Single station radar and photoelectric cooperative multi-target detection and tracking method and system

Abstract

The application relates to a single-station radar and photoelectricity cooperative multi-target detection tracking method and system; the method comprises the following steps: in an unattended mode, judging whether multiple detection targets are formation targets based on radar detection information; if the multiple detection targets are formation targets, guiding a photoelectricity detector to the formation targets for detection, and exiting the unattended mode; if the multiple detection targets are not formation targets, establishing a threat level list of the multiple detection targets according to the radar detection information, and guiding the photoelectricity detector to track and investigate the targets in sequence according to the threat level. The technical scheme of the application proposes a radar and photoelectricity single-station cooperative detection tracking technology based on target threat level, can realize day and night monitoring, tracking and investigation on suspicious personnel and vehicles on an open land in an unattended mode; when multiple targets are detected, whether the multiple targets are formation targets is judged, threat level determination is further performed, and thus targeted treatment on different conditions is realized.

Description

Technical Field

[0001] This application relates to the field of border security monitoring, specifically to a method and system for multi-target detection and tracking using a single-station radar and optoelectronic coordination. Background Technology

[0002] The persistent high incidence of asymmetric threats along land and sea borders worldwide has driven the technological development of border security monitoring systems. Several common problems must be addressed in the application of border security monitoring systems: firstly, the uncertainty of potential threats; secondly, the wide geographical area requiring control; thirdly, the complexity and diversity of application scenarios; and finally, the difficulty in unifying internal resource allocation. Currently, the optimal solution is to "adopt a combination of physical, human, and technological defenses to build a controllable, reliable, and flexibly configurable integrated electronic information system for border security that meets the needs of different application scenarios."

[0003] In order to monitor the border situation around the clock, provide early warnings before threats are detected, collect evidence and issue alerts when suspicious targets are detected, and support relevant departments in activating threat handling mechanisms, the integrated electronic information system for border security needs to deploy detection equipment along the border line according to the terrain, topography, climate, probability of threat occurrence and other scenario factors to form a complete early warning and monitoring system.

[0004] Currently, the main carriers or forms of threats in border areas are still people and vehicles moving on the ground. Common detection methods include ground-based surveillance radar and long-range electro-optical detection equipment. Ground-based surveillance radar has superior low-interception performance and anti-jamming capabilities, can operate around the clock and in all weather conditions, and is mobile, flexible, and easy to deploy. However, its angle measurement accuracy is relatively poor, and it suffers from severe ground clutter and obstruction. Electro-optical detection equipment integrating multi-band detectors has all-weather, long-range detection capabilities, but its long-range field of view is narrow, resulting in low search efficiency.

[0005] Among related technologies, there is no mature technical solution that can achieve the detection and tracking of multiple targets. Summary of the Invention

[0006] To overcome, at least to some extent, the difficulty in detecting and tracking multiple targets in related technologies, this application provides a method and system for multi-target detection and tracking using monostation radar and electro-optical coordination.

[0007] According to a first aspect of the embodiments of this application, a multi-target detection and tracking method combining monostation radar and electro-optical coordination is provided, comprising:

[0008] In unattended mode, the radar detection information is used to determine whether multiple detected targets are formation targets;

[0009] If it is a formation target, guide the photoelectric detector to the formation target location for detection and exit the unattended mode;

[0010] If the target is not part of a formation, a threat level list is created for multiple targets based on radar detection information, and the photoelectric detectors are guided to track and collect evidence from the targets in sequence according to their threat levels.

[0011] Furthermore, the radar detection information includes: the target's direction of travel and its speed;

[0012] Multiple detection targets are considered to be in formation when they meet the following conditions:

[0013] The number of detected targets exceeds a preset threshold.

[0014] The distribution area of ​​multiple detection targets is smaller than the preset spatial range;

[0015] Multiple detection targets are traveling in the same direction; and / or,

[0016] The difference in travel speed between multiple detection targets is less than a preset speed range.

[0017] Furthermore, the radar detection information also includes: the motion state of the detected target, radar scattering information, and target classification information;

[0018] Creating a threat level list for multiple detected targets involves the following steps:

[0019] The threat level of the detected target is determined based on the target's motion state, radar scattering information, target classification information, and in conjunction with a pre-set threat target database.

[0020] Arrange the threat levels of multiple targets in order to create a threat level list.

[0021] Furthermore, the method also includes the following steps:

[0022] Determine the state vector of the detected target based on radar detection information;

[0023] Establish a nonlinear model based on state vectors;

[0024] By introducing bivariate discrete random variables into the nonlinear model, an improved nonlinear model is obtained.

[0025] State update equations and measurement update equations are established based on an improved nonlinear model;

[0026] Incomplete measurement criteria are determined based on state update equations and measurement update equations.

[0027] The measurement vector is updated based on the incomplete measurement criterion.

[0028] Furthermore, the nonlinear state transition model is as follows:

[0029]

[0030] Where Φ(k+1|k) is the target nonlinear state transition matrix, X(k)∈ Let Γ(k) be the target state vector at time k. Let u(k) be the process noise distribution matrix at time k, and let u(k) be the process noise vector ∈ It is Gaussian white noise with zero mean and variance Q(k); Z(k)∈ H(k) is the system observation vector; H(k) is the measurement equation, and the measurement noise vector v(k) ∈ It is Gaussian white noise with zero mean and variance R(k).

[0031] Furthermore, a binary discrete random variable g(k) is introduced, where g(k) = 1 indicates normal measurement and g(k) = 0 indicates incomplete measurement; then the measurement equation in the nonlinear state transition model is: Z(k) = g(k)H(k)X(k) + v(k);

[0032] Measurement noise is expressed as:

[0033]

[0034] In the formula, g(k) = 0 corresponds to the limit value of δ, that is, δ → ∞.

[0035] Furthermore, the state update equation is:

[0036]

[0037] P(k+1 / k)=Φ(k)P(k / k)Φ T (k)+Γ(k)Q(k)Γ T (k);

[0038] The measurement update equation is:

[0039]

[0040] K(k+1)=P(k+1 / k)H T (k+1)[H(k+1)P(k+1 / k)H T (k+1)+R(k+1)] -1 ;

[0041]

[0042] P(k+1 / k+1)=[IK(k+1)H(k+1)]P(k+1 / k).

[0043] Furthermore, the predicted residual of the measurement vector at time k+1 As a criterion for incomplete measurement:

[0044]

[0045] In the formula, C is the residual detection threshold; when the above formula is true, g(k) = 1, indicating that no incomplete measurement has occurred; when the above formula is false, g(k) = 0, indicating that incomplete measurement has occurred.

[0046] Furthermore, the measurement update equation is expressed as:

[0047]

[0048]

[0049] According to a second aspect of the embodiments of this application, a multi-target detection and tracking system combining single-station radar and optoelectronic coordination is provided, comprising: radar, photoelectric detector, and fusion processing module;

[0050] The radar is used to perform dynamic scanning and acquire information on the range, azimuth, speed, heading, and classification of the detected target.

[0051] The photodetector is used to identify and confirm the designated detection target;

[0052] The fusion processing module is used to determine whether multiple detection targets are in formation based on radar detection information in unattended mode. If they are in formation, the module guides the photoelectric detector to the formation target for detection and exits the unattended mode. If they are not in formation, the module establishes a threat level list for multiple detection targets based on radar detection information and guides the photoelectric detector to track and collect evidence from the targets in sequence according to their threat levels.

[0053] The technical solutions provided by the embodiments of this application have the following beneficial effects:

[0054] This invention proposes a radar-electro-optical single-station collaborative detection and tracking technology based on target threat levels. In unattended mode, it can monitor, track, and verify suspicious personnel and vehicles on the ground in open areas day and night. When multiple targets are detected, it determines whether they are part of a formation and then assesses their threat level, enabling targeted handling of different situations. This invention possesses multi-sensor single-station collaborative detection and tracking capabilities and can perform single-station intelligence fusion processing. This invention significantly reduces false alarms in border security monitoring systems, alleviates the workload of border personnel, and allows them to focus on high-value threat targets, providing both unattended front-end operation and remote back-end monitoring capabilities.

[0055] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Attached Figure Description

[0056] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0057] Figure 1 This is a schematic diagram illustrating the composition of integrated processing and display control software according to an exemplary embodiment.

[0058] Figure 2 This is a flowchart illustrating a multi-target detection and tracking method using a single-station radar and electro-optical coordination according to an exemplary embodiment.

[0059] Figure 3 This is a flowchart illustrating a multi-target cooperative detection and tracking process according to an exemplary embodiment.

[0060] Figure 4 This is a schematic diagram of an IMM algorithm flow according to an exemplary embodiment.

[0061] Figure 5 This is a schematic diagram of an MHT algorithm logic according to an exemplary embodiment. Detailed Implementation

[0062] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of methods and systems consistent with some aspects of this application as detailed in the appended claims.

[0063] This application addresses the challenges posed by long border lines and weak infrastructure in border areas by proposing an unmanned border high-point monitoring system. First, the system composition and subsystem design are presented. Then, the key technologies for engineering implementation are outlined: radar-electro-optical single-station collaborative detection and tracking technology, and a multi-target tracking algorithm under incomplete measurement conditions. The implementation and demonstration tests of the prototype system demonstrate the scientific rationality of the system design, achieving the tactical and technical indicators and unmanned operation capabilities for high-point monitoring.

[0064] In embodiments of this invention, a single-station integrated design of ground surveillance radar and optoelectronic equipment is adopted. The radar performs wide-field-of-view dynamic scanning, and then transmits the target distance and coarse azimuth information acquired by the radar to the optoelectronic equipment turntable in real time, driving the optoelectronic equipment to align with the target for identification and confirmation. This significantly improves the target detection and identification capabilities. Therefore, the unmanned border defense high-point monitoring system uses radar-optoelectronic collaborative detection as its basic means. The system mainly consists of subsystems such as ground surveillance radar, optoelectronic detectors, integrated power supply, and integrated processing and display software.

[0065] The following section provides an expanded explanation of the solution proposed in this application, using specific application scenarios as examples.

[0066] (1) Radar selection

[0067] Factors for selecting ground surveillance radar include technical and tactical indicators such as operating range, detection accuracy, and multi-target capability. Considering that the radar should be unattended for long periods of time in the field, it should have low transmission power, low overall power consumption, few mechanical servo mechanisms, and strong adaptability to harsh environments such as high temperature and sandstorms.

[0068] This solution selects a highly integrated, maintenance-free, tri-proof Ku-band frequency-modulated continuous wave ground surveillance radar. The radar employs an all-solid-state electronic scanning technology, with a peak transmit power of less than 5W. It eliminates components with limited lifespan, such as rotating hinges, achieving an IP66 protection rating. The equipment is simple, and maintenance and operating costs are low throughout its lifespan. The radar features horizontal electronic scanning, with each array covering 110° horizontally, allowing for rapid expansion of auxiliary arrays. Simultaneously, the radar has a wide elevation beam, capable of detecting vertically moving targets in mountainous areas and low-altitude, slow-moving small aircraft.

[0069] (2) Optoelectronic Design

[0070] The design of photoelectric detectors for border surveillance mainly considers factors such as operating range, field of view, image quality, device lifespan, cost, and adaptability to field environments. Ground surveillance radar and photoelectric detectors adopt a single-station integrated design, with the photoelectric detector employing a dual-channel configuration of a visible light detector and a long-life cooled infrared detector.

[0071] The detector turntable is designed with a low-drag T-shaped frame and a servo stabilization device. It exhibits excellent resonant frequency between the turntable's azimuth / pitch axes and the frame, superior azimuth / pitch axis rotation accuracy, overall stability, and tracking accuracy. The turntable can capture clear and stable images even under wind speeds below level 10 and high-point swaying with a ±12° period of 4 seconds.

[0072] The detection range of the photoelectric detector is mainly limited by the infrared thermal imager. Therefore, a high-sensitivity MCT staring focal plane mid-wave cooled detector and a large focal length optical continuous zoom lens are used, enabling high-magnification optical continuous zoom and achieving a detection range of no less than 10km for people and no less than 20km for vehicles under line-of-sight conditions. The visible light camera uses a 1 / 1.8" CCD sensor, combined with a high-definition motorized zoom lens, to capture full HD high-resolution video within the zoom range, improving the recognition performance of the surveillance system.

[0073] (3) Integrated Energy

[0074] Based on system load conditions, and considering the abundant average sunshine hours and green resources such as wind power at the deployment location, a polycrystalline silicon photovoltaic power generation system and a vertical axis magnetic levitation wind power generation system are selected as the main energy sources, with diesel generators as backup power. A small number of energy storage battery packs are used for energy storage and regulation, forming a wind-solar-diesel-storage microgrid system as an energy solution for high-point monitoring in border areas. The system control logic is as follows:

[0075] (3.1) When there is sufficient sunlight / wind, the load is powered entirely by photovoltaics / wind turbines, and the surplus power generation is used to charge the energy storage system; (3.2) When there is no sunlight or wind at night or when the photovoltaic and wind turbines generate insufficient power, the energy storage system is used to power the load first; (3.3) When the energy storage system consumes power to the set capacity limit and sunlight / wind is still insufficient, the diesel generator is turned on to power the load, ensuring uninterrupted power supply to the load, and charging the energy storage system on the one hand.

[0076] Under various operating conditions, the system can effectively control the charging and discharging of the energy storage system and the switching on and off of the diesel engine through power monitoring, maintaining the system power balance in real time and meeting the load requirements.

[0077] (4) Integrated processing and display control

[0078] The integrated processing and display control software consists of three modules: radar display control, photoelectric display control, and integrated processing. Figure 1 As shown.

[0079] The integrated processing and display software can display the radar detection interface, photoelectric images, and comprehensive intelligence, and can control the detection parameters; it can perform detection, tracking, identification, and fusion processing of radar and optical target information; it can control the linkage between photoelectric equipment and radar, providing intelligent operation for operators and meeting the needs of unattended operation.

[0080] After radar detects a target, it can display it in the software interface as a dot or track. In secondary operations, information such as the target's speed, size, and distance can be indicated. Furthermore, filtering parameters can be modified to reduce clutter interference and false alarms, increasing detection accuracy. Maps can also be loaded, and alarm zones can be set based on the map, allowing observers to determine the target type and threat level based on terrain structure, thus achieving the purpose of issuing warnings.

[0081] The photoelectric detector uses its long-focus high-definition lens to display the environment and targets within its coverage area in high-definition and infrared images in the software. The software can be adapted to target observation in complex environmental and climatic conditions by setting photoelectric parameters, and can also manually control the turntable to align and identify the target based on its location.

[0082] The integrated processing and control software features three operating modes: single-target tracking, multi-target localization, and area alarm. In single-target tracking mode, it selectively and continuously tracks and collects evidence from a single target. In multi-target localization mode, it prioritizes targets based on their overall threat level and activates optoelectronic devices to sequentially collect evidence from relevant targets according to their threat level. The area alarm mode allows for the setting of alarm zones; when a target appears or crosses the warning zone, the software controls optoelectronic devices to continuously track and collect evidence from that target and trigger an alarm.

[0083] The unmanned border high-point monitoring system of the present invention is designed to monitor, track and verify suspicious personnel and vehicles on the ground in open areas day and night; it has the ability to conduct multi-sensor single-station collaborative detection and tracking, and can perform single-station intelligence fusion processing; it can be powered by various means and has the ability to be self-sufficient in energy.

[0084] (5) Other factors

[0085] Considering the Earth's curvature and terrain, border defense applications typically elevate surveillance equipment to fixed iron towers or mobile lifting towers to ensure visibility. The tower design should prioritize low structural wind resistance, strong wind capacity, and high wind resistance. Components should undergo shot blasting and zinc plating to extend their service life. Furthermore, construction conditions in border areas must be considered, and rapid deployment designs should be adopted whenever possible.

[0086] Communication systems are also a problem that border defense applications must face. Considering the long border, the huge construction period and cost, the communication system should be comprehensively planned to fully consider business needs, have high reliability and scalability, and in suitable areas, existing communication equipment should be fully integrated and utilized as much as possible.

[0087] To make the objectives, technical solutions, and advantages of this invention clearer, the key technologies of this invention will be further described in detail below with reference to the accompanying drawings.

[0088] Figure 2This is a flowchart illustrating a multi-target detection and tracking method using monostation radar and electro-optical coordination, according to an exemplary embodiment. The method may include the following steps:

[0089] Step S1: In unattended mode, determine whether multiple detected targets are formation targets based on radar detection information;

[0090] Step S2: If it is a formation target, guide the photoelectric detector to the formation target location for detection and exit the unattended mode;

[0091] Step S3: If the target is not in formation, a threat level list is created for multiple targets based on radar detection information, and the photoelectric detector is guided to track and collect evidence from the targets in sequence according to their threat levels.

[0092] The technical solution of this invention proposes a radar-electro-optical single-station collaborative detection and tracking technology based on target threat level. In unattended mode, it can monitor, track, and verify suspicious personnel and vehicles on the ground in open areas day and night. When multiple targets are detected, the threat level is determined by judging whether they are formation targets, thereby realizing targeted handling for different situations. The solution of this invention has multi-sensor single-station collaborative detection and tracking capabilities and can perform single-station intelligence fusion processing.

[0093] It should be understood that, although Figure 2 The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order in which these steps are executed, and they can be performed in other orders. Figure 2 At least some of the steps in the process may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.

[0094] To further elaborate on the technical solution of this application, the relevant technical concepts will first be explained.

[0095] Assuming the state vector of the detected object is position x and velocity v, the measured values ​​are obtained through radar detection, and the predicted values ​​are based on the x and v values ​​at one moment to predict the x and v values ​​at the next moment. Due to various errors, neither the measured nor the predicted values ​​are accurate. Therefore, the two are weighted and fused to obtain a new quantity: the state value. The measured values ​​are also called traces, which are a series of points detected by the radar on the same target (assuming the target is stationary, but due to errors, more than one point will be detected). The track corresponds to the state value, which is a weighted fusion of the traces and the predicted values. The input to radar track processing is the set of traces sent by the radar, and the output after processing is the set of tracks.

[0096] One of the key technologies: cooperative detection and tracking.

[0097] In practical applications, when ground surveillance radar and photoelectric detectors work together, the radar searches its area of ​​responsibility and provides guidance information, while the photoelectric detector automatically locks onto and tracks the target within the designated area. Simultaneously, the integrated processing and display software identifies the tracked target and synthesizes the target information.

[0098] In multi-target environments, due to limitations in the scheduling of photoelectric detector location resources, this scheme introduces the concept of dense multi-target (formation) and threat level criteria. In this case, the radar-photoelectric cooperative detection and tracking process is as follows: Figure 3 As shown in the diagram, a Frame represents each frame of data or one scan cycle. A gate is a spatial and logical range of possible measurement values ​​for a target, derived from information such as target threat level, target status (position, speed, heading, etc.), and historical trajectory. Falling into a relevant gate indicates that the measurement is the latest measurement of the target confirmed by this scheme, and will participate in the state update calculation of the target tracking algorithm. It also signifies that this is a genuine, purposeful target with a continuous trajectory, and should be monitored.

[0099] In some embodiments, the radar detection information includes the direction of travel and speed of the detected targets. Multiple detected targets are determined to be a formation target when the following conditions are met: the number of detected targets is greater than a preset number threshold; the distribution range of the multiple detected targets is smaller than a preset spatial range; the multiple detected targets have the same direction of travel; and / or, the difference in speed between the multiple detected targets is less than a preset speed range.

[0100] First, a group of at least three groups of targets moving in the same direction at similar speeds within a certain space is defined as a formation target. The distance between individual targets within a formation target is much smaller than the distance between individual targets in different formation targets.

[0101] It should be noted that the specific criteria for determining "formation" differ depending on the target. For example, medium tanks are generally 8-10 meters long and travel at speeds typically between 30-50 km / h, allowing for a formation spacing of up to 100 meters. Civilian vehicles, on the other hand, are generally less than 5 meters long and travel at speeds exceeding 60 km / h on the road; terrorist convoys on the border follow closely, typically maintaining a distance of less than 50 meters. Furthermore, the size of the formation varies considerably, resulting in significant differences in spatial range.

[0102] The definition of formation target given in this embodiment is the theoretical basis for the algorithm to determine whether the actual measured value is part of the formation. In actual judgment, specific constraints can be set according to the specific application scenario, and these constraints or judgment logic are also related to the radar used in the actual application. Therefore, the data recorded in the above embodiments is merely an example and should not be construed as a limitation of the present invention.

[0103] In some embodiments, the radar detection information further includes: the target's motion state, radar scattering information, and target classification information. The target threat level is derived by combining the target's motion state, radar scattering information, target classification information, warning area information, and a threat target database; then, the threat levels of multiple detected targets are arranged sequentially to establish a threat level list. The target's motion state includes its position, speed, acceleration, heading, and even pitch and roll angles. Radar scattering information is the target's RCS (Radar Cross Section), calculated based on the echo signal strength during radar detection. Target classification information refers to whether the target is a person, vehicle, tank, container truck, small boat, cruise ship, etc. The warning area refers to the geographical area most likely to contain threatening targets, or the geographical area where high-value targets require protection, which is of concern to border personnel.

[0104] Once the radar detects a high-value formation target, it immediately exits the unattended mode and guides the photoelectric detector to identify and collect evidence of the suspicious target. In general multi-target environments, a threat level list is established for the points / tracks. Once the conditions are met, the photoelectric detector is guided to identify, photograph, and collect evidence of targets with high threat levels in sequence (the evidence collection time for each target can be customized).

[0105] The second key technology: incomplete measurement.

[0106] Although the high-point monitoring system has achieved collaborative detection and tracking of radar and photoelectric multi-source sensors, the complex ground environment where the target is located, the obstruction of obstacles, weather conditions, and the vibration of the monitoring platform inevitably cause the loss of measurement data. At this time, false alarms or outliers may also occur in the system, which introduces the problem of incomplete measurement.

[0107] The method of this application also includes the following steps to solve the incomplete measurement problem: determining the state vector of the target based on radar detection information; establishing a nonlinear model based on the state vector; introducing a binary discrete random variable into the nonlinear model to obtain an improved nonlinear model; establishing a state update equation and a measurement update equation based on the improved nonlinear model; determining the incomplete measurement criterion based on the state update equation and the measurement update equation; and updating the measurement vector based on the incomplete measurement criterion.

[0108] Ground target maneuvers are often difficult to describe using linear systems. To balance algorithm maturity and complexity, this scheme employs the Extended Kalman Filter (EKF) to address this issue. In some embodiments, the system's nonlinear model is represented as follows:

[0109]

[0110] In the formula, Φ(k+1|k) is the target nonlinear state transition matrix, and X(k)∈ Let Γ(k) be the target state vector at time k. Let u(k) be the process noise distribution matrix at time k, and let u(k) be the process noise vector ∈ It is Gaussian white noise with zero mean and variance Q(k); Z(k)∈ H(k) is the system observation vector; H(k) is the measurement equation, and the measurement noise vector v(k) ∈ It is Gaussian white noise with zero mean and variance R(k). X(k)∈ X(k) represents an n×1 matrix, which is an n-dimensional vector; the other variables are similar and will not be elaborated further.

[0111] When incomplete measurements exist, a binary discrete random variable g(k) is introduced, where g(k) = 1 indicates normal measurement and g(k) = 0 indicates incomplete measurement. In this case, the measurement equation in (1) is:

[0112] Z(k)=g(k)H(k)X(k)+v(k) (2)

[0113] Its measurement noise can be expressed as

[0114]

[0115] In the formula, g(k) = 0 corresponds to the limit value of δ, that is, δ → ∞.

[0116] The system uses the EKF algorithm to track the target in two steps:

[0117] Step 1, Status Update

[0118]

[0119] P(k+1 / k)=Φ(k)P(k / k)Φ T (k)+Γ(k)Q(k)Γ T (k) (5)

[0120] The second step is measurement and update.

[0121]

[0122] K(k+1)=P(k+1 / k)H T (k+1)[H(k+1)P(k+1 / k)H T (k+1)+R(k+1)] -1 (7)

[0123]

[0124] P(k+1 / k+1)=[IK(k+1)H(k+1)]P(k+1 / k) (9)

[0125] Here, we use the prediction residual of the measurement vector at time k+1, which is easy to implement in engineering. As a criterion for incomplete measurement:

[0126]

[0127] In the formula, C is the residual detection threshold. When the above formula holds, g(k) = 1; otherwise, g(k) = 0 indicates incomplete measurement. In this case, the measurement update can be further expressed as:

[0128]

[0129]

[0130] Equations (4)-(6) and (11)-(12) constitute the EKF algorithm under incomplete measurement. Based on the EKF algorithm under incomplete measurement, the cooperative detection and tracking system further adopts interactive multi-model (IMM) and multi-hypothesis tracking (MHT) techniques to solve the problem of diverse target motion state characteristics and high-maneuverability tracking.

[0131] The third key technology: interactive multi-model.

[0132] Because ground-based moving targets are relatively maneuverable and operate in complex environments, especially in multi-target tracking scenarios, their motion cannot be well modeled by a single state equation. To achieve good accuracy while tracking maneuvering targets, the tracking algorithm essentially requires that the target motion be described by different state models at different times. This is the Interacting Multiple Model (IMM) estimation method, a multi-model filter with Markov switching coefficients proposed based on the generalized pseudo-Bayesian algorithm and the Kalman filter. Multiple models operate in parallel, switching between them based on a Markov chain, and the target state is the result of the interaction of multiple filters. The IMM algorithm consists of three parts: interaction, filtering, and combination. Figure 4 As shown.

[0133] The IMM algorithm assumes that a given model is valid at the current time step. It obtains the initial conditions for filters matching this specific model by mixing the state estimates of all filters from the previous time step. Then, it performs regular filtering (prediction and correction) steps in parallel for each model. Finally, it updates the model probability based on the model-matching likelihood function and combines all the corrected state estimates of the filters in a weighted sum form to obtain the state estimate. The probability of a model being valid plays a crucial role in the weighted summation calculation of the state estimate and covariance.

[0134] The characteristics of the IMM algorithm include: First, the filtering algorithm utilizes Gaussian distributions to approximate Gaussian sum distributions in several places; therefore, it is also known as a Gaussian sum adaptive estimation algorithm. Second, the utilization of measurement information is reflected not only in the filtering estimation but also in the model probability, achieving adaptive model adjustment through changes in the model probability. Third, the algorithm is modular; depending on the application, the filtering module can employ various linear and nonlinear filtering algorithms. Fourth, the filtering modules in the algorithm are computed in parallel, improving computational efficiency.

[0135] Key technology four: multi-hypothesis tracking.

[0136] When multiple targets and clutter exist, the source of each measurement data point is uncertain. Therefore, the source of each measurement data point must be determined before it can be used for state updates. This invention employs a multiple hypothesis tracking (MHT) data association method.

[0137] The "hypotheses" in the multi-hypothesis method have the following characteristics:

[0138] (1) When forming a hypothesis, not only should the possibility of a false alarm be considered for any valid echo, but also the possibility of a new target appearing; (2) The hypothesis at time k is obtained by associating the hypothesis at time k-1 with the current accumulated measurement set.

[0139] The multi-hypothesis tracking method is characterized by continuously maintaining the hypothesis structure of each frame and expanding or deleting hypotheses when new observation data arrives, using measurements from multiple scans to solve the data correlation problem.

[0140] Whenever the algorithm receives a new measurement, it forms a new hypothesis about the target and the acquired measurement based on the existing hypothesis, and iterates through and evaluates the posterior probability of all possible related hypotheses and the target state estimate corresponding to the hypothesis.

[0141] The MHT algorithm provides a unified processing framework for all aspects of multi-target tracking, such as track splitting, data association, track initiation, and track maintenance. This is something that probabilistic data association algorithms do not have. Their disadvantage is that they rely too much on prior knowledge of the target and clutter.

[0142] In practical applications, the Structured Bifurcation Theory (SBT) algorithm can be divided into hypothesis-oriented and track-oriented SBT algorithms. The latter does not continuously store hypotheses but updates the formed tracks in each frame and predicts the position of the surviving tracks for the next frame, repeating this process. Track-oriented SBT algorithms initialize, update, and evaluate the quality of tracks using a fractional function before forming hypotheses. This way, tracks with low probability are deleted before hypothesization in the next stage. Due to the success of structured branching theory, track-oriented SBT algorithms are more feasible in practice than hypothesis-oriented SBT algorithms.

[0143] In summary, this invention's border high-point monitoring system, from a practical application perspective, explores application modes under unattended conditions. It employs a collaborative detection and tracking system using ground-based surveillance radar and a remote photoelectric detector. The ground-based surveillance radar performs continuous 24 / 7 scanning and reconnaissance. When a suspicious target is detected and forms a track, the system's integrated processing and display terminal issues an alarm, while simultaneously guiding the photoelectric detector to the suspicious target for further identification and verification. To support this operating mode, a radar-photoelectric single-station collaborative detection and tracking technology based on target threat alarm criteria and a multi-target tracking algorithm under incomplete measurement conditions were developed. Demonstration and verification experiments based on typical applications validated the system's functional performance and its ability to support radar-photoelectric linkage operations under unattended conditions.

[0144] Embodiments of this application also provide a multi-target detection and tracking system combining single-station radar and optoelectronic coordination, comprising: radar, optoelectronic detector, and fusion processing module.

[0145] The radar is used for dynamic scanning to acquire the distance and azimuth information of the target. The photoelectric detector is used to identify and confirm the designated target.

[0146] The fusion processing module is used to determine whether multiple detection targets are in formation based on radar detection information in unattended mode. If they are in formation, the module guides the photoelectric detector to the formation target for detection and exits the unattended mode. If they are not in formation, the module establishes a threat level list for multiple detection targets based on radar detection information and guides the photoelectric detector to track and collect evidence from the targets in sequence according to their threat levels.

[0147] Regarding the multi-target detection and tracking system in the above embodiments, the specific steps of the fusion processing module's operation have been described in detail in the embodiments related to the method, and will not be elaborated further here. The fusion processing module can be implemented entirely or partially through software, hardware, or a combination thereof. The fusion processing module can be embedded in or independent of the processor in a computer device in hardware form, or it can be stored in the memory of a computer device in software form, so that the processor can call and execute the operation corresponding to the fusion processing module.

[0148] It is understood that the same or similar parts in the above embodiments can be referred to each other, and the contents not described in detail in some embodiments can be referred to the same or similar contents in other embodiments.

[0149] Any process or method described in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing a particular logical function or process, and the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the function involved, as will be understood by those skilled in the art to which embodiments of this application pertain.

[0150] It should be understood that various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0151] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

[0152] Furthermore, the functional units in the various embodiments of this application can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0153] The storage media mentioned above can be read-only memory, disk, or optical disk, etc.

[0154] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0155] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. A method for multi-target detection and tracking using monostation radar and electro-optical coordination, characterized in that, include: In unattended mode, the radar detection information is used to determine whether multiple detected targets are formation targets; The radar detection information includes: the direction of travel and speed of the detected target; Multiple detection targets are considered to be in formation when they meet the following conditions: The number of detected targets exceeds a preset threshold. The distribution area of ​​multiple detection targets is smaller than the preset spatial range; Multiple detection targets are traveling in the same direction; and / or, The difference in travel speed between multiple detection targets is less than a preset speed range; The radar detection information also includes: the motion state of the detected target, radar scattering information, and target classification information; Creating a threat level list for multiple detected targets involves the following steps: The threat level of the detected target is determined based on the target's motion state, radar scattering information, target classification information, and in conjunction with a pre-set threat target database. Arrange the threat levels of multiple targets in order to create a threat level list; If it is a formation target, guide the photoelectric detector to the formation target location for detection and exit the unattended mode; If the target is not part of a formation, a threat level list is created for multiple targets based on radar detection information, and the photoelectric detector is guided to track and collect evidence from the targets in sequence according to their threat level.

2. The method according to claim 1, characterized in that, It also includes the following steps: Determine the state vector of the detected target based on radar detection information; Establish a nonlinear model based on state vectors; By introducing bivariate discrete random variables into the nonlinear model, an improved nonlinear model is obtained. State update equations and measurement update equations are established based on an improved nonlinear model; Incomplete measurement criteria are determined based on state update equations and measurement update equations; The measurement vector is updated based on the incomplete measurement criterion.

3. The method according to claim 2, characterized in that, The nonlinear state transition model is as follows: ； in, Let be the target nonlinear state transition matrix. for The target state vector at any given time. for Time-matrix process noise distribution matrix, process noise vector The mean is 0 and the variance is Gaussian white noise; This is the system observation vector; For the measurement equation, measure the noise vector. The mean is 0 and the variance is Gaussian white noise.

4. The method according to claim 3, characterized in that, Introducing a binary discrete random variable , This indicates that the measurement is normal. Indicating incomplete measurement, the measurement equation in the nonlinear state transition model is: ; Measurement noise is expressed as: ； In the formula, correspond Take the limit value, that is .

5. The method according to claim 4, characterized in that, The state update equation is: ； ； The measurement update equation is: ； ； ； 。 6. The method according to claim 5, characterized in that, by Predicted residuals of measurement vectors at time points As a criterion for incomplete measurement: ； In the formula, C This is the residual detection threshold; when the above formula holds true... This indicates that no incomplete measurement occurred; when the above formula is not true... This indicates that an incomplete measurement has occurred.

7. The method according to claim 6, characterized in that, The measurement update equation is expressed as: 。 8. A multi-target detection and tracking system combining single-station radar and optoelectronic coordination, characterized in that, include: Radar, photoelectric detectors, and fusion processing modules; The radar is used to perform dynamic scanning and acquire information on the range, azimuth, speed, heading, and classification of the detected target. The photodetector is used to identify and confirm the designated detection target; The fusion processing module is used to determine whether multiple detected targets are formation targets based on radar detection information in unattended mode; If the target is a formation, the photoelectric detector is guided to the formation target for detection and exits the unattended mode; if the target is not a formation, a threat level list is established for multiple detected targets based on the radar detection information, and the photoelectric detector is guided to track and collect evidence from the targets in order of threat level. The radar detection information includes: the direction of travel and speed of the detected target; Multiple detection targets are considered to be in formation when they meet the following conditions: The number of detected targets exceeds a preset threshold. The distribution range of multiple detection targets is smaller than the preset spatial range; Multiple detection targets are traveling in the same direction; and / or, The difference in travel speed between multiple detection targets is less than a preset speed range; The radar detection information also includes: the motion state of the detected target, radar scattering information, and target classification information; Creating a threat level list for multiple detected targets involves the following steps: The threat level of the detected target is determined based on the target's motion state, radar scattering information, target classification information, and in conjunction with a pre-set threat target database. Arrange the threat levels of multiple targets in order to create a threat level list.

Images