Track management method based on multi-frame energy accumulation

Abstract

The application discloses a track management method based on multi-frame energy accumulation. The method mainly solves the problem that the prior art is difficult to maintain the track of a long-distance low signal-to-noise ratio target under the condition of a low false track maintenance probability, and the implementation scheme is as follows: a predicted gate is established by using tracking information to predict the target state; low threshold detection is performed in the predicted gate; the detection result is data-associated with a historical track; the associated track is continuously tracked or terminated according to different situations of the associated track; if the associated track meets the n / m logic method criterion, the track is continuously tracked, otherwise the track is terminated; for the continuously tracked track, it is judged whether the current associated result is empty; if the associated result is empty, the next frame of data is processed, otherwise the associated result is tracked and filtered; whether the filtered track is a false track is judged; if the track is a false track, the track is cancelled, otherwise the track is continuously tracked. The application can improve the target track maintenance probability under the condition of a low false track maintenance probability, and can be used for the detection and tracking of a low signal-to-noise ratio target.

Description

Technical Field

[0001] This invention belongs to the field of radar technology, specifically a track management method that can be used for radar target detection and tracking. Technical Background

[0002] Traditional radar detection and tracking are unidirectional. During the detection phase, prior information about targets within the monitored area is typically lacking, requiring a high detection threshold to reduce false alarms. However, a high threshold makes it difficult to detect low signal-to-noise ratio (SNR) targets at long range. In contrast, after track initiation, the system possesses sufficient prior target information. Based on this prior information, more stable target tracking can be achieved. Specifically, the tracking information can be used to predict the target state at the next moment, establishing a prediction gate, and performing low-threshold detection within the gate. This approach can control the number of false alarms while improving the detection performance of low SNR targets at long range, thereby increasing the probability of track maintenance. However, low-threshold detection may cause the track to be associated with false alarm points, generating some false tracks and affecting the analysis of the target state. Therefore, a track management method is needed to cancel false tracks while maintaining the target track with low SNR.

[0003] Patent application number 202110224659.5 discloses a multi-target tracking algorithm based on a track management method. It measures track lifetime and track score by the duration of track not being updated; short updates increase track lifetime and score, while short updates decrease them. Tracks are terminated when the lifetime is 0 or the score is negative. If the number of historical associations for terminated tracks is low, the track is considered a false track; otherwise, it is identified as a target track and output. By introducing the track not-updated duration, track interruption and loss are prevented; and the number of historical associations suppresses false tracks. However, under low-threshold detection conditions, the algorithm suffers from increased false points within the gate, making it easier for false points to associate with tracks. This results in shorter update durations, making it difficult to terminate false tracks. Furthermore, the increased number of historical associations for false tracks makes it impossible to effectively revoke false tracks by limiting the number of historical associations.

[0004] Patent application number 201811134571.9 discloses a heterogeneous multi-sensor multi-target tracking method. This method employs a variable evidence set approach, defining track position status, carrier frequency, repetition frequency, and pulse width as the effective evidence set. Based on the type of candidate measurement, it determines whether the candidate measurement possesses position, signal characteristics, and other information. The method automatically reduces the associated evidence set according to the measurement type, determining the measurement-related evidence set. Position correlation function and signal feature correlation function are calculated separately, and a comprehensive measurement correlation function is obtained by combining the evidence. This comprehensive correlation function effectively utilizes multi-sensor, multi-dimensional measurement information to calculate the track's comprehensive score, making the data association results more accurate and improving the continuity of the target tracking track. However, as the detection threshold decreases, the number of false points detected by a single sensor increases. It becomes difficult to distinguish between the target and false points using the position and signal characteristics of candidate measurements, thus significantly increasing the probability of association between false points and the track. The probability of false tracks also increases, making it impossible to effectively revoke false tracks under low-threshold detection conditions. Summary of the Invention

[0005] The purpose of this invention is to address the shortcomings of the prior art by proposing a track management method based on multi-frame energy accumulation, which can improve the detection performance of distant low signal-to-noise ratio targets and increase the track maintenance probability at long distances while ensuring a low false track maintenance probability.

[0006] The technical approach to achieve the objective of this invention is as follows: by using the tracking information of the trajectory, the target state at the next moment is predicted, and a prediction gate is established. Low-threshold detection is performed within the gate to improve the detection performance of targets with low signal-to-noise ratio. By accumulating energy of the trajectory over multiple frames, a threshold for low false trajectory maintenance probability is set. Trajectories below the decision threshold are identified as false trajectories and are revoked. This improves the target trajectory maintenance probability at long distances while ensuring a low false trajectory maintenance probability.

[0007] Based on the above approach, the trajectory management method of the present invention, which is based on multi-frame energy accumulation, includes the following steps:

[0008] (1) Estimate the target state based on the last frame after the start of the track. Its state covariance matrix D k|k Predict the target state in the next frame Its covariance matrix D l|k ;

[0009] (2) Based on the forecast information and D l|k Establish a predictive gate and set the false alarm probability to P. fa Low-threshold detection is performed on the data that has already undergone signal processing and falls within the predicted gate to obtain the detection result Z at the current time l.l ;

[0010] (3) The detection result Z at the current time l l By correlating the data with historical flight paths, the correlation result z is obtained. l And determine whether the associated track satisfies the n / m logical criterion:

[0011] If the conditions are not met, the flight path is terminated.

[0012] If satisfied, proceed to step (4);

[0013] (4) Determine the correlation result z l Is it empty?

[0014] If the association result z l If it is empty, then let time l = l + 1 and return to step (1);

[0015] If the association result z l If not empty, continue tracking and execute step (5);

[0016] (5) Perform tracking filtering on the associated points and calculate the estimated target state at time l. With covariance matrix D l|l and update the track;

[0017] (6) Let the energy accumulation window length be N. Perform multi-frame energy accumulation on the track segments within the energy accumulation window after the update, and obtain the energy accumulation value Y. l ;

[0018] (7) Set the probability of maintaining a false track to P. t|fa Calculate the corresponding track decision threshold v t_track ;

[0019] (8) Determine the energy accumulation value Υ l Is it below the track decision threshold v? t_track

[0020] If the value is below the threshold, the track segment within the window is considered a false track and is cancelled.

[0021] Otherwise, continue tracking, set time l = l + 1, and return to step (1).

[0022] Compared with the prior art, the present invention has the following advantages.

[0023] Firstly, this invention accumulates energy across multiple frames of a track to cancel tracks that are below the false track determination threshold, thereby ensuring a low probability of maintaining false tracks.

[0024] Secondly, this invention uses tracking information to assist in establishing a predictive gate for low-threshold detection. While controlling the number of false alarms, it can ensure the detection performance of targets with low signal-to-noise ratio, thereby improving the probability of track maintenance at long distances. Attached Figure Description

[0025] Figure 1 This is a flowchart illustrating the implementation of the present invention.

[0026] Figure 2 This is a comparison chart of the track maintenance probability of the present invention and traditional detection and tracking algorithms at different distances. Detailed Implementation

[0027] To enable those skilled in the art to better understand the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0028] It should be noted that the step numbers in the specification and claims of this invention are only for the purpose of clearly describing the embodiments of this invention and facilitating understanding, and their order is not limited.

[0029] Reference Figure 1 The implementation steps for this example are as follows:

[0030] Step 1: Estimate the target state based on the last frame after the track starts. Its state covariance matrix D k|k Predict the target state in the next frame Its covariance matrix D l|k .

[0031] The prediction refers to obtaining the possible state of the target at time l based on its motion state at time k and its motion trend, thus obtaining the prediction information of the target at time l.

[0032] The specific implementation of this step is as follows:

[0033] (1.1) Estimate the state of the target at time k Predict the target state at time l

[0034]

[0035] Among them, F l|k Let k be the transition matrix from the target state at time k to time l.

[0036] (1.2) Based on the target state covariance matrix D at time k k|k Calculate the target predicted state covariance matrix D at time l. l|k :

[0037]

[0038] Among them, Q k|k Let be the process noise covariance matrix from time k to time l.

[0039] Step 2, based on the prediction information and D l|k A prediction gate is established, and low-threshold detection is performed on the data that has already undergone signal processing and falls within the prediction gate to obtain the detection result Z at the current time l. l .

[0040] The prediction gate refers to the gate based on the target prediction information at time l. and D l|k It estimates the area where the target may appear. Since the prediction gate only covers a portion of the input observation space, although low-threshold detection leads to a higher false alarm probability, the number of false alarms is limited, while improving the detection performance for low signal-to-noise ratio targets.

[0041] The specific implementation method for this step is as follows:

[0042] (2.1) Target prediction state Switching to the measurement space, denoted as

[0043]

[0044] Among them, h l Let l be the target measurement function at time l;

[0045] (2.2) Based on the target prediction state covariance matrix D at time l l|k Calculate the target measurement status The prediction covariance matrix B l|k :

[0046]

[0047] Among them, H l Let h be the target measurement function at time l. l Jacobian matrix, H represents l transpose;

[0048] (2.3) Based on the predicted target measurement status With the predicted covariance matrix B l|k Establish the following prediction gate:

[0049]

[0050] Among them, y l Indicates the possible location of the target in the measurement space. Representation matrix B l|k The reverse,

[0051] g is the prediction gate parameter, which represents the gate size and is used to ensure that the target falls into the prediction gate with a high probability.

[0052] (2.4) Set the false alarm probability to P fa Calculate its corresponding detection threshold v t

[0053] v t =-ln(P fa );

[0054] (2.5) Compare the data after signal processing within the prediction gate with the detection threshold v. t Size:

[0055] If the energy of the dots in the data exceeds the detection threshold v t Then this part of the dot pattern is recorded as Let m be the j-th point at time l, and m be the number of points detected at time l.

[0056] If the energy of the dots in the data does not exceed the detection threshold v t If the data is not a target, then it is considered to be discarded.

[0057] Step 3, take the detection result Z at the current time l. l By correlating the data with historical flight paths, the correlation result z is obtained. l And determine whether the associated track satisfies the n / m logic criterion.

[0058] The data association mentioned refers to the algorithm for associating detected points with historical tracks, including but not limited to nearest neighbor, global nearest neighbor, probabilistic data association algorithm and joint probabilistic data association algorithm.

[0059] The n / m logic criterion refers to determining whether the track has successfully associated with no less than n target points within the last m frames. If it has failed to successfully associate with no less than n target points, it means that the target has not been effectively observed within the m frames, and therefore the track can be terminated.

[0060] The specific implementation method for this step is as follows:

[0061] (3.1) The association result z is obtained by using the association algorithm to associate the detection result at the current time l with the historical track. l ,

[0062]

[0063] in, For historical flight track information, Z l This represents the detection result at time l, and association represents the data association algorithm.

[0064] (3.2) For the association result z l and historical navigation trails The total number of target frames detected from frame l-m+1 to frame l is s, where m is the window length of the time window in the n / m logic criterion.

[0065] (3.3) Compare the target frame number s with the detection threshold n in the n / m logic criterion:

[0066] If s≥n, then the n / m logical criterion is satisfied, and step 4 is executed.

[0067] If s < n, then the n / m logical rule is not satisfied, and the track is terminated.

[0068] Step 4, determine the association result z l Is it empty? Select to return to step 1 or proceed to step 5.

[0069] If the current association result is empty, it means that no target was detected at the current moment, and the data at the next moment needs to be detected. Therefore, we need to return to step 1 and process the data at the next moment.

[0070] The specific implementation method for this step is as follows:

[0071] If the association result z l If it is empty, then let time l = l + 1 and return to step (1);

[0072] If the association result z l If not empty, continue tracking and execute step (5);

[0073] Step 5: Perform tracking filtering on the associated points and calculate the estimated target state at time l. With covariance matrix D l|l And update the track.

[0074] The tracking filtering mentioned in this step refers to tracking filtering algorithms, including but not limited to Kalman filtering, extended Kalman filtering, and unscented Kalman filtering.

[0075] The specific implementation method for this step is as follows:

[0076] (5.1) Target estimation state

[0077] (5.1.1) Measure the noise covariance matrix R at time l. l The predictive covariance matrix B of the target measurement state l|k Calculate the new information covariance matrix: S l =B l|k +R l ;

[0078] (5.1.2) Based on the target prediction state covariance matrix D at time l l|k Target measurement function h at time l l Jacobian matrix H l With the new information covariance matrix S l Calculate the Kalman gain:

[0079] (5.1.3) Predict the state based on the target at time l Kalman gain K l The correlation result z at time l l Target prediction state in measurement space Calculate the posterior mean of the target state at time l.

[0080]

[0081] (5.2) According to the Kalman gain K l Target measurement function h at time l l Jacobian matrix H l The covariance matrix D of the target predicted state at time l l|k Calculate the posterior covariance matrix D of the target state at time l. l|l

[0082] D l|l =D l|k -K l H l D l|k

[0083] (5.3) Update track information

[0084] Step 6: Perform multi-frame energy accumulation on the updated track segments within the energy accumulation window to obtain the energy accumulation value Y. l .

[0085] The accumulated energy window mentioned in this step refers to the frame interval composed of N points successfully associated at the end of the track. Its significance is that if only the energy of the last point is considered, it is difficult to distinguish whether it is a false point or a target point with low signal-to-noise ratio; if the points of the entire track are considered, when the track is associated with a false point, it will be affected by many real target points in the track, which will lead to the false track being canceled in a timely manner.

[0086] The specific implementation method for this step is as follows:

[0087] Given an energy accumulation window length of N, calculate the energy accumulation value Y across multiple frames. l :

[0088]

[0089] Where γ i Let l-num(l) represent the energy of the point at time i in the track. If the track is not associated with any point in a frame, the energy of the point in that frame is considered to be 0. l-num(l) represents the frame number of the first target among the N consecutive successfully associated targets, including time l.

[0090] Step 7, set the probability of maintaining a false track to P. t|fa Calculate the corresponding track decision threshold v t_track .

[0091] The significance of the false track maintenance probability mentioned in this step is the probability that when the accumulated energy from multiple frames corresponds to a false point, i.e., the track segment is a false track, it is still determined to be the target track. Since false tracks can affect the determination of the target track, it is necessary to control the false track maintenance probability. The specific implementation steps include the following:

[0092] (7.1) The noise signal-to-noise ratio (SNR) is modeled and follows an exponential distribution with parameter 1, i.e.:

[0093] f(SNR) = e -SNR (SNR>0)

[0094] (7.2) The first detection threshold v will be passed. t Let ξ be the signal-to-noise ratio of the false dots, and its probability density function is a truncated version of the noise signal-to-noise ratio probability density function f(SNR). Based on the form of the noise signal-to-noise ratio probability density function f(SNR), the dotted dot is passed through the first detection threshold v. t The probability density function of the signal-to-noise ratio ξ of the spurious traces is expressed as:

[0095]

[0096] (7.3) Let the intermediate variable μ i =ξ i -vt,i , where ξ i v represents the signal-to-noise ratio of spurious dots that pass the first detection threshold in the i-th frame. t,i Represents the first detection threshold of the i-th frame, based on the threshold v. t The probability distribution form of the signal-to-noise ratio ξ of the spurious dots, with the intermediate variable μ i The probability density distribution is expressed as:

[0097]

[0098] That is, the intermediate variable μ i It follows an exponential distribution with parameter 1;

[0099] (7.4) Let the sum of the signal-to-noise ratios of the spurious points in multiple frames be... Then it and the intermediate variable μ i The relationship is:

[0100]

[0101] Among them, v t,i Let be the threshold for detecting the processed data at time i.

[0102] (7.5) Let the sum of the intermediate variables For η, since μ i They follow an exponential distribution with parameter 1, and are independent and identically distributed among themselves, so η follows a gamma distribution with parameter (N,1).

[0103] (7.6) Let the false alarm probability of η be the false track maintenance probability P. t|fa Based on its probability density distribution, the false alarm probability P of η is... t|fa The expression is as follows:

[0104]

[0105] Among them, the lower limit of integration Let Γ(N) be the initial threshold value, and Γ(N) = (N-1)!

[0106] (7.7) According to step (7.4), we obtain Based on this relationship, the probability of maintaining a false track is calculated to be P. t|fa In the case of multiple frames of false traces, the track decision threshold θ is:

[0107]

[0108] Among them, v t,i Let be the threshold for detecting the processed data at time i.

[0109] Step 8, determine the energy accumulation value Yl Is it below the track decision threshold v? t_track Choose whether to cancel the flight path.

[0110] If the cumulative energy value of a track across multiple frames is lower than the track decision threshold, it means that it has 1-P t|fa The probability of this is a false track. Since false tracks need to be suppressed, the probability of maintaining a false track is generally set to a low value. Therefore, this track has a high probability of being a false track and needs to be canceled.

[0111] The specific implementation of this step is to convert the energy accumulation value Y. l With track decision threshold v t_track Comparison:

[0112] If Υ l <v t_track If so, the track segment within the window is considered a false track and is cancelled;

[0113] If Υ l ≥v t_track Continue tracking, set time l = l + 1, and return to step (1).

[0114] The effects of this invention can be further illustrated by the following simulation experiments:

[0115] I. Simulation Conditions

[0116] The radar is set to operate at a frequency of 1 GHz with a bandwidth of 1 MHz, θ 3dB The initial position is 2°, the data rate is 2 seconds, the prediction gate parameter g is 4, the first detection threshold is 12.8 dB at the beginning of the track; the initial target position is [80 km, 80 km]. T With an azimuth angle of 45°, the target moves away from the radar at a speed of 200 m / s. The target's RCS fluctuation model is Swerling I. The average signal-to-noise ratio of the target at 200 km is set to 12 dB, and the average target signal-to-noise ratio is inversely proportional to the fourth power of the distance.

[0117] The data association algorithm used in this invention is the nearest neighbor algorithm, which adopts the 4 / 10 logic criterion. The track filtering algorithm adopts the transformed measurement Kalman filter algorithm. In the track decision, the multi-frame energy accumulation window length N=4.

[0118] Traditional single-frame detection and tracking algorithms assume a false alarm probability of 10 for their detection threshold. -6 Its data association algorithm selects the nearest neighbor algorithm, its track filtering algorithm selects the transformed measurement Kalman filter algorithm, and if no target is detected for three consecutive frames, the track is determined to terminate.

[0119] II. Simulation Content

[0120] Under the above simulation conditions, 100,000 Monte Carlo experiments were conducted using both the detection and tracking algorithms of this invention and the traditional ones. The relationship between the track maintenance probability and distance was statistically analyzed, and the results are as follows: Figure 2 As shown, the horizontal axis represents distance, and the vertical axis represents the track maintenance probability. Curves 1, 2, 3, and 4 are the results of the relationship between track maintenance probability and distance statistically analyzed under different frame false alarm probabilities and false track maintenance probabilities, respectively; curve 5 is the statistical result of traditional detection and tracking.

[0121] As can be seen from the comparison between curve 1 and curve 3, the larger the first detection threshold, the lower the frame false alarm probability and the worse the track maintenance effect.

[0122] As can be seen from the comparison between curve 1 and curve 2, the larger the track decision threshold, the lower the probability of false track maintenance and the worse the track maintenance effect.

[0123] Comparing curves 1 and 5, we can see that the track maintenance probability is 0.5 and the frame false alarm probability is 10. -3 The probability of maintaining a false track is 10%. -3 Under these conditions, the maintenance distance of this invention is 186.5 km, while the maintenance distance of the traditional detection and tracking algorithm is 132.3 km. Compared with the traditional algorithm, the track maintenance distance of this invention is extended by 40.1%. This is achieved with a track maintenance probability of 0.9 and a frame false alarm probability of 10. -3 The probability of maintaining a false track is 10%. -3 Under these conditions, the maintenance distance of the present invention is 149.8 km, while the maintenance distance of the traditional detection and tracking algorithm is 121.5 km. Compared with the traditional algorithm, the track maintenance distance of the present invention is extended by 23.3%.

[0124] Comparing curves 4 and 5, it can be seen that the track maintenance probability is 0.5 and the frame false alarm probability is 10. -4 The probability of maintaining a false track is 10%. -4 Under these conditions, the maintenance distance of this invention is 178.1 km, while the maintenance distance of the traditional detection and tracking algorithm is 132.3 km. Compared with the traditional algorithm, the track maintenance distance of this invention is extended by 34.6%. This is achieved with a track maintenance probability of 0.9 and a frame false alarm probability of 10. -4 The probability of maintaining a false track is 10%. -4 Under the given conditions, the maintenance distance of the present invention is 146.7 km, while the maintenance distance of the traditional detection and tracking algorithm is 121.5 km. Compared with the traditional algorithm, the track maintenance distance of the present invention is extended by 20.7%.

[0125] In summary, compared with traditional single-frame detection and tracking algorithms, this invention can improve the probability of maintaining a track at a long distance while ensuring a low probability of false track maintenance.

Claims

1. A trajectory management method based on multi-frame energy accumulation, characterized in that: include: (1) Estimate the target state based on the last frame after the start of the track. Its state covariance matrix Predict the target state in the next frame Its covariance matrix ; (2) Based on the forecast information and Establish a predictive gate and set the false alarm probability as follows: Low-threshold detection is performed on the data that has already undergone signal processing and falls within the prediction gate to obtain the current... Time detection results ; (3) The current Time detection results Data is correlated with historical flight tracks to obtain correlation results. And determine whether the associated track meets the requirements. Logical criterion: If the conditions are not met, the flight path is terminated. If satisfied, proceed to step (4). (4) Determine the association results Is it empty? If the association result If it is empty, then let the time be... Return to step (1); If the association result If not empty, continue tracking and execute step (5); (5) Perform tracking filtering on the associated points and calculate the current... Target estimation state at time 1 With covariance matrix and update the track; (6) Let the length of the accumulated energy window be... The updated track segments within the energy accumulation window are subjected to multi-frame energy accumulation to obtain the energy accumulation value. ; (7) Set the probability of maintaining a false track as Calculate the corresponding track decision threshold as follows: The implementation is as follows: (7a) Set the probability of maintaining a false track The expression is as follows: ； Among them, the lower limit of integration This is the initial threshold value. ; (7b) Based on the initial threshold value Calculate the trajectory decision threshold ; ； in, for A threshold for constantly detecting data that has undergone signal processing; (8) Determine the energy accumulation value Is it below the track decision threshold? , If the value is below the threshold, the track segment within the window is considered a false track and is cancelled. Otherwise, continue tracking, keeping track of the situation. Return to step (1).

2. The method according to claim 1, characterized in that, Step (1) Estimate the target state based on the last frame after the track starts. Its state covariance matrix Predict the target state in the next frame Its covariance matrix The implementation is as follows: (1a) According to Time-based target estimation state predict Target state at time ； in, for Always towards The target state transition matrix at time t; (1b) According to Time-state covariance matrix ,calculate The target prediction state covariance matrix at time 1 : ； in, for Always towards The process noise covariance matrix at time step 1.

3. The method according to claim 1, characterized in that, In step (2), based on the prediction information and Establish a predictive gate, as follows: (2a) Transform the target prediction state into the measurement space, denoted as : ； in, for The target measurement function at time; (2b) Calculate the prediction covariance matrix of the target measurement state: , in, Represented as Time-based target measurement function Jacobian matrix, express transpose; (2c) Based on the predicted target measurement status With the predicted covariance matrix Establish the following prediction gate: ； in, Indicates the possible location of the target in the measurement space. Representation matrix The reverse, The predicted gate parameter represents the gate size and is used to ensure that the target falls into the predicted gate with a high probability.

4. The method according to claim 1, characterized in that, In step (2), the false alarm probability of the data that has been processed and falls within the prediction gate is calculated as follows: The low-threshold detection is implemented as follows: (2d) Calculate the false alarm probability Corresponding detection threshold ; ； (2e) Compare the data after signal processing within the prediction gate with the detection threshold. Size: If the energy of the dots in the data exceeds the detection threshold Then this part of the dot pattern is recorded as , In order to be in Time of the first A dot mark, for The number of dots detected at any given time; If the energy of the dots in the data does not exceed the detection threshold If the data is not a target, then it is considered to be discarded.

5. The method according to claim 1, characterized in that, Step (3) for the current time The correlation results are obtained by correlating the detection results with historical flight tracks. The following is a description: ； in, For historical flight track information, For the present The detection results at any given time This represents a data association algorithm.

6. The method according to claim 1, characterized in that, In step (3), it is determined whether the associated track meets the requirements. The logical criterion is implemented as follows: (3a) For the association results and historical navigation trails Overall statistics Frame to Total number of target frames detected ,in for The window length of the time window in the logical method criteria; (3b) Target frame number and Threshold for the number of tests in the logic method criterion Comparison: Then it satisfies Logical criterion If not, then it does not satisfy Logical principles.

7. The method according to claim 1, characterized in that, In step (5), calculate the current Target estimation state at time 1 With covariance matrix And update the track, as follows: (5a) Target estimation state : (5a1) According to Time measurement noise covariance matrix The predictive covariance matrix of the target measurement state Calculate the new information covariance matrix: ; (5a2) According to The target prediction state covariance matrix at time 1 , Time-based target measurement function Jacobian matrix With the new covariance matrix Calculate the Kalman gain: ; (5a3) According to Target prediction state at time 1 Kalman gain , Time-related results Target prediction state in measurement space calculate The posterior mean of the target state at time 1 : ； (5b) Based on Kalman gain , Time-based target measurement function Jacobian matrix and The target prediction state covariance matrix at time 1 ,calculate The posterior covariance matrix of the target state at time step ; ； (5c) Update track information .

8. The method according to claim 1, characterized in that, The energy accumulation value is obtained in step (6). , is represented as follows, ； in In the flight path The energy of the point corresponding to a given time. If the track is not associated with a point in a certain frame, the energy of the point in that frame is considered to be 0. The energy window for accumulating the flight path is long. Representatives include Continuous including moments The frame number of the first target among the successfully associated targets.

Images