the present invention has the following beneficial effects:

[0011] The automatic labeling method of radar point trace data comprises the following steps: first, acquiring radar point trace data; then, performing feature sensitivity analysis on the radar point trace data, establishing a point trace classification sensitive feature space, and screening out a number of sensitive features with separability above a preset value; secondly, taking the sensitive features as the input of a CFDP algorithm, clustering the radar point trace data to obtain a plurality of clusters; thirdly, merging clusters that meet intra-cluster distance conditions and inter-cluster distance conditions to obtain a merged result, and judging whether the merged result meets an overall DBI condition; finally, when the merged result meets the overall DBI condition, outputting the merged result as a final class.

[0012] The improved CFDP algorithm is used to divide the point data, which better covers the original categories of the radar point data, can meet the requirements of sample accuracy in the automatic labeling of radar point data, can provide an effective data basis for radar point quality evaluation, and can stably and accurately perform automatic labeling of radar point data, which has very good practical value.

[0013] According to some embodiments of the present invention, the performing feature sensitivity analysis on the radar point trace data comprises the steps of:

[0014] For each radar point trace feature of the radar point trace data, a scatter plot of true and false point traces and a frequency distribution statistical diagram corresponding to the radar point trace feature are used to perform qualitative and quantitative analysis on the contribution of the radar point trace feature to the classification of true and false radar point traces, and all the radar point trace features are ranked for their separability into true and false radar point traces.

[0015] According to some embodiments of the present invention, the using the sensitive feature as an input of the CFDP algorithm comprises the steps of:

[0016] The number of traces, azimuth spread, range spread, amplitude variance, distance to the nearest trace and trace duration are taken as inputs to the CFDP algorithm.

[0017] According to some embodiments of the present invention, clustering the radar point trace data to obtain a plurality of clusters comprises the steps of:

[0018] By iterating the “local density” and the “distance to the higher density point”, the radar point trace data is clustered in the high-dimensional space to obtain a plurality of clusters.

[0019] According to some embodiments of the present invention, the step of merging clusters that satisfy the intra-cluster distance condition and the inter-cluster distance condition to obtain a merged result includes the steps of:

[0020] When two clusters have similar average distances within the clusters, and the minimum distance between the two clusters is similar to the average distances within the two clusters, the two clusters are marked as clusters to be merged;

[0021] If the overall DBI value of the clusters to be merged decreases after merging, the clusters to be merged are merged to obtain the merged result.

[0022] According to some embodiments of the present invention, the two clusters are C i , C j Indicates that C i , C j The following conditions should be met for clusters with similar average distances:

[0023]

[0024] The average distance within a cluster is the average distance between all objects in the cluster and their nearest neighbor objects;

[0025]

[0026] neigh(x i ,2) is a nearest neighbor function, which means that the distance x is excluding itself i nearest object;

[0027] The minimum distance between the two clusters and the average distance within the two clusters should be close to each other if the following conditions are met:

[0028]

[0029] The distance between clusters is represented by the minimum distance.

[0030]

[0031] According to some embodiments of the present invention, clustering the radar point trace data to obtain a plurality of clusters comprises the steps of:

[0032] In the outlier selection stage, the outlier conditions are relaxed so that the CFDP algorithm tends to generate more classes.

[0033] In a second aspect, the present invention provides a radar point trace data automatic labeling system, comprising:

[0034] Feature sensitivity analysis module, used to perform feature sensitivity analysis on radar point trace data, establish sensitive feature space for point trace classification, and screen out several sensitive features with separability above a preset value;

[0035] A CFDP algorithm clustering processing module is communicatively connected with the feature sensitivity analysis module, and the CFDP algorithm clustering processing module is used to perform clustering processing on the radar point trace data to obtain multiple clusters;

[0036] A merging module, which is in communication connection with the CFDP algorithm clustering processing module, and is used to merge clusters that meet the intra-cluster distance condition and the inter-cluster distance condition to obtain a merging result;

[0037] A DBI condition judgment module, which is in communication with the merging module, and is used to judge whether the merging result meets the overall DBI condition;

[0038] An output module is communicatively connected to the DBI condition judgment module, and is used to output the merged result that meets the overall DBI condition as a final class.

[0039] In a third aspect, the present invention provides a radar point trace data automatic labeling system, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the radar point trace data automatic labeling method as described in any one of the first aspects is implemented.

[0040] In a fourth aspect, the present invention provides a computer-readable storage medium, characterized in that the computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are used to enable a computer to execute the radar point track data automatic labeling method as described in any one of the first aspects.

[0041] Additional aspects and advantages of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] The above and / or additional aspects and advantages of the present invention will become apparent and easily understood from the description of the embodiments in conjunction with the following drawings, wherein the abstract drawings are identical to one of the drawings in the specification:

[0043] Figure 1 A flowchart of a method for automatically labeling radar point trace data provided by an embodiment of the present invention;

[0044] Figure 2 A flowchart of a method for automatically labeling radar point trace data provided by another embodiment of the present invention. DETAILED DESCRIPTION

[0045] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0046] It should be noted that, although the functional modules are divided in the system schematic diagram and the logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than the module division in the system or the order in the flowchart. The terms "first", "second", etc. in the specification, claims and the above drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence.

[0047] The present invention provides a method for automatic labeling of radar point trace data. By dividing the point trace data using an improved CFDP algorithm, the original categories of the radar point trace data are better covered, the requirements for sample accuracy in the automatic labeling of radar point trace data can be met, an effective data basis can be provided for radar point trace quality evaluation, and the radar point trace data can be automatically labeled stably and accurately, which has very good practical value.

[0048] The embodiments of the present invention are further described below in conjunction with the accompanying drawings.

[0049] Reference Figure 1 , Figure 1 This is a flowchart of a radar point trace data automatic labeling method provided by an embodiment of the present invention. The radar point trace data automatic labeling method includes but is not limited to steps S110 to S150.

[0050] Step S110, obtaining radar point trace data;

[0051] Step S120, performing feature sensitivity analysis on the radar point trace data, establishing a point trace classification sensitive feature space, and screening out a number of sensitive features with separability above a preset value;

[0052] Step S130, using the sensitive features as the input of the CFDP algorithm, clustering the radar point data to obtain multiple clusters;

[0053] Step S140, merging clusters that meet the intra-cluster distance condition and the inter-cluster distance condition to obtain a merged result, and determining whether the merged result meets the overall DBI condition;

[0054] Step S150: When the merged result meets the overall DBI condition, the merged result is output as the final class.

[0055] In one embodiment, the automatic labeling method of radar point trace data includes the following steps: first, obtaining radar point trace data; then, performing feature sensitivity analysis on the radar point trace data, establishing a point trace classification sensitive feature space, and screening out several sensitive features with separability above a preset value; second, using the sensitive features as the input of the CFDP algorithm, clustering the radar point trace data to obtain multiple clusters; second, merging the clusters that meet the intra-cluster distance condition and the inter-cluster distance condition to obtain a merged result, and judging whether the merged result meets the overall DBI condition; finally, when the merged result meets the overall DBI condition, outputting the merged result as the final class. The improved CFDP algorithm divides the point trace data, covers the original categories of the radar point trace data well, can meet the requirements for sample accuracy in the automatic labeling of radar point trace data, can provide an effective data basis for radar point trace quality assessment, can stably and accurately perform automatic labeling of radar point trace data, and has very good practical value.

[0056] In one embodiment, the automatic labeling method of radar point trace data includes the following steps: first, obtaining radar point trace data; then, performing feature sensitivity analysis on the radar point trace data, establishing a point trace classification sensitive feature space, and screening out several sensitive features with separability above a preset value; second, using the sensitive features as the input of the CFDP algorithm, clustering the radar point trace data to obtain multiple clusters; second, merging the clusters that meet the intra-cluster distance condition and the inter-cluster distance condition to obtain a merged result, and judging whether the merged result meets the overall DBI condition; finally, when the merged result meets the overall DBI condition, outputting the merged result as the final class. Performing feature sensitivity analysis on radar point trace data includes the following steps: for each radar point trace feature of radar point trace data, using the true and false point trace scatter plot and frequency distribution statistics corresponding to the radar point trace feature, qualitatively and quantitatively analyzing the contribution of the radar point trace feature to the classification of true and false radar point traces, and ranking the separability of all radar point trace features for true and false radar point traces.

[0057] In one embodiment, the automatic labeling method of radar point trace data includes the following steps: first, obtaining radar point trace data; then, performing feature sensitivity analysis on the radar point trace data, establishing a point trace classification sensitive feature space, and screening out several sensitive features with separability above a preset value; second, using the sensitive features as the input of the CFDP algorithm, clustering the radar point trace data to obtain multiple clusters; second, merging the clusters that meet the intra-cluster distance condition and the inter-cluster distance condition to obtain a merged result, and judging whether the merged result meets the overall DBI condition; finally, when the merged result meets the overall DBI condition, outputting the merged result as the final class. Using the sensitive features as the input of the CFDP algorithm includes the following steps: using the number of point traces, azimuth extension, distance extension, amplitude variance, nearest point trace distance, and point trace duration as the input of the CFDP algorithm.

[0058] In one embodiment, the automatic labeling method of radar point trace data includes the following steps: first, obtaining radar point trace data; then, performing feature sensitivity analysis on the radar point trace data, establishing a point trace classification sensitive feature space, and screening out several sensitive features with separability above a preset value; second, using the sensitive features as the input of the CFDP algorithm, clustering the radar point trace data to obtain multiple clusters; second, merging the clusters that meet the intra-cluster distance condition and the inter-cluster distance condition to obtain a merged result, and judging whether the merged result meets the overall DBI condition; finally, when the merged result meets the overall DBI condition, outputting the merged result as the final class. Clustering the radar point trace data to obtain multiple clusters includes the following steps: clustering the radar point trace data in the high-dimensional space to obtain multiple clusters by iterating the "local density" and the "distance from higher density points".

[0059] refer to Figure 2 , Figure 2 This is a flowchart of a radar point trace data automatic labeling method provided by another embodiment of the present invention. The radar point trace data automatic labeling method includes but is not limited to steps S210 to S220.

[0060] Step S210, when two clusters have similar average distances within the clusters, and the minimum distances between the two clusters are similar to the average distances within the two clusters, the two clusters are marked as clusters to be merged;

[0061] Step S220: If the overall DBI value of the clusters to be merged is reduced after the merging, the clusters to be merged are merged to obtain a merging result.

[0062] In one embodiment, the automatic labeling method of radar point trace data includes the following steps: first, obtaining radar point trace data; then, performing feature sensitivity analysis on the radar point trace data, establishing a point trace classification sensitive feature space, and screening out several sensitive features with separability above a preset value; second, using the sensitive features as the input of the CFDP algorithm, clustering the radar point trace data to obtain multiple clusters; second, merging the clusters that meet the intra-cluster distance condition and the inter-cluster distance condition to obtain a merged result, and judging whether the merged result meets the overall DBI condition; finally, when the merged result meets the overall DBI condition, outputting the merged result as the final class. The clusters that meet the intra-cluster distance condition and the inter-cluster distance condition are merged to obtain a merged result, including the following steps: when two clusters have similar intra-cluster average distances, and the minimum distance of the two clusters is similar to the average distance within the two clusters, marking the two clusters as clusters to be merged; if the overall DBI value of the clusters to be merged decreases after the merger, merging the clusters to be merged to obtain a merged result.

[0063] In one embodiment, the automatic labeling method of radar point trace data includes the following steps: first, obtaining radar point trace data; then, performing feature sensitivity analysis on the radar point trace data, establishing a point trace classification sensitive feature space, and screening out several sensitive features with separability above a preset value; second, using the sensitive features as the input of the CFDP algorithm, clustering the radar point trace data to obtain multiple clusters; second, merging the clusters that meet the intra-cluster distance condition and the inter-cluster distance condition to obtain a merged result, and judging whether the merged result meets the overall DBI condition; finally, when the merged result meets the overall DBI condition, outputting the merged result as the final class. The clusters that meet the intra-cluster distance condition and the inter-cluster distance condition are merged to obtain a merged result, including the following steps: when two clusters have similar intra-cluster average distances, and the minimum distance of the two clusters is similar to the average distance within the two clusters, marking the two clusters as clusters to be merged; if the overall DBI value of the clusters to be merged decreases after the merger, merging the clusters to be merged to obtain a merged result.

[0064] For the clustering result C of CFDP, consider any two clusters in C. If the two clusters Ci and Cj have similar average distances within the clusters, and the minimum distance between the two clusters is close to the average distance within the two clusters, then Ci and Cj may be two relatively close different clusters, or two subclusters in a multi-peak cluster, which are called clusters to be merged. At this time, the DBI index can be used for judgment. If merging Ci and Cj can reduce the overall DBI value (the smaller the DBI value, the better the clustering result), then the two clusters are considered to be two subclusters in the multi-peak cluster and should be merged.

[0065] Two clusters with C i , C j Indicates that C i , C j The following conditions should be met for clusters with similar average distances:

[0066]

[0067] The average distance within a cluster is the average distance between all objects in the cluster and their nearest neighbor objects;

[0068]

[0069] neigh(x i ,2) is a nearest neighbor function, which means that the distance x is excluding itself i nearest object;

[0070] The minimum distance between the two clusters and the average distance within the two clusters should be close to each other if the following conditions are met:

[0071]

[0072] The distance between clusters is represented by the minimum distance.

[0073]

[0074] DBI is defined as follows:

[0075]

[0076]

[0077]

[0078]

[0079]

[0080] The pseudo code of the improved CFDP clustering algorithm is as follows:

[0081]

[0082] The input of the algorithm is an m×n dimensional data set X, and the output is a cluster structure C. The algorithm is divided into two stages: CFDP clustering and merging. First, the CFDP algorithm is used to obtain the cluster structure C of the clustering result for the data set X, and a merge list is initialized to store the class labels corresponding to the two clusters to be merged. For the CFDP result C, consider any two clusters C i , C j , if the distance between the two clusters and the average distance within the cluster are close enough to meet the merging condition, then the two clusters are considered to be clusters to be merged. Then try to merge the clusters to be merged. If the overall DBI is reduced after the merger, the two clusters can be merged. The corresponding class labels (p i 、p j ) into the merge list mergePairs. After all cluster pairs have been traversed, C is actually merged, and cluster pairs are taken out from the mergePairs list one by one, merged, and finally returned to C.

[0083] In one embodiment, the automatic labeling method of radar point trace data includes the following steps: first, obtaining radar point trace data; then, performing feature sensitivity analysis on the radar point trace data, establishing a point trace classification sensitive feature space, and screening out several sensitive features with separability above a preset value; second, using the sensitive features as input to the CFDP algorithm, clustering the radar point trace data to obtain multiple clusters; second, merging clusters that meet the intra-cluster distance condition and the inter-cluster distance condition to obtain a merged result, and judging whether the merged result meets the overall DBI condition; finally, when the merged result meets the overall DBI condition, outputting the merged result as the final class. Clustering the radar point trace data to obtain multiple clusters includes the following steps: in the stage of selecting outliers, relaxing the outlier conditions so that the CFDP algorithm tends to generate more classes.

[0084] The present invention also provides a radar point trace data automatic labeling system, comprising: a feature sensitivity analysis module, used for performing feature sensitivity analysis on radar point trace data, establishing a point trace classification sensitive feature space, and screening out a number of sensitive features with separability above a preset value; a CFDP algorithm clustering processing module, which is communicated with the feature sensitivity analysis module, and the CFDP algorithm clustering processing module is used for performing clustering processing on radar point trace data to obtain multiple clusters; a merging module, which is communicated with the CFDP algorithm clustering processing module, and the merging module is used for merging clusters that meet intra-cluster distance conditions and inter-cluster distance conditions to obtain a merged result; a DBI condition judgment module, which is communicated with the merging module, and the DBI condition judgment module is used for judging whether the merged result meets the overall DBI condition; an output module, which is communicated with the DBI condition judgment module, and the output module is used for outputting the merged result that meets the overall DBI condition as a final class.

[0085] The present invention also provides a radar point trace data automatic labeling system, comprising: a memory, a processor and a computer program stored in the memory and executable on the processor, and the processor implements the radar point trace data automatic labeling method as described above when executing the computer program.

[0086] The processor and the memory may be connected via a bus or other means.

[0087] The memory, as a non-transient computer-readable storage medium, can be used to store non-transient software programs and non-transient computer executable programs. In addition, the memory may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some embodiments, the memory may optionally include a memory remotely disposed relative to the processor, and these remote memories may be connected to the processor via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0088] It should be noted that the radar point trace data automatic labeling system in this embodiment may include a business processing module, an edge database, a server version information register, and a data synchronization module. When the processor executes the computer program, it implements the radar point trace data automatic labeling method as applied in the radar point trace data automatic labeling system as described above.

[0089] The device embodiments described above are merely illustrative, and the units described as separate components may or may not be physically separated, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0090] In addition, an embodiment of the present invention further provides a computer-readable storage medium, which stores computer-executable instructions, and the computer-executable instructions are executed by a processor or controller, for example, by a processor in the above-mentioned terminal embodiment, so that the above-mentioned processor can execute the radar point track data automatic labeling method in the above-mentioned embodiment.

[0091] It will be appreciated by those skilled in the art that all or some of the steps and systems in the methods disclosed above may be implemented as software, firmware, hardware, and appropriate combinations thereof. Some or all physical components may be implemented as software executed by a processor, such as a central processing unit, a digital signal processor, or a microprocessor, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit. Such software may be distributed on a computer-readable medium, which may include a computer storage medium (or non-transitory medium) and a communication medium (or transient medium). As known to those skilled in the art, the term computer storage medium includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disk (DVD) or other optical disk storage, magnetic cassettes, magnetic tapes, disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and can be accessed by a computer. Furthermore, it is well known to those skilled in the art that communication media typically embodies computer readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transport mechanism, and may include any information delivery media.

[0092] The above is a specific description of the preferred implementation of the present invention, but the present invention is not limited to the above-mentioned implementation mode. Technical personnel familiar with the field can also make various equivalent deformations or substitutions without violating the spirit of the present invention. These equivalent deformations or substitutions are all included in the scope defined by the claims of the present invention.

[0093] The specific implementation of the present invention described above does not constitute a limitation on the protection scope of the present invention. Any other corresponding changes and modifications made based on the technical concept of the present invention should be included in the protection scope of the claims of the present invention.

Claims

1. A radar point data automatic labeling method, characterized in that: The following steps are involved: Get radar trace data; Performing feature sensitivity analysis on the radar point trace data, establishing a point trace classification sensitive feature space, and screening out a number of sensitive features with separability above a preset value; The sensitive features are used as input of the CFDP algorithm, and the radar point trace data are clustered to obtain multiple clusters; Merging clusters that meet the intra-cluster distance condition and the inter-cluster distance condition to obtain a merged result, and determining whether the merged result meets the overall DBI condition; When the merged result meets the overall DBI condition, the merged result is output as the final class; The feature sensitivity analysis of the radar point trace data comprises the steps of: For each radar trace feature of the radar trace data, using a true and false trace scatter plot and a frequency distribution statistical graph corresponding to the radar trace feature, a qualitative and quantitative analysis is performed on the contribution of the radar trace feature to the classification of true and false radar traces, and all the radar trace features are ranked in terms of their separability for true and false radar traces; The step of merging clusters that satisfy the intra-cluster distance condition and the inter-cluster distance condition to obtain a merging result comprises the following steps: When two clusters have similar average distances within the clusters, and the minimum distance between the two clusters is similar to the average distances within the two clusters, the two clusters are marked as clusters to be merged; If the overall DBI value of the clusters to be merged decreases after merging, merging the clusters to be merged to obtain the merged result; The two clusters are represented by C i , C j Indicates that C i , C j The following conditions should be met for clusters with similar average distances: The average distance within a cluster is the average distance between all objects in the cluster and their nearest neighbor objects; neigh(x j ,2) is a nearest neighbor function, which means that the distance x is excluding itself j nearest object; The minimum distance between the two clusters and the average distance within the two clusters should be close to each other if the following conditions are met: The distance between clusters is expressed as the minimum distance 。 2. The method for automatically labeling radar point trace data according to claim 1, characterized in that: The method of using the sensitive feature as the input of the CFDP algorithm comprises the steps of: The number of traces, azimuth spread, range spread, amplitude variance, distance to the nearest trace and trace duration are taken as inputs to the CFDP algorithm.

3. The method for automatically labeling radar point trace data according to claim 1, characterized in that: The clustering process of the radar point trace data to obtain a plurality of clusters comprises the steps of: By iterating the "local density" and "distance to higher density points", clustering processing is performed in the high-dimensional space of radar point trace data to obtain multiple clusters.

4. The method for automatically labeling radar point trace data according to claim 1, characterized in that: The clustering process of the radar point trace data to obtain a plurality of clusters comprises the steps of: In the outlier selection stage, the outlier conditions are relaxed so that the CFDP algorithm tends to generate more classes.

5. A radar point trace data automatic labeling system, used to execute the radar point trace data automatic labeling method according to any one of claims 1 to 4, characterized in that: include: Feature sensitivity analysis module, used to perform feature sensitivity analysis on radar point trace data, establish sensitive feature space for point trace classification, and screen out several sensitive features with separability above a preset value; A CFDP algorithm clustering processing module is communicatively connected with the feature sensitivity analysis module, and the CFDP algorithm clustering processing module is used to perform clustering processing on the radar point trace data to obtain multiple clusters; A merging module, which is in communication connection with the CFDP algorithm clustering processing module, and is used to merge clusters that meet the intra-cluster distance condition and the inter-cluster distance condition to obtain a merging result; A DBI condition judgment module, which is in communication with the merging module, and is used to judge whether the merging result meets the overall DBI condition; An output module is communicatively connected to the DBI condition judgment module, and is used to output the merged result that meets the overall DBI condition as a final class.

6. A radar point data automatic labeling system, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the radar point trace data automatic labeling method as claimed in any one of claims 1 to 4 when executing the computer program.

7. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are used to enable a computer to execute the radar point trace data automatic labeling method as described in any one of claims 1 to 4.

Images