Multi-target recognition method, device, equipment and storage medium

Abstract

The present invention belongs to the field of aircraft tracking, and discloses a multi-target identification method, device, equipment and storage medium, the method comprising: obtaining a measurement data set of known aircraft and aircraft to be identified within a preset measurement period, filtering out measurement values ​​carrying preset message information from the measurement data set to further determine the historical motion trajectory of the known aircraft, constructing an elliptical threshold to perform a primary clustering on measurement values ​​other than the measurement values ​​carrying the preset message information; performing a secondary clustering on the remaining measurement values ​​after the primary clustering by using preset parameters; and identifying all aircraft within the preset measurement period based on the results of the two clusterings. The present invention performs a primary clustering on known aircraft among multiple aircraft measurement values ​​by using preset messages, and then performs a secondary clustering on the remaining aircraft to complete the identification of all aircraft, thereby solving the problem of identifying multiple aircraft as the same aircraft when the distance between the aircraft is relatively close, and achieving more accurate identification of multiple aircraft at relatively close distances.

Description

Technical Field

[0001] The present invention relates to the field of aircraft tracking technology, and in particular to a multi-target recognition method, device, equipment and storage medium. Background Art

[0002] In the process of multi-target tracking, data association is needed to start and maintain the track. Data association is the key to the overall performance of the multi-target tracking system. Accurate data association can greatly reduce the generation of false tracks and track the target status more accurately.

[0003] At present, the effect of clustering algorithm in expanding target clustering in multi-target tracking is not very ideal, because the current clustering algorithm cannot distinguish the targets when the flying targets are in formation flying at a close distance. The targets flying in formation will be identified as one target, resulting in poor clustering effect, which directly affects the estimation accuracy of the target number and motion state, thereby affecting the accuracy of multi-target tracking.

[0004] The above contents are only used to assist in understanding the technical solution of the present invention and do not constitute an admission that the above contents are prior art. Summary of the invention

[0005] The main purpose of the present invention is to provide a multi-target recognition method, device, equipment and storage medium, aiming to solve the technical problems of multi-target recognition in the prior art.

[0006] To achieve the above object, the present invention provides a multi-target recognition method, which comprises the following steps:

[0007] Acquire a measurement data set corresponding to a preset measurement period, wherein the measurement data set includes measurement data of a known aircraft and an aircraft to be identified;

[0008] Filtering out measurement values ​​carrying preset message information from the measurement data set;

[0009] Determine the historical motion trajectory of the known aircraft according to the measurement value carrying the preset message information;

[0010] Constructing an elliptical threshold according to the historical motion trajectory;

[0011] Clustering the other measurement values ​​except the measurement value carrying the preset message information once by using the elliptical threshold;

[0012] Perform secondary clustering on the remaining measurement values ​​after the first clustering by using preset parameters;

[0013] Based on the two clustering results, all aircraft within the preset measurement period are identified.

[0014] Optionally, the constructing an elliptical threshold according to the historical motion trajectory includes:

[0015] Obtaining known aircraft motion information according to the historical motion trajectory;

[0016] Obtaining an inverse matrix of a residual covariance matrix of the known aircraft according to the known aircraft motion information;

[0017] Get the preset threshold value;

[0018] An elliptical threshold is established according to the inverse matrix of the residual covariance matrix, the measurement value carrying the preset message information, and the preset threshold value.

[0019] Optionally, clustering the measurement values ​​other than the measurement value carrying the preset message information by using the elliptical threshold includes:

[0020] Determining whether other measurement values ​​except the measurement value carrying the preset message information are within the elliptical threshold;

[0021] Perform a clustering based on the judgment results.

[0022] Optionally, the preset parameters include: a cutoff distance, a cutoff function, and a maximum speed corresponding to the remaining measurement values ​​after a clustering;

[0023] The second clustering of the remaining measurement values ​​after the first clustering by using preset parameters includes:

[0024] Calculate the local density and density distance of the remaining measurement values ​​according to the cutoff distance, the cutoff function and the maximum speed corresponding to the remaining measurement values ​​after the first clustering;

[0025] The remaining measurement values ​​are clustered for the second time according to the local density and density distance of the remaining measurement values.

[0026] Optionally, performing secondary clustering on the remaining measurement values ​​according to the local density and density distance of the remaining measurement values ​​includes:

[0027] Comparing the local density of the first measurement value with the local density of the second measurement value, and comparing the density distance of the first measurement value with the density distance of the second measurement value, wherein the first measurement value and the second measurement value are any measurement value in the remaining measurement values;

[0028] Select the cluster center of the secondary clustering according to the comparison results;

[0029] The remaining measurement values ​​are clustered again according to the cluster centers.

[0030] Optionally, selecting a cluster center of secondary clustering according to the judgment result includes:

[0031] When the local density of the first measurement value is greater than the local density of the second measurement value, and the density distance of the first measurement value is greater than the density distance of the second measurement value, the first measurement value is used as the measurement center;

[0032] When the local density of the first measurement value is less than the local density of the second measurement value, and the density distance of the first measurement value is greater than the density distance of the second measurement value, the first measurement value is eliminated;

[0033] When the density distance of the first measurement value is smaller than the density distance of the second measurement value, the first measurement value is classified into a neighboring cluster, and the neighboring cluster is any one of the primary cluster and the secondary cluster.

[0034] Optionally, after all aircraft within a preset measurement period are identified based on the two clustering results, the method further includes:

[0035] Get the cluster centers of the two clusterings;

[0036] The positions of all aircrafts within the preset measurement period are determined based on the cluster center.

[0037] In addition, to achieve the above object, the present invention also provides a multi-target recognition device, the device comprising:

[0038] A measurement value acquisition module, used to acquire a measurement data set corresponding to a preset measurement cycle, wherein the measurement data set includes measurement data of known aircraft and aircraft to be identified;

[0039] The measurement value acquisition module is further used to filter out the measurement value carrying the preset message information from the measurement data set;

[0040] The measurement value acquisition module is further used to determine the historical motion trajectory of the known aircraft according to the measurement value carrying the preset message information;

[0041] The measurement value acquisition module is also used to construct an elliptical threshold according to the historical motion trajectory;

[0042] A target identification module, configured to perform a clustering operation on other measurement values ​​except the measurement value carrying the preset message information by using the elliptical threshold;

[0043] The target recognition module is also used to perform secondary clustering on the remaining measurement values ​​after the primary clustering by using preset parameters;

[0044] The target recognition module is also used to recognize all aircraft within a preset measurement period based on the two clustering results.

[0045] In addition, to achieve the above-mentioned purpose, the present invention also proposes a multi-target recognition device, which includes: a memory, a processor, and a multi-target recognition program stored in the memory and executable on the processor, wherein the multi-target recognition program is configured to implement the steps of the multi-target recognition method described above.

[0046] In addition, to achieve the above-mentioned purpose, the present invention also proposes a storage medium, on which a multi-target recognition program is stored, and when the multi-target recognition program is executed by a processor, the steps of the multi-target recognition method described above are implemented.

[0047] The present invention performs a primary clustering of known aircraft among multiple aircraft measurement values ​​through a preset message and then performs a secondary clustering of the remaining aircraft to complete the identification of all aircraft, thereby solving the problem that when multiple aircraft measurement values ​​are identified, the aircraft are relatively close in distance, and adjacent measurement values ​​are clustered as the same aircraft measurement values, thereby identifying multiple aircraft as the same aircraft. When multiple aircraft are relatively close in distance, the present invention achieves the ability to distinguish measurement values ​​belonging to different aircraft among multiple measurement values ​​and cluster them, thereby more accurately identifying multiple aircraft that are relatively close in distance. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 It is a schematic diagram of the structure of a multi-target recognition device in a hardware operating environment involved in an embodiment of the present invention;

[0049] Figure 2 It is a flowchart of the first embodiment of the multi-target recognition method of the present invention;

[0050] Figure 3 This is a multi-target clustering effect diagram of an embodiment of a multi-target recognition method of the present invention;

[0051] Figure 4 A comparison diagram of multi-target clustering effects of a multi-target recognition method according to an embodiment of the present invention;

[0052] Figure 5 It is a flowchart of a second embodiment of the multi-target recognition method of the present invention;

[0053] Figure 6 A comparison diagram of the number of false alarm targets in an embodiment of a multi-target recognition method of the present invention;

[0054] Figure 7 A comparison diagram of target recognition quantity of a multi-target recognition method according to an embodiment of the present invention;

[0055] Figure 8 It is a flowchart of a third embodiment of the multi-target recognition method of the present invention;

[0056] Fig. 9 This is a structural block diagram of the first embodiment of the multi-target recognition device of the present invention.

[0057] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0058] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0059] Reference Figure 1 , Figure 1 It is a schematic diagram of the structure of a multi-target recognition device in the hardware operating environment involved in the embodiment of the present invention.

[0060] like Figure 1 As shown, the multi-target recognition device may include: a processor 1001, such as a central processing unit (CPU), a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. Among them, the communication bus 1002 is used to realize the connection and communication between these components. The user interface 1003 may include a display screen (Display), an input unit such as a keyboard (Keyboard), and the optional user interface 1003 may also include a standard wired interface and a wireless interface. The network interface 1004 may optionally include a standard wired interface and a wireless interface (such as a wireless fidelity (Wireless-Fidelity, WI-FI) interface). The memory 1005 may be a high-speed random access memory (Random Access Memory, RAM), or a stable non-volatile memory (Non-Volatile Memory, NVM), such as a disk storage. The memory 1005 may also be a storage device independent of the aforementioned processor 1001.

[0061] Those skilled in the art will understand that Figure 1 The structure shown in the figure does not constitute a limitation on the multi-target recognition device, and may include more or less components than shown in the figure, or combine certain components, or arrange the components differently.

[0062] like Figure 1 As shown, the memory 1005 as a storage medium may include an operating system, a network communication module, a user interface module, and a multi-target recognition program.

[0063] exist Figure 1In the multi-target recognition device shown, the network interface 1004 is mainly used for data communication with the network server; the user interface 1003 is mainly used for data interaction with the user; the processor 1001 and the memory 1005 in the multi-target recognition device of the present invention can be set in the multi-target recognition device, and the multi-target recognition device calls the multi-target recognition program stored in the memory 1005 through the processor 1001, and executes the multi-target recognition method provided by the embodiment of the present invention.

[0064] Based on the above multi-target recognition device, the embodiment of the present invention provides a multi-target recognition method, referring to Figure 2 , Figure 2 Schematic diagram of the process of the first embodiment of the multi-target recognition method of the present invention.

[0065] In this embodiment, the multi-target recognition method includes the following steps:

[0066] Step S10: obtaining a measurement data set corresponding to a preset measurement cycle, wherein the measurement data set includes measurement data of known aircraft and aircraft to be identified.

[0067] It should be noted that the execution subject of this embodiment can be a computing service device with data processing, network communication and program running functions, such as a mobile phone, a tablet computer, a personal computer, etc., or an electronic device or a multi-target recognition device that can realize the above functions. The following takes the multi-target recognition device as an example to illustrate this embodiment and the following embodiments.

[0068] It should be understood that the preset measurement period refers to a measurement period for obtaining a measurement data set that is preset. For example, the measurement period can be set to 1s, and the measurement data of the aircraft within 1s can be obtained. The preset measurement period can be a measurement period in which the number of measurement values ​​in the measurement data set obtained after multiple experiments is relatively suitable. It can be adjusted according to actual needs, and there is no restriction on this in this embodiment. The measurement data is obtained by calculating the signal collected by the passive passive radar. Within the preset measurement period, multiple signals emitted by the same aircraft can be collected by the passive passive radar. At the same time, signals emitted by multiple aircraft can be collected, and then the position of the aircraft sending the signal is calculated based on the signal. The measurement value can be obtained based on the position. The measurement data set is also a data set composed of multiple measurement values.

[0069] It should be emphasized that the known aircraft refers to an aircraft that carries preset message information in the measurement value. If the message information carried by the measurement value cannot be identified, the aircraft corresponding to the measurement value is an aircraft to be identified.

[0070] Step S20: Filtering out measurement values ​​carrying preset message information from the measurement data set.

[0071] It should be noted that the preset message can be ADS_B message information, which is based on the information transmitted to the ADS-B system, where the ADS-B system is the abbreviation of the automatic dependent surveillance-broadcast system, which is composed of multiple ground stations and airborne stations, and completes two-way data communication in a meshed, multi-point to multi-point manner. The ADS-B system is an information system that integrates communication and monitoring, and consists of three parts: information source, information transmission channel, and information processing and display. According to the ADS_B information, the unique identification code of the aircraft can be obtained, so that the corresponding aircraft can be known.

[0072] The signals corresponding to all the measurement values ​​in the measurement data set are interpreted by a passive radar, and the ADS_B information can be obtained after the interpretation;

[0073] It should be understood that some of the measured values ​​may have corresponding ADS_B information, all of the measured values ​​may have corresponding ADS_B information, or none of the measured values ​​may have corresponding ADS_B information.

[0074] It is understandable that the main information of the ADS_B information is the 4-dimensional position information (longitude, latitude, altitude and time) of the aircraft and other possible additional information (conflict warning information, pilot input information, track angle and route turning point information) and aircraft identification information and category information. In addition, it may also include some other additional information, such as heading, airspeed, wind speed, wind direction and aircraft external temperature.

[0075] Step S30: Determine the historical motion trajectory of the known aircraft according to the measurement value carrying the preset message information.

[0076] It should be noted that the movement trajectory of the known flight information is determined by obtaining a unique identification code for the aircraft based on the ADS_B information corresponding to the measurement value carrying the preset message information. The historical track of the aircraft can be known based on the identification code, and the corresponding movement information of the aircraft can be further obtained based on the historical track, such as speed, acceleration, movement direction, aircraft's own weight, pressure and flow rate, etc.

[0077] It should be understood that the unique corresponding aircraft can be locked based on the identification code, and the movement information and flight route of the aircraft at the previous time can be known. The historical movement trajectory of the aircraft can be obtained based on the flight track of the aircraft corresponding to the identification code at the previous time.

[0078] Step S40: constructing an elliptical threshold according to the historical motion trajectory.

[0079] It should be noted that the elliptical threshold is a method for distinguishing possible situations in the motion of the target object during target tracking. The construction of the elliptical threshold is based on any target measurement value, and the distance between the measurement value other than the target measurement value and the target measurement value is used as a judgment condition, and the measurement value that meets the judgment condition is divided into an elliptical threshold.

[0080] It should be understood that the elliptical threshold is established based on the historical motion trajectory. The reference data such as the motion direction and motion speed of the target measurement value for establishing the elliptical threshold are obtained based on the historical motion trajectory. The direction, size, etc. of the elliptical threshold are determined based on the reference data, thereby establishing the elliptical threshold.

[0081] Step S50: clustering the measurement values ​​other than the measurement values ​​carrying the preset message information by using the elliptical threshold.

[0082] It should be noted that the measurement values ​​that do not carry preset message information in the first clustering may be divided into multiple elliptical thresholds. At this time, multiple clustering possibilities of the measurement values ​​are retained. The clustering refers to classifying the measurement values ​​that meet the same condition into the same category.

[0083] It should be understood that there may be multiple measurement values ​​carrying preset message information in the measurement set, and an elliptical threshold is established according to the preset message information of the measurement value, and the measurement values ​​around the measurement value that do not carry the preset message information are clustered using the elliptical threshold.

[0084] In a specific implementation, for ease of understanding, an example is given. It is assumed that there are two known measurement values ​​carrying preset message information, wherein an elliptical threshold a is established based on the historical motion trajectory of measurement value A, and the threshold value is 5; an elliptical threshold b is established based on the historical motion trajectory of another measurement value B, and the threshold value is 7. The threshold value refers to the range of the ellipse, and the measurement values ​​that do not carry the preset message information and whose distance from A is less than or equal to 5 are classified as elliptical threshold a, and the measurement values ​​that do not carry the preset message information and whose distance from B is less than or equal to 7 are classified as elliptical threshold b. When one of the measurement values ​​that does not carry the preset message information can be attributed to both the elliptical threshold a and the elliptical threshold b, the clustering of the measurement value that does not carry the preset message information in the elliptical threshold a and the elliptical threshold b is retained at the same time.

[0085] Step S60: performing secondary clustering on the remaining measurement values ​​after the primary clustering using preset parameters.

[0086] It should be noted that the secondary clustering is to cluster the remaining measurement values ​​by the local density and density distance of the remaining measurement values. The secondary clustering divides the remaining measurement values ​​into different cluster centers, which can be the primary cluster center or the secondary cluster center. Some measurement values ​​will also be regarded as clutter and eliminated during the clustering process.

[0087] It should be understood that after one clustering, some measurement values ​​may still be outside any ellipse threshold, but some measurement values ​​still need to be further clustered.

[0088] Step S70: Identify all aircraft within a preset measurement period based on the two clustering results.

[0089] It should be noted that after the two clusterings, the measurement values ​​in the measurement set can be divided into multiple different categories according to the clustering results. According to the measurement results, it can be identified how many aircraft the measurement values ​​are issued by, and the positions of all aircraft in the preset measurement period can be determined according to the clustering centers of the first clustering and the second clustering. The clustering effect of the present invention can be referred to Figure 3 And the FDPC clustering effect can be referred to Figure 4 , where the solid dots are the measurement values ​​detected within the measurement threshold, the solid circles are the results of the first clustering after the elliptical threshold is established based on the known aircraft, and the dotted circles are the results of the second clustering. By comparing and analyzing the clustering results, it can be seen that in the same measurement value set, the present invention can effectively identify and more accurately distinguish how many aircraft are issued by multiple measurement values ​​in the case of close distance.

[0090] It should be understood that the first clustering based on the elliptical threshold constructed according to the preset message signal can identify the corresponding aircraft, and the second clustering is based on the second clustering to know how many aircraft there are.

[0091] This embodiment performs a primary clustering of known aircraft among multiple aircraft measurement values ​​through a preset message and then performs a secondary clustering of the remaining aircraft to complete the identification of all aircraft, thereby solving the problem of clustering adjacent measurement values ​​as the same aircraft measurement values ​​when the aircraft are relatively close when identifying multiple aircraft measurement values, thereby identifying multiple aircraft as the same aircraft. This embodiment achieves the ability to distinguish measurement values ​​belonging to different aircraft from multiple measurement values ​​and cluster them when the multiple aircraft are relatively close, thereby more accurately identifying multiple aircraft that are relatively close.

[0092] refer to Figure 5 , Figure 5 Schematic diagram of the flow chart of the second embodiment of the multi-target recognition method of the present invention.

[0093] Based on the above first embodiment, in this embodiment, step S40 includes:

[0094] Step S41: Obtaining known aircraft motion information according to the historical motion trajectory.

[0095] It should be noted that the historical motion trajectory is the historical motion trajectory of a known aircraft. According to the historical trajectory, the motion information of the known aircraft can be known through data detection. The motion information includes (longitude, latitude, altitude and time) and other possible additional information (conflict warning information, pilot input information, track angle, route turning point, etc.) and aircraft identification information and category information. In addition, some other additional information may also be included, such as heading, airspeed, wind speed, wind direction and aircraft external temperature.

[0096] It should be understood that the motion information is not all carried in the preset signal, but can be obtained by synthesizing preset message information of the same known aircraft received multiple times.

[0097] In a specific implementation, it is currently identified that the preset message signal carries information about a known aircraft with a speed of W, an acceleration of a, an aircraft weight of G, and a corresponding aircraft identification number of Q. Then, by performing a data query based on Q, it can be obtained that the preset message signal sent by the aircraft Q in the previous flight carries a track angle P, a route turning point N, and a heading L. Then, the weight G, track angle P, route turning point N, etc. of the aircraft Q can be comprehensively known.

[0098] Step S42: Obtaining the inverse matrix of the residual covariance matrix of the known aircraft according to the known aircraft motion information.

[0099] It should be noted that the inverse matrix of the residual covariance matrix of the known aircraft is the inverse matrix of the residual covariance matrix of the known aircraft during target tracking in the historical track.

[0100] Step S43: Obtain a preset threshold value.

[0101] It should be noted that the preset threshold is the value of the elliptical range calculated based on the known motion information of the aircraft. In a specific implementation, the speed S and acceleration A of the known aircraft are used to calculate the approximate range of the aircraft based on the speed and acceleration. After multiple data experiments, a more appropriate range is obtained.

[0102] Step S44: Establishing an elliptical threshold according to the inverse matrix of the residual covariance matrix, the measured value carrying the preset message information and the preset threshold value

[0103] It should be noted that the ellipse threshold formula is:

[0104] [z i -z s ]'S -1[z i -z s ]≤γwhere zi is any measurement value. zs is the measurement value in the measurement set containing the ADS_B message. S -1 is the inverse matrix of the residual covariance matrix. γ is the preset threshold.

[0105] It should be understood that the measurement data set is clustered once by using the elliptical threshold with the measurement value containing the ADS_B message information as the cluster center. Because the measurement data with ADS_B interpretation information are accurate data of real targets, the measurement value is used to represent the cluster center after the first clustering as the subsequent association algorithm to accurately associate the data. In the experiment, there is a certain difference in the number of false alarms and the number of identified targets after the first clustering based on the measurement value containing the ADS_B message and the second clustering. The number of false alarms refers to the number of incorrect identifications. For details, please refer to Figure 6 as well as Figure 7 It can be clearly seen that the number of aircraft identified by the present invention is closer to the actual number of aircraft, and the number of false alarms for aircraft identified by the present invention is smaller than the number of false alarms for aircraft identified by the FDCP algorithm.

[0106] This embodiment further obtains the inverse matrix of the residual covariance matrix of the known aircraft and a preset threshold value based on the historical track of the known aircraft. According to the inverse matrix of the residual covariance matrix and the preset threshold value, an elliptical threshold is established with the measurement value containing preset message information. The measurement data set is clustered once through the elliptical threshold, which can accurately identify some known aircraft and distinguish multiple detected close-range aircraft, thereby avoiding the problem of mistaking multiple close-range aircraft for the same aircraft and failing to accurately identify them.

[0107] refer to Figure 8 , Figure 8 Schematic diagram of the flow chart of the third embodiment of the multi-target recognition method of the present invention.

[0108] Based on the above embodiments, in this embodiment, after step S60, the method further includes:

[0109] Step S61: Calculating the local density and density distance of the remaining measurement values ​​according to the cutoff distance, the cutoff function and the maximum speed corresponding to the remaining measurement values ​​after one clustering.

[0110] It should be noted that the local density calculation formula is:

[0111]

[0112] where d c is the cutoff distance, and d c =TV max, V max It is the maximum flight speed of the aircraft corresponding to the remaining measurement value after clustering. It can be calculated based on the current speed and acceleration of the measurement value. T is the measurement threshold of the passive radar. The calculation formula of the truncation function X(x) is:

[0113]

[0114] The density distance δ i It means that the local density is greater than z i And the distance from the nearest point, the z i It can be any measurement value among the remaining measurement values ​​after clustering. i It is expressed as:

[0115]

[0116] It should be emphasized that considering that the residual measurement value may contain noise and the truncation density ρ c It is related to the number of clutter and its distribution in the measurement space. Assuming that the number of clutter follows a Poisson distribution with parameter λ, the probability that the number of clutter is k can be expressed as:

[0117]

[0118] Cut-off density ρ c The calculation formula is:

[0119] ρ c =round(Den max )

[0120] ρ c ≈round(0.97711n(λ)+0.2093)

[0121] Among them, Den max The maximum value of λ that indicates that the time probability of assigning 1 of the λ clutter measurements in the measurement space to the same measurement unit is much less than 0.1.

[0122] Step S62: performing secondary clustering on the remaining measurement values ​​according to the local density and density distance of the remaining measurement values.

[0123] It should be noted that the secondary clustering refers to the division based on the local density and density distance of the remaining measurement values, and the local density and density distance of any two measurement values ​​in the remaining measurement values ​​are compared, and the division is performed based on the comparison result.

[0124] It should be understood that the local density ρ i Equal to the measurement value z in the measurement space i The spatial distance is less than the cutoff distance dc The number of measurement values ​​reflects the density of the measurement values.

[0125] Density distance δ i It reflects the distance between another more dense measurement value. It can be calculated based on the local density ρ i Distance to density δ i , assuming the cutoff density is ρ c , the cut-off density distance δ c The remaining measurements are clustered by FDPC, where the cutoff density distance is δ c The calculation formula is:

[0126] δ c =2B c

[0127] It should be emphasized that the secondary clustering is achieved by comparing the local density of the first measurement value with the local density of the second measurement value, and comparing the density distance of the first measurement value with the density distance of the second measurement value, wherein the first measurement value and the second measurement value are any measurement values ​​among the remaining measurement values; selecting the cluster center of the secondary clustering according to the comparison result; and performing secondary clustering on the remaining measurement values ​​according to the cluster center.

[0128] When the local density of the first measurement value is greater than the local density of the second measurement value, and the density distance of the first measurement value is greater than the density distance of the second measurement value, the first measurement value is taken as the measurement center; when the local density of the first measurement value is less than the local density of the second measurement value, and the density distance of the first measurement value is greater than the density distance of the second measurement value, the first measurement value is eliminated; when the density distance of the first measurement value is less than the density distance of the second measurement value, the first measurement value is assigned to an adjacent cluster, and the adjacent cluster is any one of the primary cluster and the secondary cluster.

[0129] In the specific experiment, a scene with dense tracking targets collected in the project is used. When the scene radius is R = 200km, the data within a measurement threshold is clustered. The identity of the measurement data within the measurement threshold is known. The measurement period is T = 1s. Assume that the maximum flight speed of the aircraft in this scene is V max =850m / s. Therefore, the cutoff distance d of FDPC c =850m, cut-off density distance δ c =1700m. Set the cutoff density ρ c = 4. Select 30 cycles of measurement data for comparison and refer to the following table:

[0130] Table 1:

[0131] Accuracy False alarm rate Missed detection rate FDPC 91.49％ 8.51％ 11.24％ This application 95.54％ 4.46％ 2.71％

[0132] Table 2:

[0133] purity F-number Parameter coefficient FDPC 71.64％ 77.38％ 75.67％ This application 87.24％ 91.98％ 90.44％

[0134] In the table, purity, F value and parameter coefficient are all clustering performance indicators in the experimental process, wherein the parameter coefficient is the Jacarrd coefficient. By comparison, it can be seen that the present invention is superior to the FDPC algorithm in accuracy, false alarm rate, missed detection rate, purity, F value and Jacarrd coefficient.

[0135] This embodiment selects a measurement value whose local density is greater than other measurement values ​​and whose density distance is greater than other measurement values ​​as a cluster center according to the local density and density distance of the remaining measurement values, and performs secondary clustering on the remaining measurement values ​​according to the cluster center, thereby eliminating the measurement values ​​corresponding to the clutter to a certain extent, and can effectively cluster the remaining measurement values ​​and identify the aircraft in the remaining measurements.

[0136] Reference Fig. 9 , Fig. 9 This is a structural block diagram of the first embodiment of the multi-target recognition device of the present invention.

[0137] like Fig. 9 As shown, the multi-target recognition device proposed in the embodiment of the present invention includes:

[0138] The measurement value acquisition module 10 is used to acquire a measurement data set corresponding to a preset measurement cycle, wherein the measurement data set includes measurement data of a known aircraft and an aircraft to be identified;

[0139] The measurement value acquisition module 10 is further used to filter out measurement values ​​carrying preset message information from the measurement data set;

[0140] The measurement value acquisition module 10 is further used to determine the historical motion trajectory of the known aircraft according to the measurement value carrying the preset message information;

[0141] The measurement value acquisition module 10 is further used to construct an elliptical threshold according to the historical motion trajectory;

[0142] A target identification module 20, configured to perform a clustering operation on other measurement values ​​except the measurement value carrying the preset message information by using the elliptical threshold;

[0143] The target recognition module 20 is also used to perform secondary clustering on the remaining measurement values ​​after the primary clustering by using preset parameters;

[0144] The target identification module 20 is further used to identify all aircraft within a preset measurement period based on the two clustering results.

[0145] This embodiment performs a primary clustering of known aircraft among multiple aircraft measurement values ​​through a preset message and then performs a secondary clustering of the remaining aircraft to complete the identification of all aircraft, thereby solving the problem of clustering adjacent measurement values ​​as the same aircraft measurement values ​​when the aircraft are relatively close when identifying multiple aircraft measurement values, thereby identifying multiple aircraft as the same aircraft. This embodiment achieves the ability to distinguish measurement values ​​belonging to different aircraft from multiple measurement values ​​and cluster them when the multiple aircraft are relatively close, thereby more accurately identifying multiple aircraft that are relatively close.

[0146] It should be noted that the workflow described above is merely illustrative and does not limit the scope of protection of the present invention. In practical applications, technicians in this field can select part or all of them according to actual needs to achieve the purpose of the present embodiment, and no limitation is made here.

[0147] In addition, for technical details not fully described in this embodiment, reference may be made to the multi-target recognition method provided in any embodiment of the present invention, and will not be repeated here.

[0148] In one embodiment, the measurement value acquisition module 10 is also used to obtain known aircraft motion information based on the historical motion trajectory; obtain the inverse matrix of the residual covariance matrix of the known aircraft based on the known aircraft motion information; obtain a preset threshold value; and establish an elliptical threshold based on the inverse matrix of the residual covariance matrix, the measurement value carrying the preset message information, and the preset threshold value.

[0149] In one embodiment, the target identification module 20 is further configured to determine whether other measurement values ​​except the measurement value carrying the preset message information are within the ellipse threshold.

[0150] Perform a clustering based on the judgment results.

[0151] In one embodiment, the target identification module 20 is further used to determine whether other measurement values ​​except the measurement value carrying the preset message information are within the elliptical threshold; and perform clustering according to the determination result.

[0152] In one embodiment, the target recognition module 20 is further used to calculate the local density and density distance of the remaining measurement values ​​according to the cutoff distance, the cutoff function and the maximum speed corresponding to the remaining measurement values ​​after the first clustering; and perform a second clustering on the remaining measurement values ​​according to the local density and density distance of the remaining measurement values.

[0153] In one embodiment, the target recognition module 20 is further used to compare the local density of the first measurement value with the local density of the second measurement value, and compare the density distance of the first measurement value with the density distance of the second measurement value, wherein the first measurement value and the second measurement value are any measurement values ​​among the remaining measurement values; select a cluster center of secondary clustering according to the comparison result; and perform secondary clustering on the remaining measurement values ​​according to the cluster center.

[0154] In one embodiment, the target recognition module 20 is further used to use the first measurement value as the measurement center when the local density of the first measurement value is greater than the local density of the second measurement value, and the density distance of the first measurement value is greater than the density distance of the second measurement value; to eliminate the first measurement value when the local density of the first measurement value is less than the local density of the second measurement value, and the density distance of the first measurement value is greater than the density distance of the second measurement value; and to assign the first measurement value to an adjacent cluster when the density distance of the first measurement value is less than the density distance of the second measurement value, wherein the adjacent cluster is any one of the primary cluster and the secondary cluster.

[0155] In one embodiment, the target recognition module 20 is further configured to obtain cluster centers of two clusterings; and determine the positions of all aircraft within the preset measurement period based on the cluster centers.

[0156] Other embodiments or specific implementations of the multi-target recognition device of the present invention can refer to the above-mentioned method embodiments, which will not be described in detail here.

[0157] In addition, an embodiment of the present invention further provides a storage medium, on which a multi-target recognition program is stored, and when the multi-target recognition program is executed by a processor, the steps of the multi-target recognition method described above are implemented.

[0158] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or system including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or system. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or system including the element.

[0159] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0160] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as a read-only memory / random access memory, a magnetic disk, or an optical disk), and includes a number of instructions for a terminal device (which can be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to execute the methods described in each embodiment of the present invention.

[0161] The above are only preferred embodiments of the present invention, and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A multi-target recognition method, characterized in that: The multi-target recognition method comprises: Acquire a measurement data set corresponding to a preset measurement period, wherein the measurement data set includes measurement data of a known aircraft and an aircraft to be identified; Filtering out measurement values ​​carrying preset message information from the measurement data set; Determine the historical motion trajectory of the known aircraft according to the measurement value carrying the preset message information; Constructing an elliptical threshold according to the historical motion trajectory; Clustering the other measurement values ​​except the measurement value carrying the preset message information once by using the elliptical threshold; Performing secondary clustering on the remaining measurement values ​​after the primary clustering by using preset parameters, wherein the preset parameters include: a cutoff distance, a cutoff function, and a maximum speed corresponding to the remaining measurement values ​​after the primary clustering; Based on the two clustering results, all aircraft within the preset measurement period are identified; Calculate the local density and density distance of the remaining measurement values ​​according to the cutoff distance, the cutoff function and the maximum speed corresponding to the remaining measurement values ​​after the first clustering; The remaining measurement values ​​are clustered for the second time according to the local density and density distance of the remaining measurement values.

2. The multi-target recognition method according to claim 1, characterized in that: The step of constructing an elliptical threshold according to the historical motion trajectory includes: Obtaining known aircraft motion information according to the historical motion trajectory; Obtaining an inverse matrix of a residual covariance matrix of the known aircraft according to the known aircraft motion information; Get the preset threshold value; An elliptical threshold is established according to the inverse matrix of the residual covariance matrix, the measurement value carrying the preset message information, and the preset threshold value.

3. The multi-target recognition method according to claim 1, characterized in that: The clustering of other measurement values ​​except the measurement value carrying the preset message information by the elliptical threshold includes: Determining whether other measurement values ​​except the measurement value carrying the preset message information are within the elliptical threshold; Perform a clustering based on the judgment results.

4. The multi-target recognition method according to claim 1, characterized in that: The performing secondary clustering on the remaining measurement values ​​according to the local density and density distance of the remaining measurement values ​​comprises: Comparing the local density of the first measurement value with the local density of the second measurement value, and comparing the density distance of the first measurement value with the density distance of the second measurement value, wherein the first measurement value and the second measurement value are any measurement value in the remaining measurement values; Select the cluster center of the secondary clustering according to the comparison results; The remaining measurement values ​​are clustered again according to the cluster centers.

5. The multi-target recognition method according to claim 4, characterized in that: The selecting the cluster center of the secondary clustering according to the comparison result includes: When the local density of the first measurement value is greater than the local density of the second measurement value, and the density distance of the first measurement value is greater than the density distance of the second measurement value, the first measurement value is used as the measurement center; When the local density of the first measurement value is less than the local density of the second measurement value, and the density distance of the first measurement value is greater than the density distance of the second measurement value, the first measurement value is eliminated; When the density distance of the first measurement value is smaller than the density distance of the second measurement value, the first measurement value is classified into a neighboring cluster, and the neighboring cluster is any one of the primary cluster and the secondary cluster.

6. The multi-target recognition method according to claim 1, characterized in that: After all aircraft within a preset measurement period are identified based on the two clustering results, the method further includes: Get the cluster centers of the two clusterings; The positions of all aircrafts within the preset measurement period are determined based on the cluster center.

7. A multi-target recognition device, characterized in that: The multi-target recognition device comprises: A measurement value acquisition module, used to acquire a measurement data set corresponding to a preset measurement cycle, wherein the measurement data set includes measurement data of known aircraft and aircraft to be identified; The measurement value acquisition module is further used to filter out measurement values ​​carrying preset message information from the measurement data set; The measurement value acquisition module is further used to determine the historical motion trajectory of the known aircraft according to the measurement value carrying the preset message information; The measurement value acquisition module is further used to construct an elliptical threshold according to the historical motion trajectory; A target identification module, configured to perform a clustering operation on other measurement values ​​except the measurement value carrying the preset message information by using the elliptical threshold; The target recognition module is further used to perform secondary clustering on the remaining measurement values ​​after the primary clustering according to preset parameters, wherein the preset parameters include: a cutoff distance, a cutoff function, and a maximum speed corresponding to the remaining measurement values ​​after the primary clustering; The target recognition module is further used to recognize all aircraft within a preset measurement period based on the two clustering results; The target recognition module is also used to calculate the local density and density distance of the remaining measurement values ​​according to the truncation distance, the truncation function and the maximum speed corresponding to the remaining measurement values ​​after the first clustering; and perform a second clustering on the remaining measurement values ​​according to the local density and density distance of the remaining measurement values.

8. A multi-target recognition device, characterized in that: The device comprises: a memory, a processor, and a multi-target recognition program stored in the memory and executable on the processor, wherein the multi-target recognition program is configured to implement the multi-target recognition method according to any one of claims 1 to 6.

9. A storage medium, characterized in that: The storage medium stores a multi-target recognition program, which, when executed by a processor, implements the multi-target recognition method according to any one of claims 1 to 6.

Images