Fusion track maintenance capability evaluation method for air-based platform observation scene

Abstract

The invention provides a fusion track maintenance capability evaluation method and device for an air-based platform observation scene. The method comprises the following steps: obtaining fusion track data obtained based on a fusion algorithm and true value track data corresponding to the fusion track data; key fields of the fused track data comprise a track identifier, a timestamp and a target space coordinate; key fields of the true-value track data comprise a true target identifier, a timestamp and a target space coordinate; according to the fusion flight path data and the true value flight path data corresponding to the fusion flight path data, the flight path maintenance proportions of n continuous time periods T are counted, the obtained n flight path maintenance proportions are multiplied to obtain the flight path maintenance proportion of which the total time length is n time periods T, n is a positive integer, and T is the time length of a single time window; and based on the flight path maintenance proportion, performing root extraction on the total duration to obtain the flight path maintenance rate w of the fusion algorithm.

Description

Technical Field

[0001] This invention relates to the fields of multi-sensor and multi-source information fusion technology, automation technology, and spaceborne observation technology, and particularly to a method, apparatus, electronic device, computer-readable storage medium, and computer program product for evaluating the fusion track maintenance capability in spaceborne platform observation scenarios. Background Technology

[0002] In observation and tracking surveillance scenarios using airborne platforms such as aircraft (manned or unmanned) and balloons, data association is a core step in forming fused tracks. Its task is to process real-time points of targets such as ships and aircraft around the platform, acquired after detection and localization by multi-source sensors (such as radar and optical equipment on the airborne platform), to form multiple continuous tracks. The accuracy of the association results (such as track association probability and ambiguity control) directly determines the quality of the fused tracks; association failure will lead to track interruptions, decreased accuracy, or false fusion. After data association forms the tracks, further processing using algorithms such as Kalman filtering can reduce the positioning error of each point, resulting in smoother and more accurate fused tracks.

[0003] Figure 1 In a typical multi-target tracking scenario, detection point data from three sources are processed through data association and track filtering to form a fused track (red). Most civil aircraft and cargo ships periodically and automatically report their positions, which are obtained by GPS devices on these targets. These targets then broadcast their identity and location data to the surrounding environment via ADS-B (Automatic Dependent Surveillance-Broadcast) or AIS (Automatic Identification System) messages. By comparing this fused track with the actual track obtained from these ADS-B or AIS messages (the green line in the figure), it can be determined whether the fused track has failed to maintain its position, such as losing the target (the track line is interrupted and restarted as a new track) or tracking the wrong target. The actual track is the GPS location reported by the target itself.

[0004] In scenarios where the truth value is known, the accuracy of data association and fusion algorithms and systems can be evaluated to obtain relevant metrics. In scenarios where the truth value is unknown, these metrics can be used to estimate the confidence level of the fused tracks. For example, unidentified or uncooperative aircraft, ships, or other manned or unmanned moving targets may not report ADS-B or AIS data in order to conceal their movements.

[0005] Existing technologies have established a multi-dimensional evaluation index system, among which the indicators related to accuracy include:

[0006] False track ratio: The ratio of the number of false tracks to the total number of tracks.

[0007] Correct correlation probability: The probability that the measured value correctly matches the true target.

[0008] Missed correlation probability: refers to the probability that a valid measurement generated by a real target fails to be successfully correlated with any tracking trajectory.

[0009] Error-related probability: The probability that a measurement is incorrectly associated with an incorrect target, reflecting the situation of misassociation.

[0010] Cumulative track interruption count: The cumulative number of times a track is interrupted due to association failure during its lifecycle.

[0011] Trajectory lifecycle: The length of time a real target lasts from the time the tracking algorithm successfully initializes (i.e., the start of the trajectory) until the trajectory is terminated.

[0012] Target identity switching count: The number of times the tracking track identifier (ID) of a real target is incorrectly changed throughout its entire lifecycle.

[0013] Each metric represents only a single aspect and partial performance of the algorithm; even weighted averaging or multiplication cannot reflect the algorithm's overall capability. Metrics such as correct correlation probability, missed correlation probability, and incorrect correlation probability, which reflect the correlation performance of a single data point with the track at a single moment, cannot reflect the ability to maintain the fused track, nor can they be inferred from these metrics. For example, a fusion algorithm or system with an 80% "correct correlation probability" may not sound perfect. However, these 20% correlation errors might be evenly and sporadically distributed across multiple targets at multiple moments, and the algorithm's powerful track management and state estimation modules can quickly correct these transient erroneous correlations. Through track fusion (i.e., updating the target state by integrating multiple observations), the track can still maintain stable and continuous tracking of the target, and the quality of its output fused track can be very high, fully meeting practical application requirements. Conversely, an algorithm with a 95% correct correlation probability, if its 5% error is a persistent confusion between two targets (i.e., identity swapping), will generate a track that is disastrous for decision-making.

[0014] To comprehensively and accurately evaluate the performance of algorithms, it is necessary to propose more comprehensive metrics and calculation methods that have clearer physical meanings and are more closely aligned with usage scenarios. These new metrics should possess the following characteristics:

[0015] More comprehensive: It can simultaneously reflect the accuracy, continuity, and error type of the association.

[0016] The physical meaning is more explicit: it is directly linked to the needs of end users, such as "How long can I continuously track this target without losing it?" or "Is the target identification output by the system stable?"

[0017] More relevant to usage scenarios: The metrics should directly assess the quality of the final generated track, rather than just the accuracy of intermediate correlation steps. Summary of the Invention

[0018] The purpose of this invention is to address the problem of the lack of a method for comprehensive capability evaluation of data association algorithms in the prior art, and to propose a method for evaluating and calculating the fusion track maintenance capability in airborne platform observation scenarios.

[0019] To address the shortcomings of existing technologies, such as Figure 14 As shown, this invention proposes a method for evaluating the fusion track maintenance capability in airborne platform observation scenarios, including:

[0020] Step 1: Obtain the fused track data and corresponding ground truth track data obtained based on the fusion algorithm; the key fields of the fused track data include: track identifier, timestamp, and target spatial coordinates; the key fields of the ground truth track data include: real target identifier, timestamp, and target spatial coordinates.

[0021] Step 2: Based on the fused track data and the corresponding ground truth track data, calculate the track maintenance ratio for n consecutive time periods of T, and multiply the n track maintenance ratios together to obtain the track maintenance ratio for a total duration of n time periods of T. , where n is a positive integer and T is the duration of a single time window;

[0022] Step 3: Based on the track maintenance ratio Taking the square root of the total duration yields the track maintenance rate w of the fusion algorithm.

[0023] The method for evaluating fusion track maintenance capability in airborne platform observation scenarios, wherein step 2 includes:

[0024] Step 2.1: Initialize the track maintenance ratio w(0,0) = 100%;

[0025] Step 2.2: Initialize the start and end times of the first time window, t1=0, t2=T;

[0026] Step 2.3: Calculate the fusion track maintenance ratio w(t1, t2) within a single window of duration T;

[0027] Step 2.4: Iteratively update the value of w(t2), w(t2) = w(0, t2) = w(0, t1) * w(t1, t2);

[0028] Step 2.5: After calculating the track maintenance ratio within the current time window, determine whether t2 equals nT. If so, save the current value of w(t2) as the track maintenance ratio. Otherwise, continue sliding the window forward by T: t1=t2, t2=t1+T.

[0029] The method for evaluating fusion track maintenance capability in airborne platform observation scenarios, wherein step 2.3 includes:

[0030] Step 2.3.1: Initialize the numerator N2 and denominator N1 in the formula for calculating the track maintenance ratio to 0 respectively; where N1 represents the number of fused tracks that are stably tracked at the start time t1 of the current time window, and N2 represents the number of fused tracks that are successfully maintained at the end time t2 of the current time window.

[0031] Step 2.3.2: Each fused track includes multiple point data points. The fused tracks are read one by one to determine whether each track entered a stable tracking state before time t1. The determination process includes sequentially checking whether three conditions are met:

[0032] Condition 1: Calculate whether the number of points on the track up to time t1 is not less than a specified threshold.

[0033] Condition 2: Before time t1, do all the last n points of the merged track originate from the same target M? i Measurement point data;

[0034] Condition 3, retrieval target M i In the last detection data point in the detection and positioning data (the detection and positioning data before fusion), determine whether the timestamp of this point is greater than or equal to t1;

[0035] If these three conditions are met, proceed to step 2.3.3;

[0036] Step 2.3.3: Add the fused track that meets the three conditions to the denominator N1 of the track maintenance ratio;

[0037] Step 2.3.4: Analyze whether there are any fused points in the current fused track within the time window [t1, t2]. If no fused points fall within this time window, determine whether the timestamp of the last point of the current fused track is less than t2, whether the tracked target has more than a preset number of detection point data after the current fused track terminates, and whether the maximum distance between the last point of the current fused track and the real position corresponding to the subsequent detection point of the tracked target is greater than a preset multiple of the positioning error. If all the judgment results are yes, the track maintenance fails; otherwise, it is determined to be maintained successfully, and step 2.3.6 is executed.

[0038] Step 2.3.5: If there is a fusion point trace in the current fusion track within the time window [t1, t2], then retrieve the last fusion point trace data pos with timestamp ≤ t2 in the current fusion track. 融合_i_k Based on the true value of the tracked target's trajectory, the pos is found through timestamp matching. 融合_i_k The target's true position at the corresponding moment is calculated. The distance between the fused point and the true position is calculated to be no more than a preset multiple of the positioning error. If so, the track is considered successfully maintained and step 2.3.6 is executed; otherwise, the track maintenance fails.

[0039] Step 2.3.6, increase N2 by 1;

[0040] Step 2.3.7: After all merged tracks have been traversed, calculate the track maintenance rate within the current time window: .

[0041] like Figure 15 As shown, this invention also proposes a fusion track maintenance capability assessment device for airborne platform observation scenarios, comprising:

[0042] Module 1: Obtain fused track data and corresponding ground truth track data based on the fusion algorithm; the key fields of the fused track data include: track identifier, timestamp, and target spatial coordinates; the key fields of the ground truth track data include: real target identifier, timestamp, and target spatial coordinates.

[0043] Module 2: Based on the fused track data and the corresponding ground truth track data, calculate the track maintenance ratio for n consecutive time periods of T, and multiply the n track maintenance ratios together to obtain the track maintenance ratio for a total duration of n time periods of T. , where n is a positive integer and T is the duration of a single time window;

[0044] Module 3, Based on Track Maintenance Ratio Taking the square root of the total duration yields the track maintenance rate w of the fusion algorithm.

[0045] The aforementioned fusion track maintenance capability assessment device for airborne platform observation scenarios, wherein module 2 includes:

[0046] Module 2.1: Initialize track maintenance ratio w(0,0)=100%;

[0047] Module 2.2: Initialize the start and end times of the first time window, t1=0, t2=T;

[0048] Module 2.3: Calculate the fusion track maintenance ratio w(t1, t2) within a single window of duration T;

[0049] Module 2.4: Iteratively update the value of w(t2), w(t2) = w(0, t2) = w(0, t1) * w(t1, t2);

[0050] Module 2.5: After calculating the track maintenance ratio within the current time window, determine whether t2 equals nT. If so, save the current value of w(t2) as the track maintenance ratio. Otherwise, continue sliding the window forward by T: t1=t2, t2=t1+T.

[0051] The aforementioned fusion track maintenance capability assessment device for airborne platform observation scenarios, wherein module 2.3 includes:

[0052] Module 2.3.1 initializes the numerator N2 and denominator N1 in the formula for calculating the track maintenance ratio to 0 respectively; where N1 represents the number of fused tracks that are stably tracked at the start time t1 of the current time window, and N2 represents the number of fused tracks that are successfully maintained at the end time t2 of the current time window.

[0053] Module 2.3.2: Each fused track includes multiple point data points. The fused tracks are read one by one to determine whether each track entered a stable tracking state before time t1. The determination process includes sequentially checking whether three conditions are met:

[0054] Condition 1: Calculate whether the number of points on the track up to time t1 is not less than a specified threshold.

[0055] Condition 2: Before time t1, do all the last n points of the merged track originate from the same target M? i Measurement point data;

[0056] Condition 3, retrieval target M i In the last detection data point in the detection and positioning data (the detection and positioning data before fusion), determine whether the timestamp of this point is greater than or equal to t1;

[0057] If these three conditions are met, then module 2.3.3 will be executed;

[0058] Module 2.3.3 adds the fused track that meets these three conditions to the denominator N1 of the track maintenance ratio;

[0059] Module 2.3.4 analyzes whether there are fused points in the current fused track within the time window [t1, t2]. If no fused points fall within this time window, it determines whether the timestamp of the last point of the current fused track is less than t2, whether the tracked target has more than a preset number of detection points after the current fused track terminates, and whether the maximum distance between the last point of the current fused track and the real position corresponding to the subsequent detection points of the tracked target is greater than a preset multiple of the positioning error. If all the judgment results are yes, the track maintenance fails; otherwise, it is determined to be maintained successfully, and module 2.3.6 is executed.

[0060] Module 2.3.5: If there is a fusion point trace in the current fused track within the time window [t1, t2], then retrieve the last fusion point trace data pos in the current fused track with timestamp ≤ t2. 融合_i_k Based on the true value of the tracked target's trajectory, the pos is found through timestamp matching. 融合_i_k The target's true position at the corresponding moment is calculated. The distance between the fused point and the true position is calculated to be no more than a preset multiple of the positioning error. If so, the track is considered successfully maintained and module 2.3.6 is executed; otherwise, the track maintenance fails.

[0061] In module 2.3.6, increase N2 by 1;

[0062] Module 2.3.7, after all merged tracks have been traversed, calculates the track maintenance rate within the current time window: .

[0063] The present invention also proposes a client for any of the aforementioned fusion track maintenance capability assessment devices for airborne platform observation scenarios.

[0064] The present invention also proposes an electronic device, including the aforementioned fusion track maintenance capability assessment device for airborne platform observation scenarios. The electronic device may be connected to an information display device, which is used to display the track maintenance rate using user-set display parameters, attributes, or through an artificial intelligence model.

[0065] The present invention also proposes a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of the fusion track maintenance capability assessment method for airborne platform observation scenarios.

[0066] The present invention also proposes a computer program product, including a computer program, wherein when the computer program is executed by a processor, it implements the steps of the fusion track maintenance capability assessment method for airborne platform observation scenarios.

[0067] As can be seen from the above solutions, the advantages of the present invention are:

[0068] This invention proposes a method for evaluating and calculating the fusion track sustaining capability in airborne platform observation scenarios, which can comprehensively evaluate the persistence and accuracy of multi-target tracking.

[0069] The fusion track maintenance rate proposed in this invention measures the ability of a multi-target tracking system to maintain stable target tracking in complex scenarios. It not only focuses on whether the target can be "kept," but more importantly, examines whether its identity identifier (ID) can be "maintained." A high maintenance rate means that the system can effectively resist environmental interference, target occlusion, and intersection, providing continuous and reliable target trajectories, which is crucial for downstream tasks (such as behavior analysis and situation prediction). Attached Figure Description

[0070] Figure 1 To generate a fused flight path map by associating data;

[0071] Figure 2 A diagram illustrating the judgment of correct and incorrect target tracking (track maintenance);

[0072] Figure 3 This is a diagram illustrating the situation when the fusion track is interrupted.

[0073] Figure 4 Flowchart for calculating the track maintenance ratio within a single time window;

[0074] Figure 5 This diagram illustrates the inclusion of only fused tracks that can be determined to correspond with the true value in the maintenance rate calculation.

[0075] Figure 6 This is a schematic diagram when the actual flight path is stationary;

[0076] Figure 7 This is an example scenario diagram illustrating the difficulty in evaluating the confidence level of fused tracks.

[0077] Figure 8 This is a schematic diagram of the scene. The area shown consistently contains only four targets, with no targets entering or leaving the area. As time progresses, at time t0, the four fused tracks gradually break down or track incorrect targets, indicating track maintenance failure, which is marked with a red cross.

[0078] Figure 9 As a typical scenario illustration, targets are constantly entering and exiting within the detection range (rectangle) of the airborne platform, and the fusion track maintenance of old or new targets is constantly failing.

[0079] Figure 10 This is a schematic diagram of the maintenance ratio calculation method proposed in this invention;

[0080] Figure 11 Flowchart for calculating track sustaining rate w;

[0081] Figure 12 Flowchart for calculating the track maintenance ratio w(nT);

[0082] Figure 13 This is a diagram illustrating data association.

[0083] Figure 14 This is a flowchart of the method of the present invention;

[0084] Figure 15 This is a block diagram of the device of the present invention;

[0085] Figure 16 This is a schematic diagram of the structure of the first electronic device of the present invention;

[0086] Figure 17 This is a schematic diagram of the application environment structure of the first electronic device of the present invention;

[0087] Figure 18 This is a schematic diagram of the structure of the second electronic device of the present invention.

[0088] Figure label:

[0089] A - First electronic device;

[0090] B - A fusion track maintenance capability assessment device for observation scenarios on airborne platforms;

[0091] C-Data acquisition equipment;

[0092] D-Information display device;

[0093] 1000 - Second electronic device;

[0094] Ⅰ-Computational Unit;

[0095] II-ROM;

[0096] III-RAM;

[0097] N-bus;

[0098] V-Interface;

[0099] VI - Input Unit;

[0100] VII - Output Unit;

[0101] VIII - Storage medium;

[0102] IX - Communication Unit. Detailed Implementation

[0103] It should be noted that, in this application, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus.

[0104] In the absence of further restrictions, an element defined by the phrase "comprising a..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0105] The processor described in this invention is the control center of an electronic device. It can be a single processor or a collective term for multiple processing elements. For example, it can be one or more central processing units (CPUs), application-specific integrated circuits (ASICs), or one or more integrated circuits configured to implement embodiments of this invention, such as one or more digital signal processors (DSPs), or one or more field-programmable gate arrays (FPGAs).

[0106] Alternatively, the processor can perform various functions of the electronic device by running or executing software programs stored in memory and by calling data stored in memory.

[0107] In a specific implementation, as one example, the processor may include one or more CPUs. Each of these processors may be a single-core processor or a multi-core processor. Here, "processor" can refer to one or more devices, circuits, and / or processing cores for processing data (e.g., computer program instructions). Electronic devices may include servers, desktop computers, laptops, smartphones, tablets, embedded computers, etc., where the embedded computer includes vehicles and robots, etc.

[0108] The memory is used to store the software program that executes the solution of the present invention, and the execution is controlled by the processor. For specific implementation methods, please refer to the above method embodiments, which will not be repeated here.

[0109] It should be noted that the structure of the electronic device shown in the accompanying drawings of this invention does not constitute a limitation thereof. The actual knowledge structure recognition device may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0110] The above embodiments can be implemented, in whole or in part, by software, hardware (such as circuits), firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of the present invention are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0111] It should also be understood that the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. A and B can be singular or plural. Additionally, the character " / " in this article generally indicates an "or" relationship between the preceding and following related objects, but it can also represent an "and / or" relationship. Please refer to the context for a more accurate understanding.

[0112] In this invention, "at least one" means one or more, and "more than one" means two or more. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of a single item or a plurality of items. For example, at least one of a, b, or c can represent: a, b, c, ab, ac, bc, or abc, where a, b, and c can be a single item or multiple items.

[0113] It should also be understood that, in various embodiments of the present invention, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0114] In the several embodiments provided by this invention, it should be understood that the disclosed devices, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0115] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0116] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0117] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0118] When evaluating data association algorithms, the inventors discovered that existing indicator systems lack a method for comprehensively assessing overall capabilities. Initially, when fused tracks are first started, their reliability is high. However, over time, they may misalign or lose targets, meaning the accuracy of track maintenance decreases, leading to a gradual decline in track reliability. For example, a map might divide the map into multiple boxes, with the number within each box representing the target density of that area, in units such as "targets per 10,000 square kilometers" or "targets per square kilometer." The higher the target density of the area traversed by the fused track, the more likely it is to lose or misalign other targets, resulting in a decrease in reliability. Different fusion algorithms or systems exhibit different rates of track reliability decline; therefore, a method is needed to assess this rate of decline.

[0119] An example scenario where evaluating the confidence level of fused tracks is difficult is... Figure 7 The left side shows the points of interest from three radar sources, while the right side shows the four tracks formed by fusion. When the true value is unknown, it is difficult to assess the confidence level of these four tracks: Did they mistakenly follow other adjacent tracks along the way? Did they lose track of the target midway, creating a false target track when there were actually only three targets?

[0120] A common application scenario is as follows: After fusing a track, the target is confirmed as ship A through imagery and other means. Then, the target is continuously tracked by radar: newly detected radar location points are constantly associated with this fused track to update the target's latest position. However, over time, the confidence level of this track decreases: could it mistakenly associate the locations of other nearby fishing boats, cargo ships, or unmanned vessels with ship A, causing the fused track to gradually shift towards the positions of other vessels and become a tracking track for other ships?

[0121] Track confidence decreases over time; that is, the longer the tracking time, the more likely it is to mistrack or lose track of the original target. It is also related to target density and sensor positioning accuracy. The more numerous and denser the targets, and the lower the sensor positioning accuracy, the more likely it is to mistrack or lose track of the target.

[0122] The inventors, through research into the application scenario of space-based observation, discovered that the track maintenance rate can be used as an indicator to comprehensively reflect the algorithm's capabilities. The track maintenance rate is used to evaluate the accuracy of data association algorithms and can further be used to estimate the confidence of fused tracks, thereby improving the data association algorithm and increasing the accuracy of the algorithm's output tracks. Physically, the track maintenance rate reflects the percentage of time that the fused track maintains accurate tracking of the original target, thus demonstrating both the continuity and accuracy of tracking. From the perspective of scenario requirements, this indicator addresses core concerns, reflecting the system's ability to continuously and accurately track multiple targets. Once we calculate the track maintenance rate of a fusion algorithm under a certain target density and sensor scenario when the truth value is known, we can evaluate the change in confidence of each fused track over time in scenarios where the truth value is unknown. For example, a fused track initially has a confidence of 100%. If its maintenance rate is 90% / hour, then the confidence of that track will decrease to 90% after 1 hour and to 81% after 2 hours. This provides crucial information for multi-target surveillance systems: if the confidence level of a fused track falls below a threshold, it becomes unreliable, requiring alerting system operators. If the target is of high importance, such as a monitored ship, the fused track can re-identify the target's identity using imagery and other methods after the confidence level drops below the threshold. This involves pointing cameras and other sensors from the airborne platform at the latest location of the fused track to re-identify the target, thereby restoring the track's confidence level to 100%.

[0123] To achieve the above-mentioned technical effects, the present invention proposes the following key technical points:

[0124] Key point 1: A definition of a fusion track maintenance rate index is proposed, which can comprehensively evaluate the maintenance capability of fusion tracks.

[0125] Key point 2: Propose the calculation method and formula for the fused track sustaining rate; and realize the calculation of the fused track sustaining rate based on this indicator definition;

[0126] Key point 3 proposes the calculation process for the fusion track maintenance rate.

[0127] To make the above-mentioned features and effects of the present invention clearer and easier to understand, specific embodiments are described below in conjunction with the accompanying drawings. This specification discloses one or more embodiments incorporating the features of the present invention. The disclosed embodiments are merely illustrative. The scope of protection of the present invention is not limited to the disclosed embodiments, but is defined by the appended claims.

[0128] Indicator definition:

[0129] First, the definition of correct tracking is: the distance between the target fused track and the real track is always no greater than a threshold.

[0130] When calculating track maintenance rate, the distance between the fused track and the ground truth is calculated every unit of time. If the distance is greater than a threshold, it is judged as a tracking error and maintenance failure. Figure 2 As shown, this distance thresholding method can simultaneously detect maintenance failures caused by both target loss due to track interruption and misalignment with another target mid-journey.

[0131] When the merged track is interrupted, new probes will be generated to form a new track. At this time, the original merged track will drift further and further away from the real track (shown by the yellow line in the figure below), until it exceeds a threshold, at which point it is judged as a tracking error and maintenance failure, as shown below. Figure 3 As shown.

[0132] This invention proposes a fusion track maintenance rate as an indicator to evaluate the performance of data association. The fusion track maintenance rate is a performance metric that measures the long-term association stability of a multi-target tracking system, and its definition is as follows:

[0133] Fusion track maintenance rate (Abbreviation: Maintenance Rate): The proportion of targets that are correctly maintained (without being lost or having their identities switched) in the fused track set at the start time after a unit time interval.

[0134] The success rate of track maintenance (abbreviated as: maintenance rate) The percentage of targets that are still correctly tracked (without being lost or having their identities changed) in the fused track set starting at time 0 at time t:

[0135]

[0136] In the above formula, This represents the number of targets that have been batched and are being correctly tracked at the initial moment. Representing time t, The number of tracks that remain correctly tracked in the initial fused track.

[0137] In addition, the proportion of fused tracks correctly maintained within the defined time period [t1, t2] is... The percentage of targets that are still correctly tracked (not lost or switched identities) in the fused track set at initial time t1 at time t2:

[0138]

[0139] This means The number of times the track is maintained correctly. This means The number of times the track is maintained correctly.

[0140] yes Special forms:

[0141] w(t) = w(0, t)

[0142] This invention proposes track maintenance rate, an index representing the rate at which the proportion of correctly maintained fused tracks decreases over time, as a comprehensive performance parameter for the fusion algorithm and monitoring system, and proposes a calculation scheme. The track maintenance rate w is the derivative of the proportion of correctly maintained fused tracks w(t) with respect to time t, and can be approximately estimated by taking the t-th power:

[0143]

[0144] Maintain ratio This is the foundation for calculating the maintenance rate, and also the most challenging aspect. Figure 8 For example, ideally, there are 4 targets initially, represented by 4 arrows. As time passes, track maintenance fails, meaning it is interrupted or the wrong target is followed, indicated by a red cross. By time t3, only the fused track of one target is correct. Therefore:

[0145]

[0146] The above calculations are based on ideal conditions. The process involves calculating the track maintenance ratio, but in practice, it faces numerous problems that hinder its application. When, in the formula The calculation presents several challenges:

[0147] 1. A longer time interval t needs to be calculated to improve the accuracy of the maintenance rate, such as 1 hour or 10 hours. If only the change in the number of correctly tracked targets before and after 1 minute is calculated, the number of targets is likely not to have changed, resulting in an incorrect result of 100% maintenance rate, that is, it is difficult to determine the correct and appropriate time interval size;

[0148] 2. In airborne reconnaissance scenarios, the platform's field of view is limited, and various targets such as aircraft and ships constantly enter and exit the area, causing changes in the numerator and denominator of the track maintenance ratio formula. For example, if three ships are detected at the initial time 0, the corresponding formula would be... =3. Half an hour later, all three ships left the field of view and could no longer be detected. Simultaneously, four other ships entered the field of view. When evaluating the accuracy of data association, the data from these three ships was consistently correctly associated. =3. Assuming t = 10 hours, then =3 will overestimate the fusion track maintenance capability of the algorithm, mistakenly believing that the correlation has been continuously correct for 10 hours.

[0149] Figure 9 Here's an example: initially, there are two fused tracks. At time t1, one target leaves the rectangular field of view and becomes undetectable. This isn't due to the data association algorithm, so it can't be considered a maintenance failure or algorithmic error. Starting at time t2, new targets begin entering the region, indicated by dashed arrows. However, these newly entered targets were not present at the initial time... If the failure information of maintenance of each target cannot be reflected in the maintenance ratio and maintenance rate calculation, it will cause the maintenance rate result to be biased.

[0150] To address the aforementioned challenges, this invention proposes a method that calculates the fused track maintenance ratio within shorter time windows, then multiplies the maintenance ratios of adjacent windows together to obtain the maintenance ratio over a longer period. Finally, by taking the square root of the maintenance ratio with respect to the duration, a track maintenance rate index independent of time length is obtained, which can be used to evaluate the overall performance of data association.

[0151] by Figure 10 Taking an example, we will explain the maintenance ratio calculation method proposed in this invention.

[0152] At the initial time t0: there are two objectives, represented by solid arrows.

[0153] At time t1, one target leaves the area (gray arrow), and the maintenance ratio is still 100%.

[0154] Time t2: A new target enters the area (dashed arrow), bringing the total number of targets to two. After a period of time, tracking of one of the targets fails, at which point the overall tracking success rate drops to 50% (from t0 to t2).

[0155] Time t3: Two more new targets enter the area, bringing the total number of targets to three. However, tracking of one of these targets fails after a period of time, resulting in a maintenance rate of 2 / 3 within the corresponding time window. Multiplying this number by the maintenance rate of 50% up to time t2 gives an overall maintenance rate of 33% (from time t0 to t3).

[0156] The specific process of this invention is as follows: Figure 11 As shown:

[0157] The calculation process for track maintenance rate w is shown in the figure.

[0158] Step 1: Read the fused track and ground truth track from the database or file into memory;

[0159] This step reads the fused track data to be evaluated and the corresponding ground truth track data from the predetermined data storage medium into the computer memory, completing the data initialization of the evaluation system.

[0160] After reading the data, it is parsed according to a predefined data pattern to extract key fields. The key fields of the fused track data include: track identifier, timestamp, and target spatial coordinates.

[0161] The key fields of the true trajectory data include: true target identifier, timestamp, and target spatial coordinates.

[0162] The parsed data is transformed into structured data objects in memory, such as arrays, linked lists, or hash tables, for efficient access and processing later.

[0163] Step 2: Calculate the track maintenance ratio w(nT) over a time period of total duration nT.

[0164] Calculate the track maintenance ratio for a continuous nT time period. Where n≥1, and T is the length of a single time window, for example, 10 minutes.

[0165] Step 3: Calculate the track maintenance rate w based on the track maintenance ratio w(nT) from the previous step;

[0166] Convert the track maintenance ratio over nT duration into track maintenance rate per unit time:

[0167] 4. Output result: Track maintenance rate w.

[0168] Calculation process for track maintenance ratio w(nT)

[0169] Step 2 is the main step, which calculates the track maintenance ratio within a duration of nT. The process is as follows Figure 12 As shown in the diagram. The duration T of a single time window depends on the detection field of view of one or more airborne platforms and the movement speed of the monitored targets, ensuring that most targets do not cross the field of view within duration T. For example, if the airborne platform's field of view is a circle with a radius of 10 kilometers, and the target is a ship with a speed not exceeding 10 meters per second, then the duration T of a single time window should not exceed 10000 / 10 = 1000 seconds, and can be set to several hundred seconds. If T is set too short, there will not be enough point data and movement length within a single time window to maintain the proportionality judgment; if T is set too long, most targets will have already moved out of the field of view.

[0170] Step 2.1: Initialize the track maintenance ratio w(0,0)=100%

[0171] Step 2.2: Initialize the start and end times of the first time window: t1=0, t2=T

[0172] Step 2.3: Calculate the fusion track maintenance ratio w(t1, t2) within a single window of duration T, as detailed in the following description.

[0173] Step 2.4: Iteratively update the value of w(t2), w(t2) = w(0, t2) = w(0, t1) * w(t1, t2)

[0174] Step 2.5: After calculating the track maintenance ratio within the current time window, continue sliding the window forward by T: t1=t2, t2=t1+T

[0175] Step 2.6: After calculating all time windows, output the final track maintenance ratio. .

[0176] 2.3 Calculation process for track maintenance ratio w(t1, t2) within a single time window

[0177] In step 2.3, the calculation process for the track maintenance ratio w(t1, t2) within a single time window ranging from t1 to t2 is as follows: Figure 4 As shown.

[0178] 2.3.1 Variable Initialization

[0179] The numerator N2 and denominator N1 in the formula for calculating the track maintenance ratio are initialized to 0. Here, N1 represents the number of fused tracks that are stably tracked at the start time t1 of the current time window, and N2 represents the number of fused tracks that are successfully maintained at the end time t2 of the current time window.

[0180] 2.3.2 Determine whether the fused track is stably tracked for each individual step.

[0181] In this step, fused tracks are read one by one from the storage device (each fused track includes multiple point track data), and it is determined whether each fused track entered a stable tracking state before time t1. The determination process is as follows: three conditions are checked sequentially:

[0182] First, calculate whether the number of points on the track up to time t1 is not less than a threshold n (which can be 3 or other positive integers). If the number of points is too small (e.g., only 1 point), the fused track may be a false detection point, not corresponding to any target. Furthermore, only after accumulating multiple points can it be confirmed that the fused track has stably tracked a real target: that is, whether all n fused points correspond to the detection data of the same real target. The fused track is a curve formed by connecting multiple point data using a fusion algorithm and then filtering (reducing the positioning error of each point). During the fusion process, the initial state of the track is prone to errors: with only 1 point, speed and heading cannot be calculated, resulting in significant uncertainty. After accumulating two points, speed and heading can be calculated based on the distance and time difference between the two points, which is relatively reliable. When the fused track accumulates three points, the third point can verify the reliability of the speed and heading calculated from the first two points, filtering out some erroneous correlations. If multiple consecutive track points in the fused track correspond to the same real track point, then a stable tracking state is entered. This state can be included in the denominator of the track maintenance ratio calculation. In subsequent moments, it is continuously detected whether this stable tracking state is broken and how long it takes to be broken, i.e., whether the target has been lost or the wrong target has been followed.

[0183] The diagram of data association is as follows Figure 13 As shown, at time k, three detection point data points Z1, Z2, and Z3 are received from radar and other equipment. The position at time k is calculated from a fused track i using speed, heading, and time difference, as shown by the red circle in the figure. Then, the distance between the predicted position and the three detection point data points is calculated, and the probability is calculated for each. In this figure, Z2 is closer to the predicted position; therefore, Z2 is associated with track i, meaning that the detection point data point Z2 is considered to be from the detection result of the target corresponding to track i. When track i has only one or two points, the calculated track, speed, and acceleration are inaccurate, resulting in low reliability.

[0184] Then determine whether to track a certain true track: before time t1, whether the last n points of the merged track all come from the same target M. i The measurement point data. (Illustrated as follows) Figure 5 As shown, the black line represents the fused track, and the yellow line represents the original track. In the left figure, the last three blue dots of the fused track are associated with different detection points of the original track, and no stable association is established with a fixed single original track. Therefore, this track is not included in the subsequent maintenance rate calculation until a target can be identified for tracking.

[0185] Finally, the target M was retrieved. iThe last detected data point (detection and positioning data before fusion, fields including: detection time, longitude, latitude, error matrix, etc.) is used to determine if the timestamp of the point is greater than or equal to t1. If the last detected data point of the target is less than the starting point t1 of the current time window, it means that the airborne platform's radar and other equipment will no longer be able to detect the target M during the current time window and the time thereafter. i This means the target has left the field of view of the airborne platform. Therefore, within the current time window and thereafter, the fused track will not contain any new data points. This interruption in track fusion is due to limitations in the detection field of view, not a problem with the fusion algorithm itself. Therefore, within the current time window, this target will no longer be included in the numerator and denominator of the track maintenance ratio.

[0186] 2.3.3 Initial Track Number N1 Update

[0187] If the current fused track meets the three conditions of the previous step, then add it to the denominator N1 of the track maintenance ratio, that is, increase N1 by 1.

[0188] 2.3.4 Retrieving merged traces

[0189] Detection point trace: A radar mounted on an airborne or other platform scans the sea surface and reports multiple isolated bright spots at a certain moment. Each point contains time, distance, and azimuth information, which can be converted into latitude and longitude coordinates. This is the detection point trace. It only tells us "at this moment, there is a suspected target at a certain location," but we don't know what it is or whether it is real.

[0190] Fusion Tracks: The system strings together detection points from different times into continuous, smooth trajectories through data association and filtering algorithms. Multiple detection point data belonging to the same target (such as a ship) are strung together on the same curve. This is fusion track, which can not only clearly show how each ship "went in the past", but also "predict where it will go in the future" (because motion parameters such as speed, heading, and acceleration are calculated from multiple data).

[0191] Merged Tracks: Each point on the merged track is a merged track.

[0192] by Figure 14 For example, the purple, blue, and yellow dots represent probe track data, the red curve represents the fused track, and each dot on the red curve represents a fused track. Because track filtering reduces the positioning error in the probe track data, each fused track does not completely overlap with its corresponding probe track. During algorithm evaluation, the ground truth is known, and it is also known which real target's positioning result each probe track is simulated from.

[0193] First, determine whether there are any fusion points in the current fusion track within the time window [t1, t2].

[0194] Case 1: If no fused point falls within the time window, determine whether maintenance has failed. The method for determining maintenance failure is to check if three conditions are met simultaneously: 1. The timestamp of the last point of the fused track is less than t2; 2. After the fused track terminates, the tracked target still has m (positive integer, such as 3) or more detection point data; 3. The maximum distance between the last point of the fused track and the actual location corresponding to the subsequent detection points of the tracked target is greater than N (e.g., 4) times the positioning error.

[0195] If all three conditions are met, the track maintenance is considered a failure; otherwise, it is considered a successful maintenance.

[0196] Explanation of condition 3: When the actual flight path (as shown by the yellow line in the figure) is almost stationary, even if multiple consecutive data points (such as...) are detected... Figure 6 The black circle) failed to be associated with the fused track (as shown by the black line in the figure below). The distance between the fused track and the real track may also remain within the threshold, which can still be judged as correct tracking and successful maintenance.

[0197] Scenario 2: If a merged point falls within this time window, retrieve the last merged point data (pos) with timestamp ≤ t2 in the current merged track. 融合_i_k To determine the subsequent track maintenance status, see section 2.3.5.

[0198] 2.3.5 Track Maintenance Status Judgment

[0199] Retrieve the true track value of the tracked target from the storage device, and find the pos by matching the timestamps. 融合_i_k For the target's true location at the corresponding moment, calculate whether the distance between the fused point and the true location does not exceed N (e.g., 4) times the positioning error. If the distance between the fused point and the true location does not exceed N (e.g., 4) times the positioning error, it is considered a successful maintenance; otherwise, it is considered a failed maintenance.

[0200] 2.3.6 N2 numerical update

[0201] If the previous steps determine that the maintenance was successful, increment N2 by 1; otherwise, the value of N2 remains unchanged.

[0202] 2.3.6 Calculate the maintenance ratio

[0203] After all merged tracks have been traversed, calculate the track maintenance ratio within the current time window:

[0204] w(t1, t2) = N2 / N1

[0205] The following are system embodiments corresponding to the above method embodiments. This embodiment can be implemented in conjunction with the above embodiments. The relevant technical details mentioned in the above embodiments are still valid in this embodiment, and will not be repeated here to reduce repetition. Accordingly, the relevant technical details mentioned in this embodiment can also be applied to the above embodiments.

[0206] like Figure 15 As shown, this invention also proposes a fusion track maintenance capability assessment device for airborne platform observation scenarios, comprising:

[0207] Module 1: Obtain fused track data and corresponding ground truth track data based on the fusion algorithm; the key fields of the fused track data include: track identifier, timestamp, and target spatial coordinates; the key fields of the ground truth track data include: real target identifier, timestamp, and target spatial coordinates.

[0208] Module 2: Based on the fused track data and the corresponding ground truth track data, calculate the track maintenance ratio for n consecutive time periods of T, and multiply the n track maintenance ratios together to obtain the track maintenance ratio for a total duration of n time periods of T. , where n is a positive integer and T is the duration of a single time window;

[0209] Module 3, Based on Track Maintenance Ratio Taking the square root of the total duration yields the track maintenance rate w of the fusion algorithm.

[0210] The aforementioned fusion track maintenance capability assessment device for airborne platform observation scenarios, wherein module 2 includes:

[0211] Module 2.1: Initialize track maintenance ratio w(0,0)=100%;

[0212] Module 2.2: Initialize the start and end times of the first time window, t1=0, t2=T;

[0213] Module 2.3: Calculate the fusion track maintenance ratio w(t1, t2) within a single window of duration T;

[0214] Module 2.4: Iteratively update the value of w(t2), w(t2) = w(0, t2) = w(0, t1) * w(t1, t2);

[0215] Module 2.5: After calculating the track maintenance ratio within the current time window, determine whether t2 equals nT. If so, save the current value of w(t2) as the track maintenance ratio. Otherwise, continue sliding the window forward by T: t1=t2, t2=t1+T.

[0216] The aforementioned fusion track maintenance capability assessment device for airborne platform observation scenarios, wherein module 2.3 includes:

[0217] Module 2.3.1 initializes the numerator N2 and denominator N1 in the formula for calculating the track maintenance ratio to 0 respectively; where N1 represents the number of fused tracks that are stably tracked at the start time t1 of the current time window, and N2 represents the number of fused tracks that are successfully maintained at the end time t2 of the current time window.

[0218] Module 2.3.2: Each fused track includes multiple point data points. The fused tracks are read one by one to determine whether each track entered a stable tracking state before time t1. The determination process includes sequentially checking whether three conditions are met:

[0219] Condition 1: Calculate whether the number of points on the track up to time t1 is not less than a specified threshold.

[0220] Condition 2: Before time t1, do all the last n points of the merged track originate from the same target M? i Measurement point data;

[0221] Condition 3, retrieval target M i In the last detection data point in the detection and positioning data (the detection and positioning data before fusion), determine whether the timestamp of this point is greater than or equal to t1;

[0222] If these three conditions are met, then module 2.3.3 will be executed;

[0223] Module 2.3.3 adds the fused track that meets these three conditions to the denominator N1 of the track maintenance ratio;

[0224] Module 2.3.4 analyzes whether there are fused points in the current fused track within the time window [t1, t2]. If no fused points fall within this time window, it determines whether the timestamp of the last point of the current fused track is less than t2, whether the tracked target has more than a preset number of detection points after the current fused track terminates, and whether the maximum distance between the last point of the current fused track and the real position corresponding to the subsequent detection points of the tracked target is greater than a preset multiple of the positioning error. If all the judgment results are yes, the track maintenance fails; otherwise, it is determined to be maintained successfully, and module 2.3.6 is executed.

[0225] Module 2.3.5: If there is a fusion point trace in the current fused track within the time window [t1, t2], then retrieve the last fusion point trace data pos in the current fused track with timestamp ≤ t2. 融合_i_k Based on the true value of the tracked target's trajectory, the pos is found through timestamp matching.融合_i_k The target's true position at the corresponding moment is calculated. The distance between the fused point and the true position is calculated to be no more than a preset multiple of the positioning error. If so, the track is considered successfully maintained and module 2.3.6 is executed; otherwise, the track maintenance fails.

[0226] In module 2.3.6, increase N2 by 1;

[0227] Module 2.3.6, after all merged tracks have been traversed, calculates the track maintenance rate within the current time window: .

[0228] like Figure 16 As shown, in another embodiment of the present invention, a first electronic device A is also proposed, which includes the aforementioned fusion track maintenance capability assessment device for airborne platform observation scenarios.

[0229] like Figure 17 As shown, the first electronic device A can also be connected to the data acquisition device C and the information display device D through a wired or wireless information transmission scheme. The data acquisition device C is used to collect fused track data and the corresponding true track data, and the information display device D is used to display the track maintenance rate obtained by the present invention.

[0230] Information display device D can process and organize the data output by the first electronic device A based on an information display mechanism to improve the readability of the data. This information display mechanism can be manually preset, for example, visualizing the data output by the first electronic device A. It can present the user with specified key information, such as track maintenance rate, based on user-defined display parameters and / or attributes. Display parameters could be, for example, the data range, and display attributes could be, for example, the font, color, and whether scrolling is enabled. This allows the user to access this information more promptly without needing to navigate to secondary pages or scroll through pages, saving user effort. Alternatively, this information display mechanism can be an artificial intelligence (AI) display model, which can learn the user's key information interests based on previous usage habits, such as viewing time, number of clicks, and number of edits, and then automatically present the user with rich and necessary key information.

[0231] The present invention also provides a computer program product, which includes a computer program that can be stored on a readable storage medium. When the computer program is executed by a processor, the computer is able to execute the fusion track maintenance capability assessment method for airborne platform observation scenarios provided by the above methods.

[0232] In another embodiment, the present invention also proposes a storage medium VIII for storing a computer program that executes the fusion track maintenance capability assessment method for airborne platform observation scenarios. It should be understood that the storage medium in the embodiments of the present invention can be volatile memory or non-volatile memory, or may include both. The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory can be random access memory (RAM), which serves as an external cache. By way of example, but not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate synchronous DRAM (DDR SDRAM), enhanced synchronous DRAM (ESDRAM), synchronous linked DRAM (SLDRAM), and direct rambus RAM (DR RAM).

[0233] Figure 18 A schematic block diagram of a second electronic device 1000 that can be used to implement embodiments of the present invention is shown. The second electronic device 1000 is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The second electronic device 1000 can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein. The second electronic device 1000 may be the same as or different from the first electronic device A.

[0234] The second electronic device 1000 includes a computing unit I, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory II (ROM) or a computer program loaded from storage medium VIII into random access memory (RAM) III. The RAM III may also store various programs and data required for the operation of the device 1000. The computing unit I, ROM II, and RAM III are interconnected via bus IV. An input / output (I / O) interface V is also connected to bus IV.

[0235] Multiple components in the second electronic device 1000 are connected to I / O interface V, including: input unit VI, such as a keyboard, mouse, etc.; output unit VII, such as various types of displays, speakers, etc.; storage medium VIII, such as a disk, optical disk, etc.; and communication unit IX, such as a network card, modem, wireless transceiver, etc. Communication unit IX allows the second electronic device 1000 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0236] The computing unit I can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of computing unit I include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit I performs the various methods and processes described above, such as method steps S1-S3. For example, in some embodiments, the methods can be implemented as computer software programs tangibly contained in a machine-readable medium, such as storage medium VIII. In some embodiments, part or all of the computer program can be loaded and / or installed on device 1000 via ROM II and / or communication unit IX. When the computer program is loaded into RAM III and executed by computing unit I, one or more steps of the methods described above can be performed. Alternatively, in other embodiments, computing unit I can be configured to perform methods by any other suitable means (e.g., by means of firmware).

[0237] Although embodiments of the present invention have been disclosed above, they are not limited to the applications listed in the specification and embodiments. They can be applied to various fields suitable for the present invention. For those skilled in the art, other modifications can be easily made. Therefore, without departing from the general concept defined by the claims and their equivalents, the present invention is not limited to the specific details and illustrations shown and described herein.

Claims

1. A method for evaluating fusion track maintenance capability in airborne platform observation scenarios, characterized in that, include: Step 1: Obtain the fused track data and corresponding ground truth track data obtained based on the fusion algorithm; The key fields of the fused track data include: track identifier, timestamp, and target spatial coordinates; the key fields of the ground truth track data include: real target identifier, timestamp, and target spatial coordinates. Step 2: Based on the fused track data and the corresponding ground truth track data, calculate the track maintenance ratio for n consecutive time periods of T, and multiply the n track maintenance ratios together to obtain the track maintenance ratio for a total duration of n time periods of T. , where n is a positive integer and T is the duration of a single time window; Step 3: Based on the track maintenance ratio Taking the square root of the total duration yields the track maintenance rate w of the fusion algorithm.

2. The method for evaluating fusion track maintenance capability for airborne platform observation scenarios as described in claim 1, characterized in that, Step 2 includes: Step 2.1: Initialize the track maintenance ratio w(0,0) = 100%; Step 2.2: Initialize the start and end times of the first time window, t1=0, t2=T; Step 2.3: Calculate the fusion track maintenance ratio w(t1, t2) within a single window of duration T; Step 2.4: Iteratively update the value of w(t2), w(t2) = w(0, t2) = w(0, t1) * w(t1, t2); Step 2.5: After calculating the track maintenance ratio within the current time window, determine whether t2 equals nT. If so, save the current value of w(t2) as the track maintenance ratio. Otherwise, continue sliding the window forward by T: t1=t2, t2=t1+T.

3. The method for evaluating fusion track maintenance capability for airborne platform observation scenarios as described in claim 2, characterized in that, Step 2.3 includes: Step 2.3.1: Initialize the numerator N2 and denominator N1 in the formula for calculating the track maintenance ratio to 0 respectively; where N1 represents the number of fused tracks that are stably tracked at the start time t1 of the current time window, and N2 represents the number of fused tracks that are successfully maintained at the end time t2 of the current time window. Step 2.3.2: Each fused track includes multiple point data points. The fused tracks are read one by one to determine whether each track entered a stable tracking state before time t1. The determination process includes sequentially checking whether three conditions are met: Condition 1: Calculate whether the number of points on the track up to time t1 is not less than a specified threshold. Condition 2: Before time t1, do all the last n points of the merged track originate from the same target M? i Measurement point data; Condition 3, retrieval target M i In the last detection data point in the detection and positioning data (the detection and positioning data before fusion), determine whether the timestamp of this point is greater than or equal to t1; If these three conditions are met, proceed to step 2.3.3; Step 2.3.3: Add the fused track that meets the three conditions to the denominator N1 of the track maintenance ratio; Step 2.3.4: Analyze whether there are any fused points in the current fused track within the time window [t1, t2]. If no fused points fall within this time window, determine whether the timestamp of the last point of the current fused track is less than t2, whether the tracked target has more than a preset number of detection point data after the current fused track terminates, and whether the maximum distance between the last point of the current fused track and the real position corresponding to the subsequent detection point of the tracked target is greater than a preset multiple of the positioning error. If all the judgment results are yes, the track maintenance fails; otherwise, it is determined to be maintained successfully, and step 2.3.6 is executed. Step 2.3.5: If there is a fusion point trace in the current fusion track within the time window [t1, t2], then retrieve the last fusion point trace data pos with timestamp ≤ t2 in the current fusion track. 融合_i_k Based on the true value of the tracked target's trajectory, the pos is found through timestamp matching. 融合_i_k The target's true position at the corresponding moment is calculated. The distance between the fused point and the true position is calculated to be no more than a preset multiple of the positioning error. If so, the track is considered successfully maintained and step 2.3.6 is executed; otherwise, the track maintenance fails. Step 2.3.6, increase N2 by 1; Step 2.3.7: After all merged tracks have been traversed, calculate the track maintenance rate within the current time window: .

4. A fusion track maintenance capability assessment device for airborne platform observation scenarios, characterized in that, include: Module 1: Obtain fused track data and corresponding ground truth track data based on the fusion algorithm; The key fields of the fused track data include: track identifier, timestamp, and target spatial coordinates; the key fields of the ground truth track data include: real target identifier, timestamp, and target spatial coordinates. Module 2: Based on the fused track data and the corresponding ground truth track data, calculate the track maintenance ratio for n consecutive time periods of T, and multiply the n track maintenance ratios together to obtain the track maintenance ratio for a total duration of n time periods of T. , where n is a positive integer and T is the duration of a single time window; Module 3, Based on Track Maintenance Ratio Taking the square root of the total duration yields the track maintenance rate w of the fusion algorithm.

5. The fusion track maintenance capability assessment device for airborne platform observation scenarios as described in claim 4, characterized in that, Module 2 includes: Module 2.1: Initialize track maintenance ratio w(0,0)=100%; Module 2.2: Initialize the start and end times of the first time window, t1=0, t2=T; Module 2.3: Calculate the fusion track maintenance ratio w(t1, t2) within a single window of duration T; Module 2.4: Iteratively update the value of w(t2), w(t2) = w(0, t2) = w(0, t1) * w(t1, t2); Module 2.5: After calculating the track maintenance ratio within the current time window, determine whether t2 equals nT. If so, save the current value of w(t2) as the track maintenance ratio. Otherwise, continue sliding the window forward by T: t1=t2, t2=t1+T.

6. The fusion track maintenance capability assessment device for airborne platform observation scenarios as described in claim 5, characterized in that, Module 2.3 includes: Module 2.3.1 initializes the numerator N2 and denominator N1 in the formula for calculating the track maintenance ratio to 0 respectively; where N1 represents the number of fused tracks that are stably tracked at the start time t1 of the current time window, and N2 represents the number of fused tracks that are successfully maintained at the end time t2 of the current time window. Module 2.3.2: Each fused track includes multiple point data points. The fused tracks are read one by one to determine whether each track entered a stable tracking state before time t1. The determination process includes sequentially checking whether three conditions are met: Condition 1: Calculate whether the number of points on the track up to time t1 is not less than a specified threshold. Condition 2: Before time t1, do all the last n points of the merged track originate from the same target M? i Measurement point data; Condition 3, retrieval target M i In the last detection data point in the detection and positioning data (the detection and positioning data before fusion), determine whether the timestamp of this point is greater than or equal to t1; If these three conditions are met, then module 2.3.3 will be executed; Module 2.3.3 adds the fused track that meets these three conditions to the denominator N1 of the track maintenance ratio; Module 2.3.4 analyzes whether there are fused points in the current fused track within the time window [t1, t2]. If no fused points fall within this time window, it determines whether the timestamp of the last point of the current fused track is less than t2, whether the tracked target has more than a preset number of detection points after the current fused track terminates, and whether the maximum distance between the last point of the current fused track and the real position corresponding to the subsequent detection points of the tracked target is greater than a preset multiple of the positioning error. If all the judgment results are yes, the track maintenance fails; otherwise, it is determined to be maintained successfully, and module 2.3.6 is executed. Module 2.3.5: If there is a fusion point trace in the current fused track within the time window [t1, t2], then retrieve the last fusion point trace data pos in the current fused track with a timestamp ≤ t2. 融合_i_k Based on the true value of the tracked target's trajectory, the pos is found through timestamp matching. 融合_i_k The target's true position at the corresponding moment is calculated. The distance between the fused point and the true position is calculated to be no more than a preset multiple of the positioning error. If so, the track is considered successfully maintained and module 2.3.6 is executed; otherwise, the track maintenance fails. In module 2.3.6, increase N2 by 1; Module 2.3.7, after all merged tracks have been traversed, calculates the track maintenance rate within the current time window: .

7. A client for evaluating the fusion track maintenance capability of any one of the airborne platform observation scenarios described in claims 4-6.

8. An electronic device, characterized in that, The device includes a fusion track maintenance capability assessment device for an airborne platform observation scenario as described in claims 4-6. The electronic device may be connected to an information display device, which is used to display the track maintenance rate using user-set display parameters, attributes, or through an artificial intelligence model.

9. A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the fusion track maintenance capability assessment method for an airborne platform observation scenario as described in any one of claims 1-3.

10. A computer program product, comprising a computer program, characterized in that, When executed by a processor, the computer program implements the steps of the fusion track maintenance capability assessment method for airborne platform observation scenarios as described in any of claims 1-3.

Images