A method and system for monitoring and early warning of low-altitude aircraft

Abstract

This application relates to the field of low-altitude aircraft monitoring and early warning technology, and discloses a method and system for low-altitude aircraft monitoring and early warning. The method includes: delineating sensitive areas and key nodes in a city; establishing a pre-set abnormal micro-behavior fingerprint database that associates masquerading identities with sensitive areas and key nodes; extracting target behavior data of aerial targets in real time; when an aerial target enters a sensitive area or approaches a key node, matching the target behavior data with the abnormal micro-behavior fingerprint database to determine if the aerial target exhibits abnormal behavior; and triggering an early warning if the aerial target exhibits abnormal behavior. This application, by constructing an abnormal micro-behavior fingerprint database and introducing multi-dimensional behavioral deviation analysis, achieves a shift from passive detection to proactive intent recognition of low-altitude threat aircraft disguised as legitimate entities, significantly improving the city's low-altitude security system's ability to accurately warn and respond in real-time to intelligent and concealed threats in complex environments.

Description

Technical Field

[0001] This application relates to the field of low-altitude aircraft monitoring and early warning technology, and more specifically, to a method and system for low-altitude aircraft monitoring and early warning. Background Technology

[0002] In the security system of smart cities, a comprehensive monitoring system is typically deployed to maintain the safety of urban low-altitude airspace. This system mainly consists of various sensors, such as radar units deployed on the tops of high-rise buildings or specific monitoring towers, responsible for continuously scanning a large area of ​​airspace to detect flying targets entering or crossing urban airspace. When the radar detects a target, it initially obtains its spatial position, flight speed, altitude, and other motion parameters. This preliminary data is then transmitted to a central processing platform. The central processing platform uses this data, combined with preset airspace management rules and reported flight plans, to make a preliminary judgment on the legitimacy of the target. If the target is determined to be unauthorized or a potential threat, the system immediately initiates subsequent tracking and identification processes. At this time, electro-optical tracking equipment deployed in key areas is activated, accurately pointing to the target based on the approximate location provided by the radar, and acquiring visual images of the target through high-resolution optical zoom. These images are then sent to an image recognition module to identify the specific model, appearance characteristics, and possibly even the payload it carries. Simultaneously, radio spectrum monitoring equipment will also operate to intercept control signals or data link signals that the aircraft may be emitting, in order to help determine its source and intent. After all this information is aggregated, the system will conduct a comprehensive analysis. If the target is confirmed to be an illegal intrusion or pose a threat, an early warning mechanism will be triggered to notify security personnel to take action. This process aims to achieve real-time, all-round monitoring of low-altitude airspace to ensure the safety of the city.

[0003] However, in actual urban security work, we have discovered a new challenge stemming from the evolution of potential threat creators' strategies. They are no longer simply pursuing low detectability of aircraft, but have shifted to a "highly deceptive" strategy. This means that threat aircraft do not attempt to be completely stealthy, but rather carefully design their appearance and flight patterns to highly mimic harmless low-flying objects common in the urban environment during the initial detection phase. For example, attackers might use an aircraft with special paint schemes or modified shapes to make it visually difficult to distinguish from large birds, such as eagles or falcons. Such aircraft may even be programmed to simulate typical bird flight characteristics, such as hovering, gliding, or drifting in specific wind directions, to further enhance their deceptiveness. Another common strategy is to disguise threat aircraft as common civilian or commercial drone models and simulate their routine flight missions, such as package delivery, building inspection, or aerial photography operations.

[0004] When such camouflaged aircraft enter the monitored airspace, radar systems detect their presence and acquire motion parameters such as size, speed, and altitude. However, due to their size and flight speed being highly similar to large birds or small civilian drones, the radar's initial classification process often categorizes them as "non-threatening targets" or "pending targets," lowering their priority or even filtering them out entirely. Subsequently, when electro-optical cameras are tracked, the images captured are also blurred due to the aircraft's camouflage, or misidentified as common birds or legitimate drones in the image recognition module. Even experienced security personnel may overlook them when quickly reviewing surveillance footage due to their highly realistic appearance and flight attitude. Furthermore, if such aircraft fly autonomously along a preset route or are controlled using common civilian radio protocols, radio spectrum monitoring equipment will struggle to detect abnormal signals because they do not emit any "threatening" electromagnetic signatures, or their signal characteristics are mixed in with numerous legitimate devices, making them difficult to distinguish.

[0005] Even more challenging is that these threat actors cleverly exploit the "background noise" of the urban environment to mask their true intentions. This "background noise" doesn't refer to acoustic noise, but rather to the large amount of legitimate flight activity and natural biological activity present in urban low-altitude airspace. They might choose to infiltrate near city parks, nature reserves, or during large events when there is a high volume of legitimate drones or bird activity in the air. For example, over a city park on a weekend, there might be dozens or even hundreds of recreational drones flying, accompanied by a large number of birds. Under such high background activity, monitoring systems receive a massive amount of signals from legitimate flying objects or natural organisms, causing the system's internal "normal" activity threshold to be raised. In this situation, a threat aircraft disguised as a bird or a legitimate drone, with its subtle behavioral anomalies—such as a brief, unnatural hovering over a sensitive area, or a sudden change in altitude and subsequent stillness during flight—can easily be drowned out by the massive amount of normal data and mistaken by the system for data fluctuations or variations in normal behavior.

[0006] At this point, the monitoring system faces a dilemma: it can detect targets and perform preliminary classifications, but these classifications are based on their "camouflage" rather than their "true intentions." The system lacks a mechanism to identify extremely subtle "behavioral anomalies" from seemingly normal flight data that contradict their camouflaged identity. These anomalies might include minor deviations from the flight path, such as a sudden, tiny, non-inertial zigzag jerk during straight flight; instantaneous changes in attitude, such as a sudden, precise attitude adjustment during gliding, rather than the natural body sway of a bird; atypical fluctuations in speed, such as a sudden, brief, aerodynamically inefficient acceleration or deceleration during steady flight; or brief hoverings at specific geographical locations, such as a precise hovering for several seconds in front of a window of a sensitive building. These behaviors themselves may not pose a direct threat, but they are logically inconsistent with the "identity" the target is camouflaging. For example, a bird would not maintain a perfectly still attitude in the air for an extended period, nor would a civilian drone perform precise, fixed-point hovering over an unreported sensitive area or execute abnormal flight patterns outside of its mission area. However, existing systems often focus on identifying "known threat patterns" or "significant anomalies," and are not good at identifying "minor inconsistencies beneath a normal appearance."

[0007] This highly deceptive strategy renders traditional early warning methods based on single sensors or simple data fusion inefficient. Radar, optoelectronic, and radio spectrum equipment, operating independently, can all be fooled by target camouflage. Even simple data overlay may fail to effectively distinguish between "normal variations" and "malicious camouflage" due to a lack of in-depth understanding of "behavioral patterns." For example, a system might mistake a "bird-shaped" drone briefly hovering over a sensitive area for a bird circling in the air, failing to trigger a high-level alarm. Security personnel, faced with numerous alerts for "suspected birds" or "suspected civilian drones," also struggle to quickly determine their true nature because these alerts' characteristics highly overlap with normal activity, thus missing the optimal response window. Therefore, how to meticulously analyze seemingly normal but actually deceptively complex multi-source data to identify extremely subtle behavioral patterns that contradict the target's "camouflaged identity" and use them as the basis for early warning has become a critical technical challenge urgently needing to be addressed in the field of low-altitude security for smart cities. Summary of the Invention

[0008] To address the shortcomings of existing technologies, this application discloses a method and system for monitoring and warning low-altitude aircraft, aiming to solve the technical problem that existing low-altitude aircraft monitoring and warning systems are unable to effectively identify subtle behavioral anomalies when facing highly deceptive camouflaged aircraft, leading to warning failure or delay.

[0009] The technical solution of this application is as follows: In a first aspect, this application discloses a method for monitoring and early warning of low-altitude aircraft, comprising the following steps: Sensitive areas and key nodes in the city are delineated, and an abnormal micro-behavior fingerprint database is established that associates pre-set disguised identities with sensitive areas and key nodes. The pre-set disguised identities are harmless entity identities that are imitated by threatened aircraft. Real-time extraction of target behavior data of aerial targets; when an aerial target enters a sensitive area or approaches a key node, the target behavior data is matched with an abnormal micro-behavior fingerprint database to determine whether the aerial target has abnormal behavior. If an aerial target is found to be behaving abnormally, an alert is triggered.

[0010] This technical solution enables the effective identification of subtle behavioral anomalies of camouflaged aircraft near sensitive areas or key nodes, overcoming the limitations of traditional systems in identifying highly deceptive threats, thereby improving the accuracy and timeliness of early warning.

[0011] Furthermore, the steps for delineating sensitive areas and key nodes in the city include: By importing urban planning data, architectural drawings, and distribution maps of important infrastructure, sensitive areas and key nodes are delineated in three-dimensional space in the form of vector graphics and stored in the monitoring database.

[0012] Based on this, the steps to establish a pre-defined database of abnormal micro-behavior fingerprints that associates spoofed identities with sensitive areas and key nodes include: Based on a comprehensive assessment of the normal behavior patterns of the pre-set masquerading identity in different sensitive areas and key nodes, aerodynamic principles, aircraft design specifications, and the experience of security experts, an abnormal micro-behavior fingerprint database is established. The abnormal micro-behavior fingerprint database includes a set of logical rules. Each logical rule defines a subtle behavior pattern that does not conform to the pre-set masquerading identity's behavior logic in sensitive areas and key nodes, including hovering duration, trajectory deviation magnitude, and attitude change rate.

[0013] Furthermore, the steps for extracting target behavior data of aerial targets in real time include: The system continuously receives radar, optoelectronic, and radio spectrum data from aerial targets via radar, cameras, and radio spectrum equipment, and performs timestamp alignment and data fusion processing to extract target behavior data.

[0014] Furthermore, when an aerial target enters a sensitive area or approaches a critical node, the steps to match the target's behavior data with an abnormal micro-behavior fingerprint database to determine whether the aerial target exhibits abnormal behavior include: When an aerial target enters a sensitive area or approaches a critical node, the conformity of the target's behavior data with multiple behavior fingerprints in the abnormal micro-behavior fingerprint database is evaluated. Fuzzy logic or a rule-based inference engine is used to evaluate the conformity. If the consistency between the target behavior data and at least one behavior fingerprint reaches a preset consistency threshold, it is determined that the aerial target has abnormal behavior.

[0015] As a technological improvement, if an abnormal behavior of an aerial target is detected, the steps to trigger an early warning include: Trigger an alert for abnormal target behavior and generate alert information including the target's location, identity type, and description of abnormal micro-behavior. The alert information is sent to security personnel via screen display, mobile terminal push, SMS, or telephone notification.

[0016] To enhance functionality, the warning information also includes action recommendations, such as dispatching patrol drones, activating directional jamming equipment, or notifying ground security personnel.

[0017] Building upon the above, this application further proposes that when an aerial target enters a sensitive area or approaches a critical node, the method also includes: Acquire and integrate environmental information to create a real-time environmental status map; Continuously receive physical characteristics and flight information of aerial targets; Based on the physical characteristics and flight information of aerial targets and the real-time environmental state map, the legal flight behavior of aerial targets under the real-time environmental state map is simulated to generate predicted normal behavior trajectories. Calculate the behavioral deviation score based on the target behavior data and the predicted normal behavior trajectory; If the behavior deviation score exceeds the preset deviation threshold, it is determined that the aerial target has abnormal behavior and an early warning is triggered.

[0018] More specifically, the steps for calculating the behavioral deviation score based on the target behavioral data and the predicted normal behavioral trajectory include: Extract multi-dimensional behavioral feature vectors from target behavior data, including the real-time position, velocity, acceleration, attitude angle, rate of change of attitude angle, rate of change of profile, and signal feature intensity of aerial targets; Calculate the Euclidean distance between the multi-dimensional behavioral feature vector and the feature vector corresponding to the predicted normal behavioral trajectory; Calculate the cosine similarity between the multi-dimensional behavioral feature vector and the feature vector corresponding to the predicted normal behavioral trajectory; The behavioral deviation score is calculated by combining Euclidean distance and cosine similarity.

[0019] Secondly, this application also discloses a low-altitude aircraft monitoring and early warning system, comprising: The fingerprint database establishment module is used to delineate sensitive areas and key nodes in the city, and establish an abnormal micro-behavior fingerprint database that associates the preset disguised identity with the sensitive areas and key nodes. The preset disguised identity is the identity of a harmless entity that is imitated by the threatened aircraft. The behavior judgment module is used to extract target behavior data of aerial targets in real time. When an aerial target enters a sensitive area or approaches a key node, the target behavior data is matched with the abnormal micro-behavior fingerprint database to determine whether the aerial target has abnormal behavior. The early warning triggering module is used to trigger an early warning if it determines that an aerial target is behaving abnormally.

[0020] This application provides a system for implementing the above-mentioned monitoring and early warning methods through this technical solution. Through modular design, it can efficiently complete functions such as sensitive area delineation, abnormal behavior fingerprint database establishment, real-time behavior judgment and early warning triggering, providing reliable hardware and software support for urban low-altitude security.

[0021] In summary, this application discloses a method and system for monitoring and warning low-altitude aircraft. The method delineates sensitive areas and key nodes within a city and establishes a pre-defined database of abnormal micro-behaviors associated with these areas and their disguised identities. This allows for the targeted identification of abnormal behaviors by threatened aircraft mimicking harmless entities. When an aerial target enters a sensitive area or approaches a key node, the system extracts the target's behavioral data in real time and matches it against the abnormal micro-behavior fingerprint database to determine if the target exhibits abnormal behavior. Once an anomaly is detected, an early warning is triggered. This method effectively solves the problem in existing technologies where traditional monitoring systems fail or delay warnings when facing highly deceptive disguised aircraft, whose appearance and behavior closely mimic harmless entities. By establishing a refined database of abnormal micro-behaviors, this application can identify subtle behavioral patterns inconsistent with the disguised identity from seemingly normal flight data, such as hovering duration, trajectory deviation, and attitude change rate, thus overcoming the shortcomings of existing systems in identifying subtle inconsistencies beneath a seemingly normal appearance. Therefore, this application can significantly improve the accuracy of identifying camouflaged threats and the timeliness of early warning, providing a more reliable and efficient solution for low-altitude security in smart cities. Attached Figure Description

[0022] Figure 1 This is a flowchart illustrating a low-altitude aircraft monitoring and early warning method provided in an embodiment of this application.

[0023] Figure 2 This is a schematic diagram of the structure of a low-altitude aircraft monitoring and early warning system provided in an embodiment of this application.

[0024] Labeling Explanation: 210, Fingerprint Database Establishment Module; 220, Behavior Judgment Module; 230, Early Warning Trigger Module. Detailed Implementation

[0025] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments. The components of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0026] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0027] Traditional low-altitude aircraft monitoring systems are highly susceptible to being misled in the initial assessment of threat aircraft employing deceptive tactics. These aircraft closely mimic the appearance and behavior of harmless entities, leading to the misidentification of genuine threats as harmless targets and resulting in ineffective or delayed early warnings. Furthermore, existing systems lack the ability to identify subtle inconsistencies beneath a seemingly normal appearance, struggling to discern minute behavioral anomalies from massive amounts of data that contradict their disguised identities, thus missing crucial response opportunities.

[0028] Firstly, please see Figure 1 This application proposes a method for monitoring and early warning of low-altitude aircraft, including: S1. Delineate sensitive areas and key nodes in the city, and establish an abnormal micro-behavior fingerprint database that associates pre-set disguised identities with sensitive areas and key nodes. The pre-set disguised identities are harmless entity identities that are imitated by threatened aircraft. S2. Extract target behavior data of aerial targets in real time. When an aerial target enters a sensitive area or approaches a key node, match the target behavior data with the abnormal micro-behavior fingerprint database to determine whether the aerial target has abnormal behavior. S3. If an abnormal behavior is detected in an aerial target, an early warning will be triggered.

[0029] The core of the low-altitude aircraft monitoring and early warning method in this application lies in the in-depth analysis and anomaly identification of the behavior patterns of low-altitude aircraft.

[0030] First, it is necessary to delineate sensitive areas and key nodes within the city. This can be achieved in several ways. For example, these areas and nodes can be manually marked on a map, and their coordinates input into the monitoring system. Alternatively, existing Geographic Information System (GIS) data can be imported, integrating urban planning data, architectural drawings, and maps of important infrastructure distribution to accurately delineate these sensitive areas and key nodes in three-dimensional space in vector graphics form, and then stored in the monitoring database. The delineation of these areas and nodes forms the basis for subsequently establishing an abnormal micro-behavior fingerprint database and conducting behavioral analysis.

[0031] After delineating sensitive areas and key nodes, it is necessary to establish a pre-defined fingerprint database of anomalous micro-behaviors that associates masquerading as identities with these sensitive areas and key nodes. The establishment of this fingerprint database is one of the key innovations of this application. For example, based on the experience of security experts and combined with the analysis of normal behavior patterns of different types of harmless entities (such as birds and civilian drones) in different sensitive areas and key nodes, a series of anomalous micro-behavior rules can be manually defined. Alternatively, by collecting a large amount of historical flight data and using machine learning algorithms to model normal behavior patterns, subtle behaviors that deviate significantly from these normal patterns can be identified. For aircraft disguised as civilian drones, precise hovering in non-mission areas or suddenly changing altitude and remaining stationary during flight may also be considered anomalous. These anomalous micro-behavior patterns are encoded as fingerprints and stored in the fingerprint database for subsequent matching.

[0032] Subsequently, the system needs to extract target behavior data of aerial targets in real time. This can be achieved by deploying multiple sensors. For example, radar systems can continuously scan the airspace to acquire radar data such as the target's distance, azimuth, and velocity. Cameras can capture visual images of the target and acquire photoelectric data. Radio spectrum equipment can monitor the radio signals emitted by the target and acquire radio spectrum data. After receiving this multi-source data, it needs to undergo timestamp alignment and data fusion processing to generate unified target behavior data. For example, data acquired by different sensors at the same point in time can be correlated to form a behavior feature vector containing multi-dimensional information such as target position, velocity, attitude, and signal characteristics.

[0033] When an aerial target enters a sensitive area or approaches a critical node, the system matches the target's real-time extracted behavior data with an abnormal micro-behavior fingerprint database to determine if the target exhibits abnormal behavior. This matching can be achieved through a rule-based inference engine, which presets a series of logical rules; if the target's behavior data meets a certain abnormality rule, it is determined to be abnormal. Alternatively, fuzzy logic can be used for matching, which evaluates the degree of conformity between the target's behavior data and multiple behavior fingerprints in the database; if the conformity reaches a preset threshold, it is determined to be abnormal. For example, if a target disguised as a civilian drone hovers over a sensitive area for more than a preset normal hovering time, or if its trajectory deviates from a preset normal flight path, it may be judged as exhibiting abnormal behavior.

[0034] If an aerial target is deemed to be behaving abnormally, an alert is triggered. Alert triggering can take various forms. Simultaneously, alert information can be sent to security personnel via mobile push notifications, SMS, or telephone calls to ensure they are promptly informed of the anomaly. Alert information typically includes the target's location, the type of identity it is masquerading as, and a description of specific abnormal micro-behaviors, enabling security personnel to quickly understand the situation and take appropriate action.

[0035] This application presents a low-altitude aircraft monitoring and early warning method designed to address the problem of warning failure or delay in traditional monitoring systems when facing highly deceptive threatening aircraft. Traditional systems often focus on identifying "known threat patterns" or "significant anomalies," lacking the ability to identify "minor inconsistencies beneath a normal appearance" in camouflaged targets. For example, a drone disguised as a bird might briefly hover unnaturally over a sensitive area, or suddenly change altitude and remain stationary during flight. These subtle movements are easily lost in the vast amount of normal data and mistaken by the system for data fluctuations or variations in normal behavior. Compared to existing technologies, the core innovation of this application lies in introducing the concept of an "abnormal micro-behavior fingerprint database" and associating it with "preset camouflaged identities" and "sensitive areas and key nodes." This method allows the system to move beyond simply identifying the target's physical characteristics or significant anomalies and delve into the logical analysis of the target's behavioral patterns. Therefore, this application significantly improves the ability to identify "highly deceptive" low-altitude aircraft, effectively makes up for the shortcomings of existing technologies in responding to new threat strategies, and thus improves the overall effectiveness of urban low-altitude security and the timeliness and accuracy of early warning.

[0036] Specifically, the steps for delineating sensitive areas and key nodes in the city include: By importing urban planning data, architectural drawings, and distribution maps of important infrastructure, sensitive areas and key nodes are delineated in three-dimensional space in the form of vector graphics and stored in the monitoring database.

[0037] Urban planning data can be understood as documents such as the city's master plan and detailed plans issued by government departments, which contain macro-level information such as urban functional zoning, land use, and road and transportation networks. Architectural drawings refer to detailed design drawings of various buildings, such as structural diagrams, floor plans, and elevations, providing precise geometric information such as the building's height, shape, and internal structure. Maps of important infrastructure distribution cover the geographical location and spatial extent of key nodes such as power facilities, communication base stations, water conservancy projects, and transportation hubs. The import of these data sources aims to provide a comprehensive and accurate geospatial foundation for subsequent regional delineation.

[0038] Furthermore, delineating sensitive areas and key nodes in three-dimensional space using vector graphics refers to converting the imported two-dimensional or three-dimensional data into editable vector graphic objects using Geographic Information Systems (GIS) or other 3D modeling software. These vector graphic objects can accurately represent the spatial boundaries, height information, and interrelationships of sensitive areas and key nodes in a three-dimensional coordinate system. For example, sensitive areas can be delineated as three-dimensional polygons or blocks with specific heights and ranges, while key nodes can be represented as three-dimensional points or small-scale three-dimensional regions. This three-dimensional vectorization method can more realistically and accurately reflect the actual physical space, providing a refined spatial reference for the monitoring of low-altitude aircraft.

[0039] This application's solution integrates multi-source heterogeneous geospatial data, including urban planning data, architectural drawings, and maps of critical infrastructure distribution, and transforms them into 3D vector graphics. This enables high-precision, 3D spatial delineation of sensitive areas and key nodes in cities. This delineation method overcomes the limitations of traditional 2D planar delineation in handling complex urban environments, especially high-rise buildings and multi-layered spatial structures. By precisely defining these areas and nodes in 3D space, it is possible to more accurately determine whether aerial targets are entering or approaching these protected spaces, thus providing a solid spatial foundation for subsequent anomaly detection. Storing this precisely delineated information in a monitoring database ensures data manageability, retrieval, and real-time availability, providing data support for the effective operation of the monitoring and early warning system.

[0040] Specifically, the steps for establishing a pre-defined database of abnormal micro-behavior fingerprints that associates spoofed identities with sensitive areas and key nodes include: Based on a comprehensive assessment of the normal behavior patterns of the pre-set masquerading identity in different sensitive areas and key nodes, aerodynamic principles, aircraft design specifications, and the experience of security experts, an abnormal micro-behavior fingerprint database is established. The abnormal micro-behavior fingerprint database includes a set of logical rules. Each logical rule defines a subtle behavior pattern that does not conform to the pre-set masquerading identity's behavior logic in sensitive areas and key nodes, including hovering duration, trajectory deviation magnitude, and attitude change rate.

[0041] The pre-defined masquerading identity refers to the harmless entity identity imitated by the threatened aircraft, such as a civilian drone, a mapping drone, or a logistics drone. When establishing an abnormal micro-behavior fingerprint database, it is necessary to fully consider the normal flight behavior patterns of these masquerading identities in specific sensitive areas and key nodes. Specifically, normal behavior patterns refer to the flight trajectories, speeds, altitudes, and attitudes typically exhibited by aircraft with pre-defined masquerading identities within a specific airspace and time period. For example, when a civilian mapping drone is conducting mapping operations in a specific area, its flight trajectory may exhibit a regular grid pattern, with relatively stable flight altitude and speed.

[0042] Furthermore, the establishment of an abnormal micro-behavior fingerprint database also needs to be based on aerodynamic principles and aircraft design specifications. Aerodynamic principles are used to analyze the physical limits and reasonable behavioral range of aircraft under different flight conditions, such as the maximum hovering time and minimum turning radius that a specific type of aircraft can achieve at a specific wind speed. Aircraft design specifications provide structural, performance, and operational limitations for different types of aircraft; for example, some aircraft may not have the ability to perform sustained high-speed maneuvers. By combining these principles and specifications, it is possible to more accurately define what constitutes "normal" behavior, thereby identifying "abnormal" behaviors that exceed physical or design limitations.

[0043] This application's solution, by comprehensively considering pre-defined normal behavior patterns of masquerading as civilians, aerodynamic principles, aircraft design specifications, and security expert experience, can establish a more comprehensive and refined anomaly micro-behavior fingerprint database. This multi-dimensional, multi-source information fusion judgment mechanism enables the anomaly micro-behavior fingerprint database to capture subtle but crucial behavioral anomalies exhibited by threat aircraft mimicking harmless entities. By transforming these subtle behavioral patterns into specific logical rules and storing them in the anomaly micro-behavior fingerprint database, the system can more effectively identify potential threats attempting to evade traditional monitoring methods when subsequently matching target behavior data of aerial targets. For example, even if a threat aircraft attempts to disguise itself as a civilian drone, if its hovering time, trajectory deviation, or attitude change rate in a sensitive area does not conform to the behavioral logic of a normal civilian drone, it can be accurately captured by the logical rules in the anomaly micro-behavior fingerprint database, thereby triggering an early warning.

[0044] Specifically, the steps for extracting target behavior data of aerial targets in real time include: continuously receiving radar data, optoelectronic data and radio spectrum data of aerial targets through radar, cameras and radio spectrum equipment, and performing timestamp alignment and data fusion processing to extract target behavior data.

[0045] The radar equipment is configured to continuously detect the distance, speed, and azimuth of aerial targets and generate corresponding radar data. The camera equipment is configured to continuously capture visual images and videos of aerial targets, providing optoelectronic data, which may include visual representations of the target's shape, color, and trajectory. The radio spectrum equipment is configured to continuously monitor radio signals emitted by aerial targets, such as remote control signals or data link signals, and generate radio spectrum data, which can be used to identify target types or communication characteristics. Through the coordinated operation of these multiple sensor devices, comprehensive multi-dimensional information about aerial targets can be collected.

[0046] Furthermore, since the data acquisition frequencies and times of different sensor devices may differ, it is necessary to perform timestamp alignment on the received radar data, photoelectric data, and radio spectrum data. Timestamp alignment aims to ensure that all sensor data remain synchronized in the time dimension, providing a consistent foundation for subsequent data fusion processing. After timestamp alignment is completed, this multi-source heterogeneous data will undergo data fusion processing.

[0047] This application's solution overcomes the limitations of single-sensor data acquisition by deploying multiple sensors, including radar, cameras, and radio spectrum equipment, and performing timestamp alignment and data fusion processing on the collected radar, photoelectric, and radio spectrum data. By timestamp-aligning these different data types, data synchronization across time is ensured, avoiding analytical errors caused by data asynchrony. Subsequently, through data fusion processing, these complementary data are integrated into a unified and comprehensive description of target behavior, significantly improving the completeness, accuracy, and reliability of target behavior data, providing a solid data foundation for subsequent anomaly behavior judgment.

[0048] In some of the embodiments described above in this application, for the real-time extraction of target behavior data of aerial targets, when an aerial target enters a sensitive area or approaches a key node, the target behavior data is matched with an abnormal micro-behavior fingerprint database to determine whether the aerial target has abnormal behavior. In order to more accurately identify potential threats, this application further proposes a specific judgment mechanism.

[0049] Specifically, the steps mentioned above, which involve matching the target's behavior data with an abnormal micro-behavior fingerprint database to determine whether the aerial target exhibits abnormal behavior when it enters a sensitive area or approaches a critical node, include: When an aerial target enters a sensitive area or approaches a critical node, the conformity of the target's behavior data with multiple behavior fingerprints in the abnormal micro-behavior fingerprint database is evaluated. Fuzzy logic or a rule-based inference engine is used to evaluate the conformity. If the consistency between the target behavior data and at least one behavior fingerprint reaches a preset consistency threshold, it is determined that the aerial target has abnormal behavior.

[0050] Specifically, when assessing conformity, either fuzzy logic or rule-based inference engines can be employed. Fuzzy logic is a mathematical approach that handles uncertainty and fuzzy information, allowing the system to reason with imprecise or partial matches. With fuzzy logic, a conformity score can be calculated for each behavioral fingerprint, representing the degree to which the target behavior matches the fingerprint. Rule-based inference engines rely on a set of predefined "if-then" rules, such as, "If the hovering duration exceeds X seconds and the trajectory deviates by Y meters, then the conformity increases by Z." These rules can be configured by security experts based on experience and domain knowledge to capture specific anomalous behavioral patterns.

[0051] Furthermore, the "preset compliance threshold" is a configurable value used to define whether the degree of matching between the target behavior and the abnormal fingerprint is sufficient to trigger an anomaly judgment. When the compliance of the target behavior data with at least one behavior fingerprint reaches or exceeds this threshold, the system considers that the aerial target has abnormal behavior. This threshold can be adjusted according to the needs of the actual application scenario and the tolerance for false alarms and missed alarms.

[0052] This application's solution introduces a conformity assessment mechanism, combined with fuzzy logic or a rule-based inference engine, to make the judgment of anomalies in aerial target behavior more refined and intelligent. When an aerial target enters a sensitive area or approaches a critical node, its target behavior data is no longer simply matched "yes" or "no" to an anomalous micro-behavior fingerprint database, but rather a quantified conformity score is calculated. This conformity assessment can capture subtle differences and partial matches between target behavior and anomalous fingerprints, thus avoiding false negatives due to overly strict matching conditions or false positives due to overly lenient conditions. The use of fuzzy logic enables the system to handle the inherent uncertainty and ambiguity of target behavior data in the real world, while the rule-based inference engine incorporates expert knowledge into the anomaly judgment process, improving the accuracy and robustness of the judgment. Finally, by comparing the conformity score with a preset conformity threshold, the system can definitively determine whether an aerial target exhibits abnormal behavior in a controllable and adjustable manner.

[0053] In some of the embodiments described above in this application, an early warning is triggered when an aerial target exhibits abnormal behavior. However, in practical applications, simply triggering an early warning may not provide sufficient detailed information, nor can it ensure that the warning information is delivered to relevant security personnel in a timely and effective manner, which may affect the security personnel's rapid response and accurate handling of abnormal events.

[0054] In this regard, this application further proposes that the steps for triggering an early warning if the behavior of an aerial target is determined to be abnormal include: Trigger an alert for abnormal target behavior and generate alert information including the target's location, identity type, and description of abnormal micro-behavior. The alert information is sent to security personnel via screen display, mobile terminal push, SMS, or telephone notification.

[0055] Specifically, when the system detects abnormal behavior from an aerial target, it immediately triggers a target behavior anomaly warning. This warning is not a simple notification, but rather generates a structured warning message. This warning message is designed to include several key elements to provide comprehensive context awareness. Among them, "target location" refers to the real-time geographic coordinates or area description of the aerial target whose abnormal behavior was detected. "Identity type" refers to the preset camouflage identity that the aerial target is imitating. "Description of abnormal micro-behavior" details the specific abnormal behavior patterns exhibited by the aerial target.

[0056] Furthermore, to ensure that early warning information is delivered efficiently and reliably to the security personnel responsible for handling the situation, it can be sent through multiple communication channels. For example, "screen display" refers to displaying early warning information in real time on a large screen in the central monitoring room or on workstation monitors, facilitating centralized monitoring and command. "Mobile terminal push" refers to pushing early warning information to the smartphones or tablets of security personnel via a dedicated mobile application, ensuring they receive it promptly during patrols or on-site operations. "SMS or telephone notification" serves as a supplementary or emergency communication method, directly notifying security personnel through text messages or automated voice calls, particularly suitable for scenarios with poor network conditions or requiring mandatory reminders. These multi-channel delivery methods aim to maximize the delivery rate and timeliness of early warning information.

[0057] This application's solution effectively addresses the issues of insufficient detail and delayed delivery in basic solutions by immediately generating a warning message containing detailed contextual information upon detecting abnormal behavior in an aerial target and distributing it through multiple channels. Specifically, the generated warning message includes the target's location, identity type, and descriptions of abnormal micro-behaviors, enabling security personnel to quickly understand the location of the abnormal event, the camouflage nature of the potential threat, and the specific abnormal manifestations. This provides crucial information for their initial assessment and response strategy development. Furthermore, the delivery of warning messages through various methods, such as screen display, mobile terminal push notifications, SMS, or telephone notifications, ensures that security personnel receive warnings promptly and reliably regardless of their location or work status, avoiding response delays caused by limited communication channels.

[0058] In some of the embodiments described above in this application, when an abnormal behavior of an aerial target is determined, a target behavior anomaly warning is triggered, and warning information containing the target's location, identity type, and description of the abnormal micro-behavior is generated. This warning information is then sent to security personnel via screen display, mobile terminal push notification, SMS, or telephone notification. However, simply providing warning information may not be sufficient to guide security personnel in taking rapid and effective action, especially in emergency situations where additional decision support may be needed to take immediate action, potentially delaying the optimal response time.

[0059] In response, this application further proposes that the aforementioned early warning information also includes handling recommendations, which include dispatching patrol drones, activating directional jamming equipment, or notifying ground security personnel.

[0060] Specifically, the response recommendations refer to the specific countermeasures or action plans provided by the system to security personnel in response to detected abnormal behavior of low-altitude aircraft, based on preset strategies or real-time analysis results. The purpose is to assist security personnel in making rapid decisions and taking effective action to address potential threats. Dispatching patrol drones can be understood as the system automatically, or by instruction from security personnel, dispatching patrol drones with reconnaissance, tracking, or intervention capabilities to the target area after receiving a warning. Activating directional jamming equipment means that when an abnormal aircraft poses a direct threat to sensitive areas or key nodes, the system can recommend activating directional jamming equipment such as radio jammers or GPS spoofers to disrupt the abnormal aircraft's navigation, communication, or control capabilities, forcing it to deviate from its flight path or land. Notifying ground security personnel means that in some cases, abnormal behavior may require on-site handling by ground personnel, such as inspecting landed aircraft or questioning suspicious individuals.

[0061] The solution proposed in this application integrates handling suggestions into early warning information, enabling security personnel to immediately obtain recommended response measures for the current abnormal situation upon receiving the warning. This avoids delays in response due to insufficient information or hesitant decision-making by security personnel in emergency situations. Specifically, when the system determines that an aerial target is behaving abnormally and generates early warning information, it intelligently matches corresponding handling strategies based on the type of abnormal behavior, threat level, target location, and characteristics of sensitive areas or key nodes, and includes these strategies as handling suggestions in the early warning information.

[0062] The aforementioned low-altitude aircraft monitoring and early warning method identifies abnormal behavior by matching aerial target behavior data with a pre-set database of abnormal micro-behaviors. However, in practical applications, the legitimate flight behavior of aerial targets may be affected by complex and ever-changing environmental factors, such as wind speed, airflow, and obstacle distribution. Relying solely on a pre-set fingerprint database may not fully account for the impact of these dynamic environmental factors on legitimate flight behavior, potentially leading to misjudgments of normal flight behavior or missed detections of more camouflaged abnormal behavior. Especially in sensitive areas or near critical nodes, the identification of abnormal behavior requires higher accuracy and robustness.

[0063] In this regard, this application further proposes that when the aforementioned aerial target enters the aforementioned sensitive area or approaches the aforementioned key node, the above method also includes: Acquire and integrate environmental information to create a real-time environmental status map; Continuously receive the physical characteristics and flight information of the aforementioned aerial targets; Based on the physical characteristics and flight information of the aforementioned aerial targets and the aforementioned real-time environmental state diagram, the legal flight behavior of the aforementioned aerial targets under the aforementioned real-time environmental state diagram is simulated, and the predicted normal behavior trajectory is generated. Based on the target behavior data and the predicted normal behavior trajectory, calculate the behavior deviation score; If the above behavioral deviation score exceeds the preset deviation threshold, it is determined that the above aerial target has abnormal behavior and an early warning is triggered.

[0064] Specifically, acquiring and integrating environmental information to create a real-time environmental state map involves collecting and fusing various environmental parameters of the current area, including but not limited to wind speed, wind direction, temperature, humidity, air pressure, topography, building distribution, and temporary no-fly zones, through multiple sensors or data sources such as weather stations, geographic information systems (GIS), real-time traffic data, and information on temporary activity areas. This results in a dynamic three-dimensional model that reflects all relevant environmental elements within the area at the current moment. This real-time environmental state map provides necessary contextual information for subsequent simulations of legal flight behavior.

[0065] Continuously receiving the physical characteristics and flight information of the aforementioned aerial targets can be understood as acquiring, uninterruptedly through monitoring equipment such as radar, electro-optical devices, and radio spectrum data, including the target's real-time position, speed, acceleration, attitude, altitude, heading, payload information, communication signal characteristics, and possible model identification information. This data forms the basis for assessing the target's current behavior and predicting its future legitimate actions.

[0066] In practical applications, simulating the legal flight behavior of the aerial target under the aforementioned real-time environmental state diagram, based on its physical characteristics, flight information, and the real-time environmental state diagram, and generating a predicted normal behavior trajectory, refers to using a physics simulation engine, aerodynamic model, and flight control algorithm, combined with the target's known physical characteristics and real-time flight information, and under the constraints provided by the real-time environmental state diagram, to calculate and predict the legal and reasonable flight path and behavior pattern that the aerial target should follow in the current environment. The resulting predicted normal behavior trajectory provides a dynamic and personalized benchmark for judging whether the actual target's behavior is abnormal.

[0067] Furthermore, based on the aforementioned target behavior data and the predicted normal behavior trajectory, a behavior deviation score is calculated. This involves comparing real-time acquired aerial target behavior data (such as actual position, velocity, and attitude) with the predicted normal behavior trajectory obtained through simulation, quantifying the degree of difference between the two. This behavior deviation score can be calculated using various mathematical methods, such as Euclidean distance, cosine similarity, and dynamic time warping (DTW), to comprehensively reflect the degree of deviation between the target's actual behavior and the legitimate expected behavior in multiple dimensions, including space, time, and dynamics.

[0068] This application's solution constructs a dynamic, context-aware baseline for normal behavior by incorporating real-time environmental information and the target's own physical characteristics. Traditional methods may rely solely on a pre-set static anomaly fingerprint database. In complex and ever-changing environments, behavioral changes in aircraft to maintain stability or perform legitimate tasks may be misjudged as anomalous. This solution, however, simulates the legitimate flight behavior of an aerial target under the current real-time environmental state map, generating a predicted normal behavior trajectory that highly matches the current context. This dynamic baseline can more accurately reflect the reasonable range of target behavior in a specific environment. Subsequently, the real-time observed target behavior data is compared with this dynamically generated predicted normal behavior trajectory to calculate a behavior deviation score. This comparison method can effectively distinguish between normal behavior fluctuations caused by environmental influences and truly threatening anomalous behavior, thereby avoiding false alarms and improving the ability to identify novel or disguised anomalous behaviors.

[0069] Specifically, the steps described above for calculating the behavioral deviation score based on the target behavioral data and the predicted normal behavioral trajectory can be performed in the following manner.

[0070] The steps for calculating the behavioral deviation score based on the target behavioral data and the predicted normal behavioral trajectory include: Extract multi-dimensional behavioral feature vectors from target behavior data, including the real-time position, velocity, acceleration, attitude angle, rate of change of attitude angle, rate of change of profile, and signal feature intensity of aerial targets; Calculate the Euclidean distance between the multi-dimensional behavioral feature vector and the feature vector corresponding to the predicted normal behavioral trajectory; Calculate the cosine similarity between the multi-dimensional behavioral feature vector and the feature vector corresponding to the predicted normal behavioral trajectory; The behavioral deviation score is calculated by combining Euclidean distance and cosine similarity.

[0071] Specifically, extracting multi-dimensional behavioral feature vectors from target behavior data refers to parsing data from real-time monitored aerial target behavior data to extract multiple dimensions that comprehensively characterize its current flight state and intentions. These dimensions include, but are not limited to, the aerial target's real-time position (e.g., three-dimensional coordinates), velocity (including size and direction), acceleration (reflecting velocity changes), attitude angles (e.g., pitch, roll, yaw), rate of change of attitude angles (reflecting dynamic changes in attitude), rate of change of profile (e.g., changes in the aircraft's appearance or projected area obtained through visual sensors, which may indicate that it is performing some operation or changes in the objects it is carrying), and signal feature strength (e.g., radio signal strength, radar cross-section, etc., which can be used to identify or determine its communication status). The extraction of these feature vectors aims to provide a rich and detailed data foundation for subsequent behavioral deviation analysis.

[0072] The calculation of the Euclidean distance between the multi-dimensional behavioral feature vector and the feature vector corresponding to the predicted normal behavioral trajectory involves comparing the multi-dimensional behavioral feature vector of the aerial target extracted in real time with the feature vector of the predicted normal behavioral trajectory generated in advance at the same time point or under the same state. Euclidean distance, as a commonly used distance metric, can intuitively reflect the degree of absolute difference between two vectors in multi-dimensional space, i.e., the deviation magnitude of the behavioral features.

[0073] This application's solution extracts multi-dimensional behavioral feature vectors from target behavior data and calculates Euclidean distance and cosine similarity between these vectors and the feature vectors corresponding to predicted normal behavioral trajectories. This allows for the quantification of the degree of behavioral deviation of aerial targets from multiple dimensions and perspectives. Euclidean distance measures the absolute difference in behavioral features, while cosine similarity assesses the directional consistency of behavioral patterns. By combining these two measures, various abnormal behaviors that aerial targets may exhibit during flight can be captured more comprehensively and precisely, such as abnormal deviations in position, speed, and attitude, as well as potential changes in flight patterns or intentions, thus providing solid data support for subsequent anomaly assessment.

[0074] Secondly, see Figure 2 The specific embodiments of this application also disclose a low-altitude aircraft monitoring and early warning system, the system comprising: The fingerprint database establishment module 210 is used to delineate sensitive areas and key nodes in the city, and establish an abnormal micro-behavior fingerprint database that associates the preset disguised identity with the sensitive areas and key nodes. The preset disguised identity is the identity of a harmless entity that is imitated by the threatened aircraft. The behavior judgment module 220 is used to extract target behavior data of aerial targets in real time. When an aerial target enters a sensitive area or approaches a key node, the target behavior data is matched with the abnormal micro-behavior fingerprint database to determine whether the aerial target has abnormal behavior. The early warning triggering module 230 is used to trigger an early warning if it determines that the aerial target is behaving abnormally.

[0075] The system of this application provides a more refined monitoring and early warning mechanism through the collaborative operation of the fingerprint database establishment module 210, the behavior judgment module 220, and the early warning triggering module 230. The fingerprint database establishment module 210 can establish a targeted abnormal micro-behavioral fingerprint database in sensitive areas and key nodes based on preset masquerading identities. This allows the system to identify extremely subtle behavioral patterns that do not match the target's "masquerading identity." The behavior judgment module 220 can match aerial target behavior data with this fingerprint database in real time and efficiently, thereby identifying potential threats from seemingly normal flight data. The early warning triggering module 230 ensures that early warning information is promptly and accurately transmitted to security personnel after anomalies are detected. The system of this application significantly improves the ability to identify "highly deceptive" low-altitude aircraft, effectively compensating for the shortcomings of existing technologies in dealing with new threat strategies, thereby improving the overall effectiveness of urban low-altitude security and the timeliness and accuracy of early warnings.

[0076] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for monitoring and early warning of low-altitude aircraft, characterized in that, include: Sensitive areas and key nodes in the city are delineated, and an abnormal micro-behavior fingerprint database is established that associates a preset disguised identity with the sensitive areas and key nodes. The preset disguised identity is the identity of a harmless entity that is imitated by the threatened aircraft. Real-time extraction of target behavior data of aerial targets; when the aerial target enters the sensitive area or approaches the key node, the target behavior data is matched with the abnormal micro-behavior fingerprint database to determine whether the aerial target has abnormal behavior. If the aerial target is determined to exhibit abnormal behavior, an alert will be triggered.

2. The method for monitoring and early warning of low-altitude aircraft according to claim 1, characterized in that, The steps for delineating sensitive areas and key nodes in the city include: By importing urban planning data, architectural drawings, and distribution maps of important infrastructure, the sensitive areas and key nodes are delineated in three-dimensional space in the form of vector graphics and stored in the monitoring database.

3. The method for monitoring and early warning of low-altitude aircraft according to claim 1, characterized in that, The step of establishing a pre-defined database of abnormal micro-behavior fingerprints that associates the disguised identity with the sensitive area and the key node includes: Based on a comprehensive judgment of the normal behavior patterns of the preset masquerade identity in different sensitive areas and key nodes, aerodynamic principles, aircraft design specifications, and security expert experience, an abnormal micro-behavior fingerprint database is established. The abnormal micro-behavior fingerprint database includes a set of logical rules. Each logical rule defines a subtle behavior pattern that does not conform to the behavioral logic of the preset masquerade identity in the sensitive areas and key nodes, including hovering duration, trajectory deviation magnitude, and attitude change rate.

4. The method for monitoring and early warning of low-altitude aircraft according to claim 1, characterized in that, The steps for real-time extraction of target behavior data of aerial targets include: The system continuously receives radar data, optoelectronic data, and radio spectrum data from the aerial target via radar, cameras, and radio spectrum equipment, and performs timestamp alignment and data fusion processing to extract the target behavior data.

5. The method for monitoring and early warning of low-altitude aircraft according to claim 1, characterized in that, The step of matching the target behavior data with the abnormal micro-behavior fingerprint database to determine whether the aerial target exhibits abnormal behavior when the aerial target enters the sensitive area or approaches the key node includes: When the aerial target enters the sensitive area or approaches the key node, the conformity of the target behavior data with multiple behavior fingerprints in the abnormal micro-behavior fingerprint database is evaluated, wherein fuzzy logic or a rule-based inference engine is used to evaluate the conformity. If the target behavior data matches at least one of the behavior fingerprints to a preset matching threshold, then the aerial target is determined to have abnormal behavior.

6. The method for monitoring and early warning of low-altitude aircraft according to claim 5, characterized in that, The step of triggering an early warning if the aerial target is determined to exhibit abnormal behavior includes: The system triggers an alert for abnormal target behavior and generates an alert message containing the target's location, identity type, and description of the abnormal micro-behavior. The alert message is sent to security personnel via screen display, mobile terminal push notification, SMS, or telephone notification.

7. The method for monitoring and early warning of low-altitude aircraft according to claim 6, characterized in that, The early warning information also includes handling suggestions, which include dispatching patrol drones, activating directional jamming equipment, or notifying ground security personnel.

8. The method for monitoring and early warning of low-altitude aircraft according to claim 1, characterized in that, When the aerial target enters the sensitive area or approaches the critical node, the method further includes: Acquire and integrate environmental information to create a real-time environmental status map; Continuously receive the physical characteristics and flight information of the aerial target; Based on the physical characteristics and flight information of the aerial target and the real-time environmental state diagram, the legal flight behavior of the aerial target under the real-time environmental state diagram is simulated to generate a predicted normal behavior trajectory. Calculate the behavior deviation score based on the target behavior data and the predicted normal behavior trajectory; If the behavior deviation score exceeds a preset deviation threshold, it is determined that the aerial target has abnormal behavior and an early warning is triggered.

9. A method for monitoring and early warning of low-altitude aircraft according to claim 8, characterized in that, The step of calculating the behavior deviation score based on the target behavior data and the predicted normal behavior trajectory includes: Extract a multi-dimensional behavioral feature vector from the target behavior data, including the real-time position, velocity, acceleration, attitude angle, attitude angle change rate, contour change rate, and signal feature intensity of the aerial target; Calculate the Euclidean distance between the multi-dimensional behavioral feature vector and the feature vector corresponding to the predicted normal behavioral trajectory; Calculate the cosine similarity between the multi-dimensional behavioral feature vector and the feature vector corresponding to the predicted normal behavioral trajectory; The behavioral deviation score is calculated by combining the Euclidean distance and the cosine similarity.

10. A low-altitude aircraft monitoring and early warning system, characterized in that, The system includes: The fingerprint database establishment module is used to delineate sensitive areas and key nodes in the city, and establish an abnormal micro-behavior fingerprint database that associates a preset disguised identity with the sensitive areas and key nodes. The preset disguised identity is the identity of a harmless entity that is imitated by a threatened aircraft. The behavior judgment module is used to extract target behavior data of aerial targets in real time. When the aerial target enters the sensitive area or approaches the key node, the target behavior data is matched with the abnormal micro-behavior fingerprint database to determine whether the aerial target has abnormal behavior. The early warning triggering module is used to trigger an early warning if it determines that the aerial target is behaving abnormally.

Images