Ship target correlation parameter generation method and device, electronic equipment, medium and chip

Abstract

The embodiment of the invention provides a ship target correlation parameter generation method and device, electronic equipment, a medium and a chip, and relates to the technical field of maritime affair supervision, and the ship target correlation parameter generation method comprises the steps: obtaining multi-source monitoring data of a target water area; determining at least one first detection target according to the satellite image data; determining at least one second detection target according to the ship automatic identification system data; determining at least one associated target result for the first detection target and the second detection target according to the first position data and the second position data; determining at least one multi-dimensional risk feature parameter for each associated target result; obtaining a preset weight coefficient; and performing weighted fusion calculation on the at least one multi-dimensional risk feature parameter according to a preset weight coefficient, and determining an associated feature parameter of the target ship corresponding to the associated target result. According to the scheme, the reliability of credibility judgment in the ship monitoring process is improved.

Description

Technical Field

[0001] This invention relates to the field of maritime regulatory technology, and more specifically, to a method, apparatus, electronic device, medium, and chip for generating ship target association parameters. Background Technology

[0002] In technologies utilizing Automatic Identification Systems (AIS) for ship monitoring, the core approach of multi-source data fusion is to achieve cross-validation through spatiotemporal matching of Synthetic Aperture Radar (SAR) detection and AIS reports. Related technical solutions generally employ a pre-set fixed spatiotemporal threshold for matching, and their judgment logic is essentially binary: if a matching AIS signal is found within a certain spatiotemporal range, the target ship is deemed trustworthy; otherwise, it is marked as untrustworthy.

[0003] However, this binary verification logic, which relies on the presence or absence of a signal, assumes the authenticity of the AIS signal content. It cannot effectively identify fraudulent activities such as falsifying or altering key information about a ship's identity, size, and location when the AIS is activated. The determination of the reliability of the data source is based on a single dimension, making it difficult to cope with increasingly complex maritime deception. Summary of the Invention

[0004] The purpose of this invention is to provide a method, apparatus, electronic device, readable storage medium, and chip for generating ship target association parameters, which can solve the problem of relying on a single dimension for determining the reliability of data sources when using Automatic Identification System (AIS) for ship monitoring.

[0005] In view of this, an embodiment of the first aspect of the present invention provides a method for generating ship target association parameters.

[0006] A second aspect of the present invention provides a device for generating ship target association parameters.

[0007] An embodiment of the third aspect of the present invention provides an electronic device.

[0008] An embodiment of the fourth aspect of the present invention provides a readable storage medium.

[0009] An embodiment of the fifth aspect of the present invention provides a chip.

[0010] To achieve the above objectives, an embodiment of the first aspect of the present invention provides a method for generating ship target association parameters. This method is used in a ship monitoring system to enable the system to screen, sort, or perform subsequent processing on at least one target ship. The method includes: acquiring multi-source monitoring data of a target water area, the multi-source monitoring data including at least satellite imagery data and automatic identification system (AIS) data; determining at least one first detection target and corresponding first location data based on the satellite imagery data; determining at least one second detection target, corresponding second location data, and ship attribute data based on the AIS data; performing spatiotemporal association matching on the first and second detection targets based on the first and second location data to determine at least one associated target result; for each associated target result, determining at least one multidimensional risk feature parameter based on the satellite imagery data and AIS data corresponding to the associated target result; acquiring preset weight coefficients; and performing weighted fusion calculation on the at least one multidimensional risk feature parameter based on the preset weight coefficients to determine the association feature parameters of the target ship corresponding to the associated target result.

[0011] The core of the ship target association parameter generation method provided by this invention lies in calculating the credibility of ship behavior through multi-source data fusion and quantification model.

[0012] The method for generating ship target association parameters first simultaneously acquires multi-source monitoring data, including satellite imagery and AIS-reported targets, and identifies the image-detected targets and AIS-reported targets, along with their location information. Subsequently, spatiotemporal correlation matching is used to precisely pair the same ship targets from different sources, forming associated target results.

[0013] For each target pair, risk feature values ​​reflecting multiple dimensions such as identity consistency and behavioral rationality are further extracted and calculated from the corresponding multi-source data.

[0014] In some technical solutions, optionally, at least one multidimensional risk characteristic parameter includes identity consistency characteristic value, behavioral rationality characteristic value, and location deviation characteristic value.

[0015] In this scheme, the identity consistency feature value, behavior rationality feature value and position deviation feature value are used to form a multi-dimensional risk feature parameter, and the ship's risk is quantitatively analyzed from three core dimensions: static identity, dynamic behavior and real-time positioning.

[0016] In some technical solutions, optionally, at least one multidimensional risk characteristic parameter is determined based on the satellite imagery data and the Automatic Identification System (AIS) data corresponding to the associated target result, including: determining the observation size data of the target vessel based on the satellite imagery data; determining the reported size data of the target vessel based on the vessel attribute data; calculating an identity consistency characteristic value based on the observation size data and the reported size data; obtaining the historical navigation trajectory of the target vessel from the AIS data corresponding to the associated target result; obtaining a historical trajectory dataset; comparing the historical navigation trajectory with a preset trajectory in the historical trajectory dataset to determine a behavior rationality characteristic value; calculating the original position deviation based on the first position data and the second position data corresponding to the associated target result; obtaining a regional attribute parameter corresponding to the current position of the target vessel, the regional attribute parameter being determined based on auxiliary information data, the auxiliary information data including electronic charts and preset sensitive area information; and determining a position deviation characteristic value based on the original position deviation and the regional attribute parameter.

[0017] This solution constructs a complete and executable computational chain from multi-source data to the three major risk characteristic values.

[0018] In some technical solutions, optionally, a weighted fusion calculation is performed on at least one multidimensional risk feature parameter based on a preset weight coefficient to determine the associated feature parameters of the target vessel corresponding to the associated target result. This includes: obtaining a preset benchmark score; obtaining a first weight coefficient corresponding to the identity consistency feature value; weighting the identity consistency feature value according to the first weight coefficient to determine a first weighted value; obtaining a second weight coefficient corresponding to the behavioral rationality feature value; weighting the behavioral rationality feature value according to the second weight coefficient to determine a second weighted value; obtaining a third weight coefficient corresponding to the position deviation feature value; weighting the position deviation feature value according to the third weight coefficient to determine a third weighted value; and determining the associated feature parameters of the target vessel based on the benchmark score, the first weighted value, the second weighted value, and the third weighted value.

[0019] In this scheme, a benchmark score is introduced as the starting point for calculation, and preset weight coefficients are assigned to the risk characteristic values ​​of the three dimensions of identity consistency, behavior rationality and location deviation. By weighted summation (or weighted deduction), the risk quantification values ​​of different natures and scales are integrated into a comprehensive scalar score.

[0020] In some technical solutions, optionally, after determining the associated feature parameters, the method for generating ship target associated parameters further includes: obtaining multiple preset risk level thresholds; comparing the associated feature parameters with the multiple risk level thresholds to determine the comparison result; and determining the risk level corresponding to the target ship based on the comparison result.

[0021] In this scheme, multiple preset risk level thresholds are obtained, the quantitative score is compared with these thresholds, and the specific risk level corresponding to the target vessel is determined based on the comparison results.

[0022] In some technical solutions, optionally, after determining the associated characteristic parameters of the target vessel corresponding to the associated target result, the method for generating vessel target association parameters further includes: acquiring preset response strategy data; determining the target response strategy from the preset response strategy data according to the risk level; and executing risk disposal operations corresponding to the target response strategy.

[0023] In this solution, the quantitative risk level generated in the preceding steps is directly converted into specific regulatory instructions, completing the entire process from risk identification to risk management with full automation, which significantly improves the response speed, execution standardization and overall efficiency of maritime supervision.

[0024] A second aspect of the present invention provides a vessel target association parameter generation device, comprising: a data acquisition module for acquiring multi-source monitoring data of a target water area, the multi-source monitoring data including at least satellite image data and automatic identification system (AIS) data; an image detection module for determining at least one first detection target and corresponding first location data based on the satellite image data; a data identification module for determining at least one second detection target, corresponding second location data, and vessel attribute data based on the AIS data; an association matching module for performing spatiotemporal association matching on the first and second detection targets based on the first and second location data to determine at least one associated target result; a risk feature module for determining at least one multidimensional risk feature parameter for each associated target result based on the satellite image data and AIS data corresponding to the associated target result; a weight acquisition module for acquiring preset weight coefficients; and a scoring and determination module for performing weighted fusion calculation on the at least one multidimensional risk feature parameter based on the preset weight coefficients to determine the association feature parameter of the target vessel corresponding to the associated target result.

[0025] An embodiment of the third aspect of the present invention provides an electronic device, including a processor, a memory, and a program or instructions stored in the memory and executable on the processor, wherein the program or instructions, when executed by the processor, implement the steps of the ship target association parameter generation method as described in the first aspect.

[0026] An embodiment of the fourth aspect of the present invention provides a readable storage medium on which a program or instructions are stored, which, when executed by a processor, implement the steps of the ship target association parameter generation method of the first aspect.

[0027] An embodiment of the fifth aspect of the present invention provides a chip including a processor and a communication interface, the communication interface and the processor being coupled together, the processor being used to run a program or instructions to implement the steps of the ship target association parameter generation method as described in the first aspect.

[0028] Additional aspects and advantages of the technical solutions of the present invention will become apparent in the following description or may be learned by practice of the invention. Attached Figure Description

[0029] Figure 1 One of the flowcharts for generating ship target association parameters according to the present invention is shown;

[0030] Figure 2 A second schematic flowchart of the method for generating ship target association parameters according to the present invention is shown;

[0031] Figure 3 A third schematic flowchart of the method for generating ship target association parameters according to the present invention is shown;

[0032] Figure 4 A fourth flowchart illustrating the method for generating ship target association parameters according to the present invention is shown;

[0033] Figure 5 Fifth of the flowcharts illustrating the method for generating ship target association parameters according to the present invention is shown;

[0034] Figure 6 A schematic block diagram of the ship target association parameter generation device according to the present invention is shown;

[0035] Figure 7 A schematic block diagram of the structure of the electronic device according to the present invention is shown;

[0036] Figure 8 A schematic diagram of the overall system architecture according to the present invention is shown;

[0037] Figure 9 A flowchart of data preprocessing and adaptive matching according to the present invention is shown;

[0038] Figure 10 A system interface diagram according to the present invention is shown.

[0039] Among them, 900: Ship target association parameter generation device; 902: Data acquisition module; 904: Image detection module; 906: Data recognition module; 908: Association matching module; 910: Risk feature module; 912: Weight acquisition module; 914: Scoring and determination module; 1000: Electronic device; 1109: Memory; 1110: Processor. Detailed Implementation

[0040] To better understand the above-described objectives, features, and advantages of the embodiments of the present invention, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, unless otherwise specified, the embodiments of the present invention and the features thereof can be combined with each other.

[0041] In the field of using maritime data (especially synthetic aperture radar SAR and automatic identification system AIS) for ship monitoring, behavior analysis and safety enforcement, the relevant technologies mainly revolve around data acquisition, target identification and preliminary correlation of multi-source data. The implementation schemes can be summarized into the following typical paths. However, these schemes all have significant limitations in achieving the core law enforcement and safety requirement of determining the credibility of ship signals (behaviors).

[0042] The relevant technologies have the following main technical shortcomings in determining the reliability of ship signals to support precise law enforcement and safety decision-making:

[0043] 1. The data source reliability relies on a single assumption, making it unable to cope with systemic fraud and deception:

[0044] In practice, AIS signals can be easily turned off, tampered with (e.g., falsifying ship codes, locations, dimensions, etc.), or used for "AIS spoofing," making it impossible to identify "AIS information content fraud" (i.e., turning on AIS but reporting false attributes). For example, a smuggling vessel might turn on AIS but disguise itself as a cargo ship and maintain a compliant AIS signal with its berthed mother ship; its fraudulent behavior cannot be detected by simple matching.

[0045] 2. The fusion of multi-source data is coarse, failing to address the underlying spatiotemporal and feature mismatch issues:

[0046] SAR images suffer from geometric distortions caused by satellite perspective and ocean motion, while AIS itself has positioning errors. Related methods use fixed spatiotemporal thresholds for matching, failing to consider the non-uniform distribution of these errors, leading to numerous false matches (associating different vessels) or false matches (failing to associate the same vessel). Furthermore, simple size or type comparisons are too coarse and cannot handle deliberate camouflage or measurement errors.

[0047] 3. The judgment logic remains at the level of "whether detection is present or not," lacking a comprehensive quantitative evaluation model for "credibility":

[0048] The output of related technologies is usually: "A target was detected," "The target does not match AIS," or "The target's attributes are inconsistent with AIS." This is a qualitative, black-and-white judgment, lacking a comprehensive judgment model that integrates multi-dimensional evidence and can output a quantitative confidence score.

[0049] 4. The technology is disconnected from business scenarios, making it difficult to directly empower law enforcement decision-making and report generation:

[0050] The relevant technologies are mostly independent "target detection" or "data association" modules, and their outputs are scattered technical indicators. They are not deeply integrated with the specific workflows of law enforcement, security checks, or report generation.

[0051] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, embodiments of the invention may be implemented in other ways different from those described herein. Therefore, the scope of protection of the invention is not limited to the specific embodiments disclosed below.

[0052] The following detailed description of the ship target association parameter generation method, apparatus, electronic device, readable storage medium, and chip provided by the embodiments of the present invention, with reference to specific examples and application scenarios.

[0053] This embodiment provides a method for generating ship target association parameters. This method is used in a ship monitoring system to enable the system to filter, sort, or perform subsequent processing on at least one target ship. Figure 1 As shown, the method for generating ship target association parameters includes:

[0054] Step S100: Acquire multi-source monitoring data of the target water area. The multi-source monitoring data shall include at least satellite imagery data and Automatic Identification System (AIS) data.

[0055] Step S102: Determine at least one first detection target and the first location data corresponding to the first detection target based on satellite imagery data;

[0056] Step S104: Determine at least one second detection target, the second location data corresponding to the second detection target, and the ship attribute data based on the Automatic Identification System (AIS) data;

[0057] Step S106: Perform spatiotemporal correlation matching on the first detection target and the second detection target based on the first location data and the second location data to determine at least one associated target result;

[0058] Step S108: For each associated target result, determine at least one multidimensional risk characteristic parameter based on the satellite imagery data and Automatic Identification System (AIS) data corresponding to the associated target result;

[0059] Step S110: Obtain the preset weight coefficients;

[0060] Step S112: Perform weighted fusion calculation on at least one multidimensional risk feature parameter according to the preset weight coefficient to determine the associated feature parameter of the target vessel corresponding to the associated target result.

[0061] The core of the ship target association parameter generation method provided by this invention lies in calculating the credibility of ship behavior through multi-source data fusion and quantification model.

[0062] The method for generating ship target association parameters first simultaneously acquires multi-source monitoring data, including satellite imagery and AIS-reported targets, and identifies the image-detected targets and AIS-reported targets, along with their location information. Subsequently, spatiotemporal correlation matching is used to precisely pair the same ship targets from different sources, forming associated target results.

[0063] For each target pair, risk feature values ​​reflecting multiple dimensions such as identity consistency and behavioral rationality are further extracted and calculated from the corresponding multi-source data.

[0064] Finally, by using preset weighting coefficients to perform weighted fusion calculations on these multidimensional risk characteristic parameters, a quantitative and continuous correlation characteristic parameter is output, thereby upgrading the traditional black-and-white qualitative judgment to a refined dynamic risk assessment.

[0065] Understandably, this invention introduces deep fusion analysis of multi-source data such as satellite imagery and automatic identification systems for ships into the field of ship monitoring, and constructs a complete intelligent judgment chain from target detection and matching to behavioral correlation feature parameters through a quantitative model to achieve a paradigm shift from qualitative alarms to quantitative risk assessment.

[0066] Even when a ship may activate AIS but report false information to commit systematic fraud, it can still accurately quantify its behavioral contradictions and risk levels through multi-dimensional feature analysis. This solves the limitations of existing technologies, which can only make binary judgments based on whether or not the information content is matched, cannot deal with information content fraud, and lack quantitative output to support accurate decision-making. This improves the reliability of credibility judgment during ship monitoring.

[0067] First, acquire multi-source monitoring data for the target water area.

[0068] For example, the target waters refer to a specific sea area where ship monitoring is required, such as an anchorage outside a port, an important waterway, or the boundary of a sensitive water area.

[0069] Multi-source monitoring data refers to a heterogeneous set of data obtained from different technical means that can reflect the activities of ships in the same water area.

[0070] In this invention, the multi-source monitoring data includes at least satellite imagery data, Automatic Identification System (AIS) data, and other auxiliary information.

[0071] Supporting information includes, but is not limited to: electronic nautical charts, port lists, and meteorological information.

[0072] By introducing auxiliary information data containing environmental and contextual information, richer judgment basis is provided for subsequent correlation matching and risk assessment, improving the system's ability to understand complex scenarios.

[0073] Satellite imagery data specifically refers to remote sensing images acquired by synthetic aperture radar satellites. These images have the ability to penetrate clouds and image at night, and can actively detect ship targets on the sea surface and form images containing information about the target's location and outline.

[0074] Automatic Identification System (AIS) data refers to radio message data automatically broadcast and received by the AIS equipment on board a ship. It includes the ship's identification code, real-time latitude and longitude position, speed over land, heading over land, and static attribute information such as ship type, length, and beam.

[0075] Based on satellite imagery data, at least one first detection target and its corresponding first location data are determined.

[0076] By processing satellite imagery with a ship target detection algorithm, each set of pixels identified as a ship in the imagery is called a first detection target.

[0077] The image coordinates of each first detection target are calculated and converted into real geographic coordinates through a geolocation model. This coordinate value is the first location data corresponding to the first detection target.

[0078] Simultaneously, based on data from the Automatic Identification System (AIS), at least one second detection target, the second location data corresponding to the second detection target, and the ship attribute data are determined.

[0079] Each ship that sends an AIS message is defined as a secondary detection target.

[0080] The latitude and longitude information directly parsed from the message constitutes the vessel's second location data. The vessel's maritime mobility service identification code, name, type, length, and beam, among other information parsed from the message, together form the vessel attribute data of the second detection target.

[0081] Then, based on the first location data and the second location data, spatiotemporal correlation matching is performed on the first detection target and the second detection target to determine at least one associated target result.

[0082] This step is the core of data fusion, and its purpose is to determine whether the first detection target seen in satellite imagery and the second detection target identified from AIS signals correspond to the same physical ship.

[0083] The association matching process considers both spatial proximity and temporal synchronization.

[0084] For each first detected target, the system searches among AIS targets with similar time windows for targets whose first location data and the second location data of the other target are spatially less than a preset threshold. If found, the two are considered successfully associated, forming an associated target result. An associated target result contains a first detected target from satellite imagery and a second detected target from the AIS system, representing evidence from two different sources that the system believes represent the same ship.

[0085] For example, the first detection target and the second detection target correspond to the same target ship.

[0086] Specifically, the first detection target and the second detection target are spatiotemporally correlated and matched based on the first location data and the second location data, including dynamically adjusting the spatiotemporal threshold of the matching based on auxiliary information data.

[0087] A first matching threshold is used in areas with complex traffic conditions such as ports and narrow waterways, while a second matching threshold is used in open sea areas. The value corresponding to the first matching threshold is less than the value corresponding to the second matching threshold. In other words, the requirements of the first matching threshold are more stringent than those of the second matching threshold, thereby improving the adaptability to various aquatic environments while ensuring the accuracy of the association.

[0088] For each associated target result, at least one multidimensional risk characteristic parameter is determined based on the satellite imagery data and Automatic Identification System (AIS) data corresponding to the associated target result.

[0089] The system extracts and quantifies features from the original data associated with the target pair, forming a series of values ​​that characterize different aspects of anomalies or risks, namely multidimensional risk feature parameters.

[0090] For example, by analyzing the pixel size of the vessel target in satellite imagery and comparing it with the length information reported by the vessel in the AIS message, a size consistency feature value is calculated; by analyzing whether the vessel's historical AIS trajectory deviates from the regular shipping lane, a behavioral compliance feature value is calculated; and by calculating the precise deviation between the target's aligned satellite position and its reported AIS position, a position drift feature value is calculated. These feature values ​​characterize the degree of suspicion of the vessel's behavior from different perspectives.

[0091] Obtain preset weight coefficients. The preset weight coefficients are a set of values ​​set in advance through model training or expert experience, and each weight coefficient corresponds to one of the aforementioned multidimensional risk feature parameters.

[0092] The magnitude of the weighting coefficient represents the importance of that dimension's risk characteristic in the overall credibility assessment. For example, in sensitive waters, the location drift characteristic may be given a higher weight because even minor location deception can pose a significant risk.

[0093] Finally, based on preset weighting coefficients, at least one multidimensional risk feature parameter is weighted and fused to determine the associated feature parameters of the target vessel corresponding to the associated target result. All previously obtained multidimensional risk feature parameters are multiplied by their respective preset weighting coefficients to finally calculate a comprehensive numerical result, namely the associated feature parameters of the target vessel.

[0094] The correlation characteristic parameter is a quantitative indicator, such as a score between 0 and 100. A higher score indicates a higher level of credibility and more normal behavior of the vessel based on multi-source information fusion analysis; a lower score indicates a higher risk of contradictory, abnormal, or deceptive behavior. This correlation characteristic parameter is no longer a binary judgment of normal or suspicious, but provides maritime law enforcement personnel with a continuous and hierarchical basis for decision-making.

[0095] In some embodiments, the acquisition of multi-source monitoring data for the target water area may be automatically triggered based on an external event or risk intelligence. External events include received collaborative monitoring requests, risk level upgrade instructions for a specific area, or relay tracking instructions for suspicious targets from other monitoring systems.

[0096] By introducing an external triggering mechanism, the credibility determination is expanded from a continuous monitoring mode to a proactive response mode that is activated on demand and focuses on key areas, thereby saving computing resources while achieving an efficient response to sudden high-risk events.

[0097] In some embodiments, the associated feature parameter is optionally not a single scalar value, but a structured scoring report that includes a total score, scores for each risk dimension, and their confidence intervals. By providing a structured score with confidence intervals, not only are comprehensive conclusions given, but the reliability of the score itself and the contributing factors of each risk are also revealed, providing decision-makers with a deeper and more interpretable basis for judgment, enabling them to make more prudent judgments when dealing with boundary risks.

[0098] In some embodiments, the ship target association parameter generation method can optionally not only determine target pairs at a single point in time, but also perform time-series analysis on a series of monitoring results of a target ship over a continuous period of time. By calculating and analyzing the evolution trend, fluctuation characteristics, or abrupt change points of the associated characteristic parameters of the target ship over a period of time, more covert and gradual behavioral anomaly patterns (such as slow drift deception) can be identified, thereby gaining insight into dynamic fraud strategies.

[0099] In some embodiments, optionally, after determining the associated characteristic parameters of the target vessel, the vessel target associated parameter generation method further includes: automatically pushing the associated characteristic parameters and the associated target identifier to one or more designated external regulatory or business systems according to a preset data interface specification.

[0100] External systems include vessel traffic management systems, port state control vessel selection systems, and marine insurance risk assessment platforms. Through standardized interface outputs, a closed loop is formed from technological awareness to business decision-making, proactively empowering enforcement, safety inspection, and report generation processes, thereby improving the intelligence and efficiency of maritime supervision.

[0101] In some embodiments, optionally, at least one multidimensional risk characteristic parameter includes identity consistency characteristic value, behavioral rationality characteristic value, and location deviation characteristic value.

[0102] In this embodiment, multidimensional risk characteristic parameters are constructed based on identity consistency characteristic value, behavior rationality characteristic value and position deviation characteristic value, and quantitative risk analysis of ships is carried out from three core dimensions: static identity, dynamic behavior and real-time positioning.

[0103] Understandably, this invention constructs a comprehensive quantitative assessment framework that ranges from static identity authenticity and dynamic behavioral patterns to real-time location integrity by specifically defining the multidimensional risk feature parameters as three core dimensions: identity consistency, behavioral rationality, and location deviation. This transforms abstract risk analysis into concrete and calculable feature comparison.

[0104] By fusing and calculating the features of these three dimensions, a solid and structured data foundation is provided for outputting continuous and interpretable correlation feature parameters, which significantly improves the ability of maritime monitoring to transform from passive alarms to proactive intelligent analysis.

[0105] Among them, at least one multidimensional risk characteristic parameter includes identity consistency characteristic value, behavioral rationality characteristic value, and position deviation characteristic value. These three characteristic values ​​constitute the core quantitative evidence system for conducting comprehensive and multi-faceted risk assessment of the target vessel, respectively approaching the issue from the three dimensions of static identity authenticity, dynamic behavioral patterns, and real-time position integrity, jointly supporting the final comprehensive credibility judgment.

[0106] For example, identity consistency eigenvalue is an indicator used to quantify the degree of consistency between a ship’s claimed static identity attributes and its observed physical characteristics.

[0107] For example, the behavioral rationality characteristic value is an indicator used to quantitatively assess whether a vessel's dynamic navigation trajectory and behavioral patterns conform to the usual operating or navigation practices of vessels of a certain type in a given waterway.

[0108] For example, the position deviation eigenvalue is an indicator used to quantify the severity of the deviation between the position reported by the Automatic Identification System (AIS) and its position independently detected by satellite imagery.

[0109] In summary, the identity consistency feature value, behavior rationality feature value, and location deviation feature value provide quantifiable contradiction detection results from the three fundamental questions of who you are, how you act, and where you are.

[0110] These features complement each other, enabling the system to not only detect simple anomalies such as AIS signal loss, but also to accurately identify complex fraudulent activities, laying a solid multi-dimensional evidentiary foundation for building an in-depth and sophisticated credibility assessment model.

[0111] In some embodiments, optionally, such as Figure 2 As shown, at least one multidimensional risk characteristic parameter is determined based on the satellite imagery data and the Automatic Identification System (AIS) data corresponding to the associated target results, including:

[0112] Step S1080: Determine the observation dimensions of the target vessel based on satellite imagery data;

[0113] Step S1082: Determine the reported size data of the target vessel based on the vessel attribute data;

[0114] Step S1084: Calculate the identity consistency feature value based on the observed size data and the reported size data;

[0115] Step S1086: Obtain the historical navigation trajectory of the target vessel from the Automatic Identification System (AIS) data corresponding to the associated target results;

[0116] Step S1088: Obtain historical trajectory dataset;

[0117] Step S1090: Compare the historical navigation trajectory with the preset trajectory in the historical trajectory dataset to determine the rationality feature value of the behavior;

[0118] Step S1092: Calculate the original position deviation based on the first and second position data corresponding to the associated target results;

[0119] Step S1094: Obtain the area attribute parameters corresponding to the current position of the target vessel. The area attribute parameters are determined based on auxiliary information data, which includes electronic nautical charts and preset sensitive area information.

[0120] Step S1096: Determine the position deviation characteristic value based on the original position deviation and regional attribute parameters.

[0121] In this embodiment, a complete and executable computational chain is constructed from multi-source data to the three major risk characteristic values.

[0122] First, by comparing the satellite observation size with the AIS report size, identity consistency is quantified into a calculable difference value. Second, by matching the ship's historical trajectory with preset typical patterns, behavioral rationality is transformed into a measurable similarity value. Finally, by combining the original positional deviations between AIS and satellite data with regional sensitivity coefficients determined based on auxiliary information such as electronic charts, positional deviations are transformed into a scenario-based risk quantification value. This generates specific and interpretable structured evidence for the system, including identity suspicion, abnormal behavior, and position fraud.

[0123] Understandably, this invention provides a clear and executable calculation path and quantitative basis for multidimensional risk analysis by specifying detailed calculation steps for the three characteristic values ​​of identity consistency, behavioral rationality, and location deviation. This enables the system to produce structured quantitative evidence (such as "SAR size difference of 45%" and "location deviation of 2.3 nautical miles"), thus laying a solid and transparent technical foundation for the final generation of interpretable and traceable associated characteristic parameters. This promotes a profound transformation in maritime monitoring from experience-based judgment to data-driven approaches and from fuzzy qualitative analysis to precise quantitative analysis, improving the objectivity and refinement of the assessment.

[0124] Specifically, based on the satellite imagery data, Automatic Identification System (AIS) data, and the ship attribute data contained therein corresponding to the associated target results, at least one multidimensional risk characteristic parameter is determined. The specific implementation process is as follows, aiming to transform abstract risk quantification into concrete and executable calculation steps:

[0125] First, determine the identity consistency feature value. This process is based on the two types of data associated with the already associated target results.

[0126] On the one hand, the system uses image processing and target measurement algorithms to deduce the physical size of the ship target from the pixel outline and size of the ship target in the satellite image based on the satellite image data corresponding to the associated target results, and obtains the observed size data of the target ship, such as the estimated ship length.

[0127] On the other hand, the system parses the ship attribute data from the Automatic Identification System (AIS) data corresponding to the same associated target result, and extracts the report size data of the target ship broadcast by the ship itself through the AIS device, such as the length field in the AIS message.

[0128] Subsequently, the system substitutes the observed size data and the reported size data into the preset difference calculation formula.

[0129] For example, the formula for calculating the difference includes, but is not limited to, the absolute error formula. For instance, the identity consistency feature value can be calculated using the absolute relative error formula: |observation size - report size| / report size.

[0130] Identity consistency features directly quantify the degree of discrepancy between a ship’s physical appearance and its claimed identity.

[0131] Secondly, determine the behavioral rationality characteristic values. This process focuses on the analysis of the ship's dynamic navigation patterns.

[0132] The system first extracts the ship's continuous position, speed, and heading records over a period of time from the Automatic Identification System (AIS) data corresponding to the associated target results, forming the target ship's historical navigation trajectory.

[0133] Meanwhile, the system obtains preset trajectories from a pre-built historical trajectory dataset that match the current waters, vessel type, and time (such as season). These preset trajectories represent legal and conventional navigation patterns under these conditions.

[0134] Next, the system uses trajectory matching algorithms (such as dynamic time warping, Fraser distance calculation, etc.) to compare the similarity of the target ship's historical navigation trajectory with the preset trajectory in the historical trajectory dataset.

[0135] Based on the similarity score or deviation measure obtained from the comparison results, the system determines a quantitative characteristic value for the rationality of the behavior.

[0136] The behavioral rationality characteristic value reflects the degree of consistency between the current ship behavior and the historical normal pattern, and is the core basis for discovering dynamic risks such as abnormal wandering, deviation from the course, and illegal operation.

[0137] Finally, the position deviation characteristic value is determined. This process is used to assess the immediate reliability of the position reported by the Automatic Identification System (AIS).

[0138] The system first calculates the straight-line distance between the first detected target (i.e., the satellite detection position after geometric correction) and the second detected target (i.e., the AIS reported position) in the associated target results to obtain the original position deviation.

[0139] At the same time, the system needs to assess the severity of this deviation in conjunction with the environmental context. To this end, the system obtains the area attribute parameters corresponding to the current position of the target vessel.

[0140] The regional attribute parameters are determined based on auxiliary information data, which specifically includes electronic nautical charts and preset sensitive area information.

[0141] For example, the current location can be determined from the electronic nautical chart whether it is open water, a narrow channel, or a port anchorage; and whether the current location is close to a military restricted area, a marine protected area, or a maritime facility can be determined from the preset sensitive area information.

[0142] The system assigns a sensitivity coefficient to the area based on this information (e.g., a high coefficient in ports and a low coefficient in the open sea).

[0143] Ultimately, the location deviation characteristic value is not simply equal to the original location deviation, but is determined jointly by the original location deviation and regional attribute parameters, usually by multiplying the original location deviation by a regional sensitivity coefficient. This means that in sensitive areas, even a small physical location deviation will produce a high location deviation characteristic value, thus carrying greater weight in credibility assessments and accurately depicting the regulatory logic of the same deviation but different risks.

[0144] Through the three specific and coherent calculation steps described above, the three core multidimensional risk characteristic parameters are transformed from concepts into practically calculable indicators, thus constructing a complete risk evidence production pipeline.

[0145] In some embodiments, optionally, identity consistency feature values ​​are calculated based on observed size data and reported size data, and the calculation model used is compatible with size reporting specification differences for various ship types.

[0146] For example, the system adapts different length estimation algorithms and error tolerance ranges for different types of ships (such as container ships, bulk carriers, and tankers), thereby improving the universality and accuracy of identity consistency judgment across different ship types.

[0147] In some embodiments, the preset trajectories in the historical trajectory dataset are optionally constructed based on different seasons, different time periods (such as day / night), and different meteorological conditions.

[0148] By comparing the target vessel’s historical navigation trajectory with the most matching preset trajectory subset that conforms to the current time and weather scenario, the rationality assessment of behavior can dynamically adapt to periodic changes and the influence of the external environment, thereby improving the contextual relevance of the judgment.

[0149] In some embodiments, the calculation of regional attribute parameters is optionally based not only on static electronic charts and preset sensitive area information, but also on real-time dynamic information.

[0150] Dynamic information includes current traffic density in the target waters, information on temporarily designated navigation control zones, or real-time severe weather warnings.

[0151] By combining dynamic scene information, the location deviation feature value can reflect immediate risks, thereby enhancing the response sensitivity to sudden high-threat scenarios.

[0152] In some embodiments, optionally, after determining the position deviation feature value, the ship target association parameter generation method further includes: performing a trajectory extrapolation consistency check based on the vector direction of the original position deviation and historical AIS heading / speed information. If the position deviation direction is physically inconsistent with the heading and speed reported by the ship, an additional penalty weighting is applied to the position deviation feature value. By introducing kinematic-based consistency verification, the ability to identify malicious position spoofing (such as coordinate jumps) is further improved.

[0153] In some embodiments, optionally, corresponding confidence assessments are generated for the calculated identity consistency feature value, behavioral rationality feature value, and location deviation feature value, respectively.

[0154] Confidence assessment is determined based on the quality of the source data it relies on (such as image signal-to-noise ratio, AIS signal strength, and historical data completeness) and the stability of the computation process. By outputting feature values ​​with accompanying confidence, meta-information on the reliability of evidence in each dimension is provided for subsequent weighted fusion calculations, making the final association feature parameters more robust and interpretable.

[0155] In some embodiments, optionally, such as Figure 3 As shown, at least one multidimensional risk feature parameter is weighted and fused according to a preset weighting coefficient to determine the associated feature parameters of the target vessel corresponding to the associated target result, including:

[0156] Step S1120: Obtain the preset benchmark score;

[0157] Step S1122: Obtain the first weight coefficient corresponding to the identity consistency feature value;

[0158] Step S1124: Weight the identity consistency feature values ​​according to the first weight coefficient to determine the first weight value;

[0159] Step S1126: Obtain the second weight coefficient corresponding to the behavioral rationality feature value;

[0160] Step S1128: Weight the behavioral rationality feature value according to the second weighting coefficient to determine the second weighting value;

[0161] Step S1130: Obtain the third weight coefficient corresponding to the position deviation feature value;

[0162] Step S1132: Weight the position deviation feature value according to the third weighting coefficient to determine the third weighting value;

[0163] Step S1134: Determine the associated characteristic parameters of the target vessel based on the benchmark score, the first weighted value, the second weighted value, and the third weighted value.

[0164] In this embodiment, a benchmark score is introduced as the starting point for calculation, and preset weight coefficients are assigned to the risk characteristic values ​​of the three dimensions of identity consistency, behavior rationality and location deviation. By weighted summation (or weighted deduction), the risk quantification values ​​of different natures and dimensions are integrated into a comprehensive scalar score.

[0165] Understandably, this process not only aggregates multi-dimensional risk analysis results into intuitive decision indicators, but more importantly, it solidifies risk preferences in different monitoring scenarios (such as port security inspections focusing more on identity and open water inspections focusing more on behavior) into configurable business rules through a preset weighting coefficient mechanism. This solves the core defects of existing technologies, such as their single judgment logic and inability to output quantitative risk levels. It provides a direct mathematical basis for precise hierarchical and classified supervision, and improves the accuracy of data fusion and the reliability of credibility judgment.

[0166] Furthermore, by associating feature parameters, regulators can clearly distinguish between low-confidence targets that require close attention and high-risk targets that require immediate action, achieving a fundamental shift from fragmented alerts to data-driven, tiered intelligent assessment.

[0167] Among them, the specific implementation process of weighted fusion calculation of at least one multidimensional risk characteristic parameter according to the preset weight coefficient to determine the associated characteristic parameter of the target vessel corresponding to the associated target result is a key step in integrating scattered risk quantification evidence of different dimensions into a comprehensive evaluation result.

[0168] Through a structured weighted calculation framework, three independent risk indicators—identity consistency, behavioral rationality, and location deviation—are integrated according to their importance in the overall risk assessment, ultimately outputting an intuitive and single correlation feature parameter.

[0169] First, the system obtains a preset baseline score. The baseline score is an initial, theoretical maximum score representing a completely trustworthy or risk-free state, usually set to a fixed value (e.g., 100 points). All subsequent risk deduction calculations will be based on this score.

[0170] Next, the system obtains the first weight coefficient corresponding to the identity consistency feature value, the second weight coefficient corresponding to the behavior rationality feature value, and the third weight coefficient corresponding to the location deviation feature value.

[0171] These weighting coefficients (first weighting coefficient, second weighting coefficient, and third weighting coefficient) are parameters that are pre-determined through model training, expert experience, or business rule configuration.

[0172] The magnitude of each weight coefficient directly reflects the relative importance of its corresponding risk dimension in the entire credibility assessment model.

[0173] For example, in port security scenarios, identity consistency may be given a higher weight (i.e., a larger first weight coefficient) because impersonation is a serious violation; while in navigation monitoring in open sea areas, the weight of behavioral rationality (second weight coefficient) may be higher, used to detect abnormal behavior that deviates from the course.

[0174] Then, the system performs a weighted calculation.

[0175] The identity consistency feature values ​​are weighted according to the first weighting coefficient to determine the first weighting value.

[0176] Specifically, the calculated identity consistency feature value (a numerical value representing the degree of identity discrepancy, such as 0.3 representing a 30% size difference) is multiplied by the first weighting coefficient to obtain a weighted risk contribution value, namely the first weighted value.

[0177] Similarly, the system weights the behavior rationality feature value according to the second weight coefficient to determine the second weight value, that is, the behavior rationality feature value (such as a similarity value between 0 and 1, whose reciprocal or complement can represent the degree of irrationality) is converted into the weighted second weight value.

[0178] Similarly, the system weights the position deviation feature values ​​according to the third weight coefficient to determine the third weight value, and transforms the scened position deviation feature values ​​into the third weight value.

[0179] Finally, the system determines the associated characteristic parameters of the target vessel based on the baseline score, the first weighted value, the second weighted value, and the third weighted value.

[0180] The general calculation logic is to subtract the first weighted value, the second weighted value, and the third weighted value from the benchmark score in sequence.

[0181] For example, the associated feature parameter = benchmark score - (first weighted value + second weighted value + third weighted value).

[0182] Through calculation, the independent risk evidence from the three dimensions is integrated into a single, comprehensive correlation feature parameter according to its pre-defined importance.

[0183] The associated feature parameters intuitively reflect the overall credibility of the target vessel after multi-dimensional and weighted evaluation. The higher the score, the more credible the vessel, and the lower the score, the higher the risk. This provides law enforcement officers with a clear and quantitative basis for decision-making, completely changing the traditional binary alarm mode of black and white.

[0184] In some embodiments, the preset benchmark score and the first weighting coefficient, the second weighting coefficient, and the third weighting coefficient are not fixed values, but are dynamically adjusted according to the real-time monitoring scenario of the target vessel.

[0185] For example, when the system determines that a target vessel has entered sensitive waters or triggered a high-risk event, it automatically lowers the baseline score and increases the weighting coefficient of the position deviation feature value, making the overall score more sensitive to position anomalies. This allows the credibility assessment model to adapt to the differentiated risk tolerance of different monitoring areas.

[0186] In some embodiments, the weighted fusion calculation process may not be a simple linear weighted summation, but may also include a nonlinear transformation step. For example, for any weighted feature value exceeding a preset threshold, an exponential or piecewise function is used to amplify the value, thereby imposing a more significant risk penalty on extreme outliers and enhancing the model's ability to identify high-risk signals.

[0187] In some embodiments, optionally, after calculating the associated characteristic parameters, the scoring trend or volatility of the target vessel is also calculated based on its historical associated characteristic parameter sequence. Combining the associated characteristic parameters with the scoring trend or volatility generates a comprehensive risk posture index. This allows the assessment to not only focus on the risk at the present moment but also capture risk evolution patterns and identify progressively deteriorating behaviors.

[0188] In some embodiments, the determination of the first weighting coefficient, the second weighting coefficient, and the third weighting coefficient is optionally based on machine learning training on a large number of historical verification samples. The historical verification samples include various ship behavior cases verified by humans and their correct risk level labels. Optimizing the weights through a data-driven approach makes the model more objective and closer to actual regulatory experience.

[0189] In some embodiments, optionally, the method outputs a weighted contribution analysis of each dimension simultaneously when outputting the final associated feature parameters. The weighted contribution analysis clearly indicates the contribution ratio of the three dimensions—identity consistency, behavioral rationality, and location deviation—to the final score, providing users with an interpretable basis for judgment and explaining where the risk mainly originates.

[0190] In some embodiments, optionally, such as Figure 4As shown, after determining the association feature parameters of the target vessel corresponding to the associated target result, the method for generating vessel target association parameters further includes:

[0191] Step S1140: Obtain multiple preset risk level thresholds;

[0192] Step S1142: Compare the associated feature parameters with multiple risk level thresholds to determine the comparison results;

[0193] Step S1144: Determine the risk level corresponding to the target vessel based on the comparison results.

[0194] In this embodiment, by introducing a risk level threshold comparison and mapping mechanism, a key transformation from continuous associated feature parameters to discrete risk classification is achieved.

[0195] By acquiring multiple preset risk level thresholds, the quantitative score is compared with these thresholds, and the specific risk level corresponding to the target vessel is determined based on the comparison results.

[0196] Understandably, transforming abstract numerical scores into explicit instructions that can directly drive differentiated regulatory actions enables high-risk targets to be quickly identified and trigger emergency responses, while medium- and low-risk targets enter the corresponding monitoring levels, thereby achieving precision and maximization of regulatory effectiveness under the premise of limited resources.

[0197] Furthermore, by rapidly identifying high-risk targets and automatically associating them with emergency response procedures, while allocating monitoring resources of appropriate intensity to medium- and low-risk targets, the optimal allocation of regulatory resources and a significant improvement in response efficiency have been achieved.

[0198] After calculating the quantitative correlation characteristic parameters of the target vessel, the method for generating vessel target correlation parameters further includes a risk level classification step. This transforms continuous, numerical correlation characteristic parameters into discrete, business-oriented risk classifications, thereby achieving a crucial link from data analysis to regulatory action.

[0199] First, the system obtains multiple preset risk level thresholds.

[0200] These risk level thresholds are one or more numerical demarcation points pre-set by system administrators or domain experts based on actual regulatory strategies and risk tolerance.

[0201] For example, risk level thresholds include, but are not limited to, high-risk thresholds (e.g., 30 points) and medium-risk thresholds (e.g., 70 points).

[0202] The high-risk threshold and medium-risk threshold divide the entire scoring range (e.g., 0-100 points) into three consecutive intervals, each corresponding to a different risk level.

[0203] The setting of thresholds reflects business rules and determines the system's sensitivity to risks and the granularity of its response.

[0204] Next, the system compares the associated feature parameters with multiple risk level thresholds to determine the comparison results.

[0205] The system will compare the calculated associated characteristic parameters of the target vessel (e.g., 45 points) with multiple preset risk level thresholds (high risk threshold 30, medium risk threshold 70) in turn.

[0206] The comparison result is a logical judgment used to determine which threshold range the associated feature parameter falls into. For example, the comparison can determine that: 45 points > the high-risk threshold of 30, and 45 points < the medium-risk threshold of 70.

[0207] Finally, the system determines the risk level of the target vessel based on the comparison results.

[0208] This step, based on the comparison results from the previous step, executes a mapping rule.

[0209] For example, if the comparison result is "score < high risk threshold", then the risk level of the target vessel is determined to be high risk;

[0210] If the result is "high risk threshold ≤ score < medium risk threshold", then it is determined to be medium risk;

[0211] If the result is "score ≥ medium risk threshold", then it is determined to be low risk.

[0212] The determined risk level is a clear classification label that directly summarizes the overall risk status of the target vessel.

[0213] In some embodiments, the preset risk level thresholds are optionally not globally fixed values, but are dynamically adjusted based on at least one factor among the target vessel's type, the type of waterway currently in which it is located, or historical violation records. For example, a stricter high-risk threshold is used for dangerous goods vessels such as oil tankers near sensitive waters. This allows risk classification to adapt to the differentiated regulatory requirements of different vessels and different regions, improving the accuracy and applicability of the classification.

[0214] In some embodiments, the risk level classification may optionally include not only qualitative labels such as "high risk," "medium risk," and "low risk," but also a numerical risk index range. For example, the "medium risk" level can be further refined into a "60-70" range representing moderate to high risk and a "70-85" range representing moderate to low risk. By providing more granular numerical ranges, users are given a more detailed risk stratification basis than the traditional three-level classification, supporting more precise resource allocation.

[0215] In some embodiments, the risk level thresholds are optionally obtained by machine learning training on historical ship behavior data and final verification results. The system continuously collects historical correlation feature parameters and their corresponding real ship behavior characteristics (normal / non-compliant), and dynamically adjusts the thresholds through optimization algorithms to ensure that the risk level classification matches the actual risk distribution to the greatest extent possible, thereby improving the model's adaptability to new risks.

[0216] In some embodiments, optionally, while determining the risk level corresponding to the target vessel, a confidence level for that level classification is also output. The confidence level is calculated based on the proximity of the associated feature parameters to the threshold boundary and the quality of the feature data of each dimension used to calculate the score. By providing the confidence level, the user is informed of the reliability of the current risk level judgment, assisting them in making more prudent decisions when dealing with boundary situations.

[0217] In some embodiments, optionally, such as Figure 5 As shown, after determining the risk level corresponding to the target vessel, the method for generating vessel target association parameters also includes:

[0218] Step S1160: Obtain preset response strategy data;

[0219] Step S1162: Determine the target response strategy from the preset response strategy data based on the risk level;

[0220] Step S1164: Execute the risk management operation corresponding to the target response strategy.

[0221] In this embodiment, the defined automated response steps achieve a closed loop from intelligent risk assessment to precise business execution. By matching predefined response strategies (pre-defined response strategy data) strictly bound to different risk levels with the real-time determined ship risk level, the corresponding risk handling operations are automatically triggered and executed. The quantified risk level generated by the preceding steps is directly converted into specific regulatory instructions, completing the entire process of automating from risk identification to risk handling, significantly improving the response speed, execution standardization, and overall efficiency of maritime supervision.

[0222] After determining the risk level corresponding to the target vessel, the vessel target association parameter generation method further includes automatic response and handling steps. This aims to automatically transform the risk assessment conclusions generated by the aforementioned steps into actionable regulatory or monitoring actions, thereby significantly improving response efficiency and reducing human decision-making delays.

[0223] First, the system obtains the preset response strategy data.

[0224] The preset response strategy data is a structured, configurable strategy rule base or database, which predefines standardized operating procedures, instruction templates, task types, and execution objectives corresponding to different risk levels.

[0225] For example, risk registration includes, but is not limited to, high risk, medium risk, and low risk.

[0226] For example, the preset response strategy data may explicitly stipulate: a strategy that immediately issues an alarm and generates an investigation report when a high-risk level is associated with a risk level;

[0227] The strategy of generating verification task orders based on medium-risk levels and pushing them to the regional monitoring center;

[0228] The strategy of linking low-risk levels to daily monitoring logs.

[0229] Next, the system determines the target response strategy from the preset response strategy data based on the risk level.

[0230] The risk level of the identified target vessel (e.g., "high risk") is used as the query key to match and search within the preset response strategy data, thereby determining the corresponding specific and unique target response strategy. For example, the matched target response strategy might be strategy number "RESP-HIGH-01," which fully defines a series of operational instructions, information filling templates, and output channels that need to be executed for high-risk vessels.

[0231] Finally, the system executes the risk management operations corresponding to the target response strategy.

[0232] The system automatically parses and executes all operational instructions defined in the target response strategy. These risk management operations are specific tasks that can be performed by the system or personnel, and typically include, but are not limited to, the following types:

[0233] Alarm and notification operation: Trigger the audible and visual alarm device, or send real-time alarm information containing vessel details and risk level to designated law enforcement personnel's mobile terminals or the command center's large screen.

[0234] Task generation and dispatch: Automatically create a structured on-site inspection task sheet or key monitoring task, which automatically fills in the name, location, associated characteristic parameters, and main risk characteristics of the target vessel (such as "size difference 45%)", and dispatches the task to the nearest patrol unit or monitoring post in charge of the sea area.

[0235] Automated report generation: By calling a report template, the system automatically integrates the current vessel's AIS information, SAR evidence images, associated feature parameter curves, and a multi-dimensional risk feature analysis list to generate a standardized report with a complete chain of evidence for law enforcement personnel to archive or use as a basis for law enforcement.

[0236] Data recording and status update operations: Update the risk level, assessment time, and handling status of the vessel to the central database and include it in the corresponding level of continuous monitoring list, such as marking it as a high-risk vessel dynamic monitoring list for key tracking.

[0237] Through an automated response chain, it is ensured that every risk assessment conclusion receives timely, standardized, and traceable business feedback, directly transforming advanced data analysis capabilities into tangible regulatory effectiveness.

[0238] In some embodiments, the strategy defined in the preset response strategy data may include logic for dynamic adjustment based on the confidence level of the target risk level. For example, for a high-risk determination with low confidence, the strategy may add an additional step requiring manual verification before proceeding with subsequent operations. This makes the automatic response mechanism both efficient and prudent, reducing the risk of triggering a major response directly due to model misjudgment.

[0239] In some embodiments, optionally, after the risk management operation is performed, the status changes of the target vessel are continuously monitored, and the effectiveness of the management is automatically evaluated and feedback is provided based on subsequent monitoring data. For example, after an inspection task is assigned, the system tracks whether the vessel accepts the inspection as instructed or whether its behavior returns to normal, and feeds back this result to evaluate the accuracy of the initial assessment and the effectiveness of the response strategy.

[0240] In some embodiments, optionally, performing risk management operations may also include automatically generating call instructions for external interactive data interfaces to push key assessment conclusions (such as vessel ID, risk type, and location) to other related external business platforms, such as port scheduling systems or marine insurance risk assessment platforms, in a standard format. Through cross-platform data linkage, the risk perception capabilities of this invention are integrated into a broader shipping ecosystem management system.

[0241] In one specific embodiment, the present invention optionally provides an intelligent system and method for determining the credibility of ship monitoring information, aiming to solve the core problems of crude judgment, inability to quantify risks, and disconnect from reality in related technologies.

[0242] Unlike related technologies that only perform simple data comparisons, this invention constructs a complete closed-loop system from comprehensive data processing to intelligent judgment and then to business triggering.

[0243] Its core lies in the fact that the system does not directly assert whether a ship is legal or illegal. Instead, it uses a quantitative model to calculate the degree of consistency between the ship's current reported information and multi-source observation information, i.e., the reliability score, thereby providing law enforcement officers with a graded and interpretable basis for decision-making.

[0244] The following is combined Figure 8 The system architecture shown illustrates how this solution achieves this goal through four core components working in tandem:

[0245] Step 1: Multi-source data processing and adaptive alignment center

[0246] Technical means: This is the system data input layer, i.e. the data preparation workshop.

[0247] It simultaneously accesses three types of data: satellite imagery (SAR); Automatic Ship Information System (AIS); and auxiliary information (such as electronic charts, port lists, and meteorological information).

[0248] The biggest difference from traditional methods is that this section has a built-in adaptive alignment mechanism.

[0249] The adaptive alignment mechanism first automatically analyzes potential systematic positional deviations (distortions) in satellite imagery due to factors such as shooting angle, and then estimates different positional correction parameters for different regions of the image. The geometric distortion correction formula is as follows:

[0250] For each detected target in the SAR image, its actual geographical location It can be estimated using the following model:

[0251] ;

[0252] ;

[0253] in, This is a satellite view, where h represents altitude. and The correction amount is calculated based on the imaging model. The x-coordinate of the ship's position in the SAR image report. The vertical coordinates of the ship's position are used to report the SAR image.

[0254] Then, the positions of ships detected on SAR images and those reported by AIS are dynamically calibrated and paired, rather than performing simple, fixed range matching on the raw data. This is an adaptive spatiotemporal matching logic, where the matching threshold is not fixed but dynamically adjusted according to the error model of the target's region.

[0255] If the following conditions are met:

[0256] ;

[0257] If the match is successful, then the match is successful; otherwise, the match is unsuccessful.

[0258] in, This is the estimated value of the overall positioning error for the region. The x-coordinate of the ship's position in the SAR image report. The ordinate of the ship's position is used to report the SAR image. The x-coordinate of the ship's position reported by AIS. Provide the longitudinal coordinates of the vessel's position for AIS reporting.

[0259] This ensures that subsequent analysis is based on accurate correlations, solving the root cause of misjudgments caused by inaccurate matching in related technologies.

[0260] Step 2: Multi-dimensional Feature Fusion and Profile Construction Module

[0261] Technical means: This is the core processing layer of the system, namely the intelligence processing unit.

[0262] It extracts features from multiple dimensions from precisely matched data to build a comprehensive behavioral profile for each ship.

[0263] Specific tasks and functions: The multi-dimensional feature fusion and profile building module not only checks whether the location matches, but also comprehensively checks: identity consistency: whether the estimated size and shape of the ship on the satellite image match the type and size of the AIS report.

[0264] This invention proposes to define dimensional difference degree :

[0265] ;

[0266] like A value greater than 0.3 indicates a questionable identity.

[0267] in, This refers to the observation size data corresponding to the satellite image. This refers to the report size data corresponding to the Automatic Identification System (AIS) data for ships.

[0268] Behavioral rationality: Whether the ship's movement trajectory and speed conform to common behavioral patterns in the area (such as waterways and anchorages).

[0269] This invention proposes a method to calculate the similarity between the current trajectory and typical patterns based on a historical trajectory pattern database. (0~1), values ​​below 0.6 are considered abnormal.

[0270] Contextual information: current time, weather, whether in sensitive waters or near other suspicious vessels.

[0271] These multi-dimensional features are integrated into a structured feature profile, thus upgrading the single location matching problem into a comprehensive scan of all credible signs of a ship, enabling the discovery of more concealed risks such as falsified identities and abnormal behavior.

[0272] Step 3: Credibility Intelligent Assessment Engine

[0273] Technical means: This is the brain of the system. It receives the above feature profiles and outputs a quantified credibility score (i.e., associated feature parameters) (e.g., 0-100 points) and key reasons.

[0274] Specific functions and roles: The credibility intelligent assessment engine has a built-in learnable assessment model.

[0275] This assessment model has learned how to weigh the impact of various contradictory characteristics on "credibility" by analyzing a large number of historical cases (including normal ships and known cases of fraud and violations).

[0276] For example, it might determine that slight AIS position drift on the high seas might result in fewer points deductions; however, near restricted areas, the same drift combined with an inappropriate vessel type would incur significant point deductions. Ultimately, it calculates a composite score for each vessel. This fundamentally changes the binary "yes or no" alert model of related technologies, providing a continuous risk spectrum that allows law enforcement to distinguish between "low-confidence targets requiring attention" and "high-risk targets requiring immediate action."

[0277] The specific comprehensive credibility scoring model is as follows:

[0278] ;

[0279] in, These are the weighting coefficients. This is the normalized value of the positional deviation. Indicate whether it is in a sensitive area (yes=1, no=0). To score credibility, For size variation, This represents the similarity between the current trajectory and the typical pattern.

[0280] in This involves converting the original, dimensional physical position deviation into a dimensionless value between 0 and 1 (or 0% to 100%), which is used to characterize the severity level of the physical position deviation under the current specific environment.

[0281] There are two possible scenarios: Scenario A (high seas): The SAR detection location of a ship differs from its AIS reported location by 500 meters. Scenario B (busy port): The location of another ship also differs by 500 meters. In the relevant technology (fixed threshold matching), these two "500-meter" discrepancies are treated equally.

[0282] However, in actual supervision, a 500-meter deviation within a port is far more suspicious and dangerous than a 500-meter deviation on the high seas. This is because the positioning accuracy requirements are higher within ports, and the vessels are densely packed; a small deviation could indicate a collision risk or deliberate violation of berthing regulations.

[0283] Therefore, we cannot directly deduct points based on the original distance difference (e.g., 500 meters). We need a method that can determine the severity of this deviation based on the current scenario.

[0284] Risk levels are classified as follows:

[0285] High risk: Score < 30;

[0286] Medium risk: 30 ≤ Score < 70;

[0287] Low risk: Score ≥ 70.

[0288] Step 4: Business Decision Support and Automated Response Module

[0289] Technical means: These are the system's hands and feet, responsible for transforming intelligent analysis results into actual action instructions.

[0290] Specific functions and roles: The business decision support and automated response module includes a configurable response rule base. Based on credibility scores and identified risk types (such as "suspicious identity" or "abnormal behavior"), the system automatically triggers different levels of response processes.

[0291] Low risk (high score): Automatically recorded and included in routine monitoring.

[0292] Medium risk (medium score): Automatically generate a verification task order and push it to nearby patrol boats or shore-based command centers, prompting them to pay close attention.

[0293] High risk (low score): Immediately triggers an audible and visual alarm and automatically generates a structured assessment report.

[0294] The structured analysis report automatically integrates vessel information, evidence screenshots (SAR images, AIS tracks), credibility scores, analysis of key risk points, and preliminary handling suggestions, enabling law enforcement personnel to make rapid decisions.

[0295] The Business Decision Support and Automated Response module achieves a seamless transition from risk identification to risk management. It directly embeds technological capabilities into law enforcement workflows, significantly improving efficiency from anomaly detection to response initiation, and providing data-rich and well-supported reporting for action.

[0296] In summary, this invention constructs a complete technology chain through the cascading and collaboration of the four components described above, encompassing "precise data alignment, comprehensive feature profiling, quantitative risk scoring, and automated business response." It is no longer merely a tool providing fragmented alerts, but an intelligent auxiliary system capable of comprehensive analysis and outputting clear decision support information, fundamentally improving the precision and intelligence of maritime supervision.

[0297] In one specific embodiment, optionally, the data preprocessing and adaptive matching flowchart provided by the present invention is as follows: Figure 9 As shown, it includes:

[0298] Step S200: Input SAR imagery, AIS data, and auxiliary information;

[0299] Step S202: Data cleaning and standardization;

[0300] Step S204: Calculate the geometric distortion parameters of the SAR image region;

[0301] Step S206: Evaluate AIS signal quality and error;

[0302] Step S208: Does a sensitive area exist?

[0303] If the judgment result of step S208 is yes, then execute step S210: adopt a stricter matching threshold;

[0304] If the judgment result of step S208 is negative, proceed to step S212: adopt an adaptive matching threshold;

[0305] Step S214: Perform dynamic spatiotemporal matching;

[0306] Step S216: Output the pairing results and confidence levels.

[0307] After the process begins, three types of raw data are input. The first type is satellite synthetic aperture radar imagery, which is image data containing echo signals of ships on the sea surface obtained through radar sensor observation of the Earth. The second type is Automatic Identification System (AIS) data, which is message information automatically broadcast by ships, containing dynamic and static attributes such as the ship's identity, position, course, and speed. The third type is auxiliary information, mainly including electronic nautical charts, pre-set sensitive area boundary information, port directories, and real-time meteorological and hydrological data. This information provides geographical and contextual references for subsequent processing.

[0308] The three types of raw input data are preprocessed to eliminate noise and inconsistencies, providing a high-quality and standardized data foundation for subsequent accurate analysis.

[0309] Due to the side-looking characteristics of satellite imaging and factors such as the Earth's curvature and terrain undulations, the geographical location of targets in satellite synthetic aperture radar (SAR) images exhibits a systematic deviation, a phenomenon known as geometric distortion. Based on the satellite's orbital parameters, imaging mode, and digital elevation model, the geographical location correction amount corresponding to each pixel in the image is calculated, thereby eliminating the impact of distortion on subsequent positioning and matching accuracy.

[0310] The positioning accuracy of Automatic Identification System (AIS) signals is not constant; its error is affected by various factors such as the coverage area of ​​the receiving station, the signal propagation environment, and the performance of the ship's reporting equipment. By analyzing the temporal continuity, the rationality of position jumps, and the deviation from a known fixed reference point of the AIS signals, the overall reliability level and typical error range of the AIS signals in the current sea area are dynamically evaluated, providing a basis for subsequent adaptive matching.

[0311] The system retrieves electronic nautical charts and preset area lists from auxiliary information to determine whether the sea area covered by the currently processed satellite synthetic aperture radar image contains areas defined as sensitive areas, such as military restricted zones, offshore oil and gas platforms, ecological protection zones, and busy shipping lane intersections.

[0312] If the assessment indicates the existence of a sensitive area, a stricter matching threshold is applied. Within sensitive areas, the tolerance for errors in monitoring ship behavior is extremely low. Therefore, the system will automatically reduce the spatiotemporal matching threshold used to correlate targets detected by satellite synthetic aperture radar with targets reported by the Automatic Identification System (AIS). This means that only when two targets are closer in both time and space will they be identified as the same ship, thereby minimizing the risk of incorrectly associating different ships in critical areas.

[0313] If the determination result indicates that no sensitive area exists, an adaptive matching threshold is used. In general, non-sensitive sea areas, the system employs a dynamic, non-fixed matching threshold.

[0314] The geometrically corrected satellite synthetic aperture radar (SAR) ship target positions are correlated with the quality-screened AIS (Automatic Identification System) reported target positions. Under the matching threshold conditions determined in steps four and five, the correlation calculation is performed, and the output is multiple correlated target pairs (i.e., correlated target results). Each target pair links a SAR detection point with an AIS reported point, and is regarded as evidence of different sources of the same ship.

[0315] Finally, the process outputs the pairing results and confidence levels.

[0316] In one specific embodiment, optionally, the system interface diagram is as follows: Figure 10As shown, the system interface is a visual operation platform that transforms the quantitative results and structured evidence output by the backend intelligent analysis engine into a form that law enforcement officers can intuitively understand and make quick decisions.

[0317] The Vessel Monitoring System - Law Enforcement Command Interface (i.e., Vessel Behavior Credibility Judgment System - Law Enforcement Command Interface) clarifies the functional positioning of the current view, namely, a dedicated interactive interface serving maritime law enforcement command.

[0318] The target list is a tabular view window that centrally displays all vessels processed by the system within the currently monitored sea area and their core risk assessment results. The target list contains four columns of data: vessel name, credibility, status, and operations.

[0319] The vessel name column displays the vessel's identifier, such as "Vessel 123" or "Vessel 888". These names are usually derived from the vessel's name or maritime mobile communication service identifier in the Automatic Identification System (AIS) message.

[0320] The credibility column displays the quantitative credibility score calculated for each ship by the core credibility intelligent assessment engine of this invention. The credibility score is a value from 0 to 100, such as 85 or 18. The higher the score, the higher the credibility of the ship's current behavior and reported information.

[0321] The status column is a visual label automatically determined based on a credibility score, displaying risk levels such as "Low Risk," "Medium Risk (Pending Verification)," and "High Risk (Alert)." This determination is directly linked to the preset risk level thresholds and response rules in the business decision support and automated response modules. Different status labels may trigger different backend monitoring strategies.

[0322] The operation bar provides an interactive button to view details, allowing law enforcement officers to click and obtain detailed assessment information about a particular vessel.

[0323] When law enforcement officers click the "View Details" button for a vessel (e.g., "Vessel 888") in the target list, the system will dynamically load and display an in-depth analysis report of the selected target vessel.

[0324] The report begins by prominently highlighting a summary of the core risk issues posed by the vessel, such as "suspicious identity," "abnormal behavior," and "AIS signal fraud." These summary items are distillations of key risk conclusions from the structured feature profile output by the multi-dimensional feature fusion and profiling module.

[0325] The list of key evidence presents the multi-dimensional and quantifiable evidence data supporting the above risk conclusions in an itemized format. Each piece of evidence corresponds to one or more risk feature values ​​extracted from the technical solution and their calculation results.

[0326] SAR Size Difference from AIS Report: 45%: This evidence comes directly from the calculation of the "Identity Consistency Feature Value". The figure of 45% is the specific numerical value of the size difference, indicating that there is a significant difference of 45% between the ship size estimated from synthetic aperture radar imagery and the size reported by the Automatic Identification System (AIS).

[0327] Position offset exceeds threshold: 2.3 nautical miles: This is the specific value of the original position deviation, that is, the actual physical distance between the synthetic aperture radar detected position and the position reported by the automatic identification system after adaptive alignment correction.

[0328] The system determined that this distance exceeded the matching threshold dynamically calculated based on the error model of the area, constituting direct evidence of location fraud or anomaly.

[0329] Number of historical behavioral anomalies: 3 times: This evidence is closely related to the calculation of the behavioral rationality characteristic value. 3 times indicates the number of times the similarity between the selected target's vessel's recent historical trajectory and typical patterns in the historical trajectory pattern library is lower than a preset threshold, reflecting the continued anomalousness of its behavior.

[0330] The current trajectory does not match the declared purpose: This is a logical analysis of the current navigation trajectory by combining destination information in the Automatic Identification System (AIS) message with multi-dimensional feature fusion, and is another concrete manifestation of the judgment of the rationality of the behavior.

[0331] By integrating information such as electronic nautical charts and real-time ship positions, the system determined that the traffic density in the target's current area was significantly higher than the historical normal or the level of similar areas, which was an anomaly in the context and may indicate sensitive behaviors such as assembly or transfer.

[0332] The three function buttons located at the bottom of the interface—Generate Enforcement Report, Assign Verification Task, and Initiate Real-time Alert—serve as the direct entry point to the Business Decision Support and Automated Response module in the user interface.

[0333] Law enforcement officers can manually trigger or confirm the system-suggested response process based on the status in the target list and the evidence in the details of the selected target.

[0334] For example, for high-risk (alert) vessel 888, clicking the "Generate Enforcement Report" button will automatically call up the report template, fill in all the evidence and vessel information in the details area, and generate a structured draft enforcement document.

[0335] Clicking "Dispatch Verification Task" will create a verification instruction containing the vessel's location and risk characteristics, and send it to the designated patrol unit. This achieves a seamless connection from risk assessment to response action.

[0336] like Figure 6As shown in the figure, this embodiment of the invention also provides a ship target association parameter generation device 900, which includes: a data acquisition module 902, used to acquire multi-source monitoring data of the target water area, the multi-source monitoring data including at least satellite image data and automatic identification system (AIS) data; an image detection module 904, used to determine at least one first detection target and first location data corresponding to the first detection target based on the satellite image data; a data identification module 906, used to determine at least one second detection target, second location data corresponding to the second detection target, and ship attribute data based on the AIS data; an association matching module 908, used to perform spatiotemporal association matching of the first detection target and the second detection target based on the first location data and the second location data to determine at least one associated target result; a risk feature module 910, used to determine at least one multi-dimensional risk feature parameter for each associated target result based on the satellite image data and AIS data corresponding to the associated target result; a weight acquisition module 912, used to acquire a preset weight coefficient; and a scoring and judgment module 914, used to perform weighted fusion calculation on at least one multi-dimensional risk feature parameter based on the preset weight coefficient to determine the association feature parameter of the target ship corresponding to the associated target result.

[0337] like Figure 7 As shown, this embodiment of the invention also provides an electronic device 1000, including a processor 1110, a memory 1109, and a program or instructions stored in the memory 1109 and executable on the processor 1110. When the program or instructions are executed by the processor 1110, they implement the various processes of the above-described embodiment of the ship target association parameter generation method and achieve the same technical effect. To avoid repetition, they will not be described again here.

[0338] Optionally, the processor 1110 is used to acquire multi-source monitoring data of the target water area, the multi-source monitoring data including at least satellite imagery data and automatic identification system data of ships;

[0339] Optionally, the processor 1110 is configured to determine at least one first detection target and first location data corresponding to the first detection target based on satellite imagery data;

[0340] Optionally, the processor 1110 is configured to determine at least one second detection target, second location data corresponding to the second detection target, and ship attribute data based on the ship automatic identification system data;

[0341] Optionally, the processor 1110 is configured to perform spatiotemporal correlation matching on the first detection target and the second detection target based on the first location data and the second location data, and determine at least one associated target result;

[0342] Optionally, the processor 1110 is configured to determine at least one multidimensional risk feature parameter for each associated target result, based on satellite imagery data and Automatic Identification System (AIS) data corresponding to the associated target result;

[0343] Optionally, the processor 1110 is used to obtain preset weight coefficients;

[0344] Optionally, the processor 1110 is used to perform weighted fusion calculation on at least one multidimensional risk feature parameter according to a preset weight coefficient, and determine the associated feature parameter of the target vessel corresponding to the associated target result.

[0345] The memory 1109 can be used to store software programs and various data. The memory 1109 may primarily include a first storage area for storing programs or instructions and a second storage area for storing data. The first storage area may store the operating system, application programs or instructions required for at least one function (such as sound playback, image playback, etc.). Furthermore, the memory 1109 may include volatile memory or non-volatile memory, or both. The non-volatile memory may be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. Volatile memory can be random access memory (RAM), static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDRSDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM), and direct memory bus RAM (DRRAM). The memory 1109 in this embodiment includes, but is not limited to, these and any other suitable types of memory.

[0346] This invention also provides a readable storage medium storing a program or instructions. When executed by a processor, the program or instructions implement the various processes of the above-described ship target association parameter generation method embodiments and achieve the same technical effects. To avoid repetition, these will not be described again here. Furthermore, the readable storage medium improves the data storage capacity and data processing speed of the ship target association parameter generation method in this invention.

[0347] A computer-readable storage medium can be a tangible device that holds and stores instructions for use by an instruction execution device. A computer-readable storage medium can be an electronic storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing, but is not limited thereto. A non-exhaustive list of more specific examples of computer-readable storage media includes: portable computer floppy disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable and programmable read-only memory (EPROM or flash memory), static random access memory (SRAM), portable optical disc read-only memory (CD-ROM), digital universal disk (DVD), memory cards, floppy disks, encoding mechanical devices (e.g., punched cards or grooves with raised structures for recording instructions), and any suitable combination of the foregoing. The computer-readable storage medium used herein should not be construed as the transmission of signals themselves, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through waveguides or other transmission media, or electrical signals transmitted through wires.

[0348] The processor is the processor in the electronic device described in the above embodiments. The readable storage medium includes computer-readable storage media, such as computer read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk.

[0349] This invention also provides a chip, which includes a processor and a communication interface. The communication interface and the processor are coupled. The processor is used to run programs or instructions to implement the various processes of the above-described ship target association parameter generation method embodiments, and can achieve the same technical effect. To avoid repetition, it will not be described again here. Furthermore, the chip improves the data processing speed of the ship target association parameter generation method in this invention.

[0350] It should be understood that the chip mentioned in the embodiments of the present invention may also be referred to as a system-on-a-chip, system chip, chip system, or system-on-a-chip, etc.

[0351] In this invention, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance; the term "multiple" refers to two or more unless otherwise explicitly defined. The terms "install," "connect," "link," and "fix" should be interpreted broadly. For example, "connect" can be a fixed connection, a detachable connection, or an integral connection; "link" can be a direct connection or an indirect connection through an intermediate medium. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0352] In the description of this invention, it should be understood that the terms "upper," "lower," "left," "right," "front," "rear," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or unit referred to must have a specific orientation or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0353] In the description of this specification, the terms "one embodiment," "some embodiments," "specific embodiment," etc., refer to specific features, structures, materials, or characteristics described in connection with an embodiment or example that are included in at least one embodiment or example of the present invention. 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.

[0354] The above are merely preferred embodiments of the present invention and are not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for generating ship target association parameters, characterized in that, The ship target association parameter generation method is used in a ship monitoring system to enable the ship monitoring system to screen, sort, or perform subsequent processing on at least one target ship. The ship target association parameter generation method includes: Acquire multi-source monitoring data of the target water area, wherein the multi-source monitoring data includes at least satellite imagery data and automatic identification system (AIS) data. Based on the satellite imagery data, at least one first detection target and the first location data corresponding to the first detection target are determined; Based on the data from the Automatic Identification System (AIS), at least one second detection target, second location data corresponding to the second detection target, and ship attribute data are determined. Based on the first location data and the second location data, spatiotemporal correlation matching is performed on the first detection target and the second detection target to determine at least one associated target result; For each of the associated target results, at least one multidimensional risk feature parameter is determined based on the satellite imagery data and the Automatic Identification System (AIS) data corresponding to the associated target result; Obtain the preset weight coefficients; The at least one multidimensional risk feature parameter is weighted and fused according to the preset weight coefficient to determine the associated feature parameter of the target vessel corresponding to the associated target result.

2. The method for generating ship target association parameters according to claim 1, characterized in that, The at least one multidimensional risk characteristic parameter includes identity consistency characteristic value, behavioral rationality characteristic value, and location deviation characteristic value.

3. The method for generating ship target association parameters according to claim 2, characterized in that, The step of determining at least one multidimensional risk characteristic parameter based on the satellite imagery data and the Automatic Identification System (AIS) data corresponding to the associated target results includes: The observation dimensions of the target vessel are determined based on the satellite imagery data. Determine the report size data of the target vessel based on the vessel attribute data; Calculate the identity consistency feature value based on the observed size data and the reported size data; Obtain the historical navigation trajectory of the target vessel from the Automatic Identification System (AIS) data corresponding to the associated target results; Obtain historical trajectory dataset; The historical navigation trajectory is compared with a preset trajectory in the historical trajectory dataset to determine the rationality feature value of the behavior; Calculate the original position deviation based on the first position data and the second position data corresponding to the associated target results; Obtain the regional attribute parameters corresponding to the current position of the target vessel. The regional attribute parameters are determined based on auxiliary information data, which includes electronic nautical charts and preset sensitive area information. The position deviation feature value is determined based on the original position deviation and the regional attribute parameters.

4. The method for generating ship target association parameters according to claim 3, characterized in that, The step of performing a weighted fusion calculation on the at least one multidimensional risk feature parameter according to the preset weight coefficient to determine the associated feature parameter of the target vessel corresponding to the associated target result includes: Obtain the preset benchmark score; Obtain the first weight coefficient corresponding to the identity consistency feature value; The identity consistency feature value is weighted according to the first weighting coefficient to determine the first weighting value; Obtain the second weighting coefficient corresponding to the rationality feature value of the behavior; The rationality feature value of the behavior is weighted according to the second weighting coefficient to determine the second weighting value; Obtain the third weighting coefficient corresponding to the position deviation feature value; The position deviation feature value is determined by weighting the value according to the third weighting coefficient. The associated characteristic parameters of the target vessel are determined based on the benchmark score, the first weighted value, the second weighted value, and the third weighted value.

5. The method for generating ship target association parameters according to claim 1, characterized in that, After determining the association feature parameters of the target vessel corresponding to the associated target result, the method for generating vessel target association parameters further includes: Obtain multiple preset risk level thresholds; The associated feature parameters are compared with the multiple risk level thresholds to determine the comparison result; The risk level of the target vessel is determined based on the comparison results.

6. The method for generating ship target association parameters according to claim 5, characterized in that, After determining the risk level corresponding to the target vessel, the method for generating vessel target association parameters further includes: Obtain preset response strategy data; Based on the risk level, a target response strategy is determined from the preset response strategy data; Perform risk management operations corresponding to the target response strategy.

7. A device for generating ship target association parameters, characterized in that, include: The data acquisition module is used to acquire multi-source monitoring data of the target water area, wherein the multi-source monitoring data includes at least satellite image data and automatic identification system data of ships; The image detection module is used to determine at least one first detection target and the first location data corresponding to the first detection target based on the satellite image data; The data identification module is used to determine at least one second detection target, second location data corresponding to the second detection target, and ship attribute data based on the data from the Automatic Identification System (AIS). The association matching module is used to perform spatiotemporal association matching on the first detection target and the second detection target based on the first location data and the second location data, and determine at least one associated target result; The risk feature module is used to determine at least one multidimensional risk feature parameter for each associated target result, based on the satellite imagery data and the Automatic Identification System (AIS) data corresponding to the associated target result. The weight acquisition module is used to obtain preset weight coefficients; The scoring and determination module is used to perform weighted fusion calculation on the at least one multidimensional risk feature parameter according to the preset weight coefficient, and determine the associated feature parameter of the target vessel corresponding to the associated target result.

8. An electronic device, characterized in that, It includes a processor, a memory, and a program or instructions stored in the memory and executable on the processor, wherein the program or instructions, when executed by the processor, implement the steps of the ship target association parameter generation method as described in any one of claims 1 to 6.

9. A readable storage medium, characterized in that, The readable storage medium stores a program or instructions that, when executed by a processor, implement the steps of the ship target association parameter generation method as described in any one of claims 1 to 6.

10. A chip, characterized in that, The chip includes a processor and a communication interface, the communication interface being coupled to the processor, the processor being used to run programs or instructions to implement the steps of the ship target association parameter generation method as described in any one of claims 1 to 6.

Images