A method and system for detecting and identifying a ship based on a large model

Abstract

This invention discloses a method and system for image detection and identification of ships based on a large model. The method includes: integrating multimodal information from visible light images and textual intelligence information to generate multi-source heterogeneous information; generating specific fleet group information based on the multi-source heterogeneous information; analyzing and judging the current environmental situation in conjunction with the multi-source heterogeneous information, wherein the judgment results include ship attributes, enemy combat intentions, and the range of sea areas that may be reached in the future; and generating an intelligence report based on the specific fleet group information and the judgment results.

Description

Technical Field

[0001] This invention relates to the field of image recognition technology, and more specifically, to a method and system for detecting and recognizing ships based on large-scale images. Background Technology

[0002] With the rapid development of technologies such as artificial intelligence (AI), machine learning (ML), and big data analytics, intelligent building design is gradually shifting from traditional manual modeling and experience-based judgment to data-driven and automated optimization. Against this backdrop, technologies such as Building Information Modeling (BIM) and Finite Element Analysis (FEA) are widely used in the design, evaluation, and optimization of modern buildings. Specifically, in the field of historical building restoration, due to the unique nature of these buildings, their design and construction need to optimize restoration plans without damaging the original cultural value, ensuring structural safety and long-term stability. Among these, the selection of restoration materials and the prediction of structural stress and deformation have become crucial issues in building restoration. To address this need, AI-powered intelligent analysis methods are used to comprehensively evaluate the compatibility of restoration materials with the original materials of the historical building, and to predict the potential stress and deformation effects before construction, ensuring a scientific and reasonable restoration process and avoiding unnecessary damage.

[0003] However, this method has certain limitations. In the process of historical building restoration, the compatibility analysis of restoration materials and the prediction of construction stress and deformation have long relied on manual experience and traditional engineering calculations. First, in terms of material compatibility, modern restoration materials and historical building materials have significant differences in physical and chemical properties. Existing technologies cannot accurately assess the compatibility of different materials in long-term use, which may lead to secondary damage to the restored building due to material incompatibility. In addition, during construction, due to the lack of accurate prediction of stress distribution and deformation trends in the area to be restored, new cracks and structural deformations often appear after restoration, which may even lead to a decrease in the overall stability of the building. It is difficult to fully predict the possible material aging and stress changes in the future by relying solely on manual experience and static calculations. Therefore, existing methods can often show restoration effects in the short term, but they are prone to problems in long-term use, leading to accelerated aging of the area to be restored and even damage to the original historical appearance of the building. Summary of the Invention

[0004] According to the present invention, a method and system for detecting and recognizing ships based on large-scale models are provided to solve the technical problem that existing methods often show repair effects in the short term, but are prone to problems in long-term use, leading to accelerated aging of the area to be repaired and even damage to the historical appearance of the original buildings.

[0005] According to a first aspect of the present invention, a method for detecting and recognizing ships based on large-scale models is provided, comprising:

[0006] Multimodal information from visible light images and textual intelligence information are integrated to generate multi-source heterogeneous information;

[0007] Based on the aforementioned multi-source heterogeneous information, specific fleet group information is generated;

[0008] The current environmental situation is combined with the aforementioned multi-source heterogeneous information to analyze and judge the results, which include ship attributes, enemy combat intentions, and the range of sea areas that may be reached in the future.

[0009] An intelligence report is generated based on the information and analysis results of the specific fleet group.

[0010] Optionally, multimodal information from visible light images and textual intelligence information are integrated to generate multi-source heterogeneous information, including:

[0011] Based on multi-source heterogeneous input data, multi-source heterogeneous information is generated. The multi-source heterogeneous input data includes at least: Automatic Identification System (AIS) data files, visible light satellite image data, and intelligence text information input by the user.

[0012] The first intelligent agent parses the Automatic Identification System (AIS) data file of the ship and outputs the ship's number, category, time, latitude and longitude, speed, and direction information;

[0013] The second intelligent agent performs target detection, image classification, and semantic segmentation on visible light satellite images, and outputs ship feature description information.

[0014] Visual language analysis of visible light satellite images is performed by a third-party intelligent agent to generate a comprehensive scene description that includes entity information and environmental status.

[0015] The fourth intelligent agent retrieves matching basic information from the knowledge base based on intelligence information input by the user.

[0016] Optionally, based on the multi-source heterogeneous information, specific fleet group information is generated, including:

[0017] Based on the multi-source heterogeneous information, ship technical parameters and ship captain information are generated. The ship technical parameters include: displacement, deployment location and weapon configuration.

[0018] Optionally, the current environmental situation is combined with the aforementioned multi-source heterogeneous information to analyze and assess the results. These results include ship attributes, enemy operational intentions, and the potential future sea area reached, including:

[0019] The current environmental situation is combined with the aforementioned multi-source heterogeneous information to analyze and judge the results, which include ship attributes, enemy combat intentions, and the range of sea areas that may be reached in the future.

[0020] The vessel attributes include displacement, size, equipment information, possible deployment location, activity area and historical trajectory information for the past year;

[0021] The enemy's operational intent is whether to enter our territorial waters or to demonstrate by circumventing the rules.

[0022] Optionally, based on the specific fleet group information and the analysis results, an intelligence report is generated, including:

[0023] Based on the detailed information and analysis results of the specific fleet group, a preliminary intelligence report is generated;

[0024] Conduct a standardization check on the preliminary intelligence report and provide suggestions for revision;

[0025] Integrate preliminary intelligence reports and revision suggestions, conduct logic and common sense checks, and output a complete report that supports downloading in multiple formats.

[0026] According to another aspect of the present invention, a system for detecting and recognizing ships based on large-scale models is also provided, comprising:

[0027] A multi-source heterogeneous information generation module is used to integrate multimodal information from visible light images and text intelligence information to generate multi-source heterogeneous information;

[0028] A module for generating specific fleet group information is used to generate specific fleet group information based on the multi-source heterogeneous information.

[0029] The analysis and judgment result module is used to combine the current environmental situation with the multi-source heterogeneous information to analyze and judge the results, which include ship attributes, enemy combat intentions, and the range of sea areas that may be reached in the future.

[0030] The intelligence report generation module is used to generate intelligence reports based on the detailed information and analysis results of the specific fleet group.

[0031] Optionally, a multi-source heterogeneous information module is generated, including:

[0032] A multi-source heterogeneous information generation submodule is used to generate multi-source heterogeneous information based on multi-source heterogeneous input data. The multi-source heterogeneous input data includes at least: Automatic Identification System (AIS) data files, visible light satellite image data, and intelligence text information input by the user.

[0033] The AIS data file identification submodule is used to parse the ship's Automatic Identification System (AIS) data file through the first intelligent agent and output the ship's number, category, time, latitude and longitude, speed, and direction information;

[0034] The target detection image submodule is used to perform target detection, image classification and semantic segmentation on visible light satellite images through a second intelligent agent, and output ship feature description information;

[0035] The visual analysis image submodule is used to perform visual language analysis on visible light satellite images through a third-party intelligent agent to generate a comprehensive scene description that includes entity information and environmental status.

[0036] The matching information knowledge submodule is used by the fourth intelligent agent to retrieve basic matching information from the knowledge base based on intelligence information input by the user.

[0037] Optionally, a specific fleet group information module is generated, including:

[0038] A submodule for generating specific fleet group information is used to generate ship technical parameters and ship captain information based on the multi-source heterogeneous information. The ship technical parameters include: displacement, deployment location and weapon configuration.

[0039] Optionally, the analysis and judgment results module includes:

[0040] The analysis and judgment results submodule is used to combine the current environmental situation with the multi-source heterogeneous information to analyze and judge the results, which include ship attributes, enemy combat intentions, and the range of sea areas that may be reached in the future.

[0041] The vessel attributes include displacement, size, equipment information, possible deployment location, activity area and historical trajectory information for the past year;

[0042] The enemy's operational intent is whether to enter our territorial waters or to demonstrate by circumventing the rules.

[0043] Optionally, an intelligence report generation module includes:

[0044] The module for generating a preliminary intelligence report is used to generate a preliminary intelligence report based on the detailed information and analysis results of the specific fleet group.

[0045] The Preliminary Intelligence Report Submodule is used to perform a standardization check on the preliminary intelligence report and output suggested revisions.

[0046] The Output Full Report submodule integrates preliminary intelligence reports and modification suggestions, performs logic and common sense checks, and outputs a full report that supports downloading in multiple formats.

[0047] Therefore, leveraging the large-scale model generation and reasoning capabilities, combined with knowledge from a professional knowledge base, and guided by evaluation dimension prompts, a large-scale model-based image detection and recognition application solution for maritime vessels is provided. The provided basic multimodal model possesses capabilities such as image classification, object detection, image segmentation, and image description. This basic multimodal model can integrate heterogeneous information from multiple sources, including text and images, to comprehensively analyze, understand, and judge the current environmental situation, outputting structured analysis results or reports. The analysis reports should be logically consistent, conform to common sense, and support output in formats such as JSON and XML. This provides a large-scale model-based image detection and recognition application solution for maritime vessels. Attached Figure Description

[0048] Exemplary embodiments of the present invention can be more fully understood by referring to the following figures:

[0049] Figure 1 This is a flowchart illustrating a method for detecting and recognizing ships based on a large model, as described in this embodiment.

[0050] Figure 2 This is a schematic diagram of a system for image detection and ship recognition based on a large model, as described in this embodiment. Detailed Implementation

[0051] Exemplary embodiments of the invention will now be described with reference to the accompanying drawings. However, the invention may be embodied in many different forms and is not limited to the embodiments described herein. These embodiments are provided to fully and completely disclose the invention and to fully convey its scope to those skilled in the art. The terminology used in the exemplary embodiments illustrated in the drawings is not intended to limit the invention. In the drawings, the same units / elements are referred to by the same reference numerals.

[0052] Unless otherwise stated, the terms used herein (including technical terms) have their common meaning as understood by one of ordinary skill in the art. Furthermore, it is understood that terms defined in commonly used dictionaries should be understood to have a meaning consistent with the context of their relevant field, and not to be interpreted as having an idealized or overly formal meaning.

[0053] According to a first aspect of the present invention, a method 100 for detecting and recognizing ships based on large-scale images is provided, with reference to... Figure 1 As shown, the method 100 includes:

[0054] S101: Integrate multimodal information from visible light images and textual intelligence information to generate multi-source heterogeneous information;

[0055] S102: Generate specific fleet group information based on the multi-source heterogeneous information;

[0056] S103: Combine the current environmental situation with the aforementioned multi-source heterogeneous information to analyze and judge the results, which include ship attributes, enemy combat intentions, and the range of sea areas that may be reached in the future.

[0057] S104: Generate an intelligence report based on the information of the specific fleet group and the analysis results.

[0058] Specifically, this invention addresses image detection and recognition of ships at sea based on large models. It utilizes an intelligent agent approach to perform image detection and recognition based on large models, enabling the detection of various types of ships at sea and the identification and monitoring of changes in ship type, quantity, and location.

[0059] Image classification: Classify visible light satellite images of ships to distinguish different types of ships, port areas, etc., providing a basis for subsequent analysis and reasoning.

[0060] Target detection: Automatically detects various types of vessels in images, identifies their coordinates, and counts the number of vessels in the port area.

[0061] Image segmentation: Segment the target in the visible light satellite image of the ship and select the ship's location to facilitate subsequent anomaly identification.

[0062] Image description capabilities: It can describe the feature information of visible light satellite ship images and generate structured and concise image descriptions.

[0063] Multi-source heterogeneous information integration: Based on multi-modal information such as visible light images (e.g., image information of the aircraft carrier fleet in the region) and text intelligence information (e.g., intelligence information of the aircraft carrier fleet in the region), multi-source heterogeneous information is comprehensively understood, reasoned, and judged, and integrated results and reports of comprehensive analysis are output.

[0064] A comprehensive analysis and assessment of the current environmental situation is conducted: Based on the fusion identification results, the ship's attributes, such as displacement, size, equipment information, possible deployment location, activity areas over the past year, and historical trajectory information, are comprehensively analyzed and understood. Furthermore, considering the current political environment and the deployment of surrounding supply bases, the enemy's operational intentions (whether they will enter our waters or conduct a demonstration and detour) are assessed, along with the potential sea areas they may reach in the future.

[0065] The analysis report should be logically consistent and in line with common sense: by using prompts, large model verification nodes, and the reasoning process to generate the analysis report, we can ensure that the generated analysis report is logically rigorous and consistent in terms of fact description, reasoning process, and conclusion expression, and avoid problems such as contradictions and inconsistencies with common sense.

[0066] Structured analysis results output: All target detection, recognition, analysis, judgment and reasoning processes are output in the form of structured reports, supporting JSON and XML standard formats.

[0067] (1) Agent S1-01: Document and Intelligence Analysis

[0068] Input: Multi-format Automatic Identification System (AIS) data files (xlsx, TXT, CSV, etc.).

[0069] Output: Information such as ship number, category, time, latitude and longitude, speed, and direction.

[0070] (2) Agent S1-02: Image Detection and Recognition

[0071] Input: Visible light satellite image data.

[0072] Output: Object detection, image classification, semantic segmentation results, and ship feature description information (such as the types and numbers of ships present in the detected image).

[0073] (3) Agent S1-03: Scene description generation

[0074] Input: Visible light satellite image data.

[0075] Output: Scene descriptions generated using a visual language model, including comprehensive analysis of entity information and environmental conditions.

[0076] (4) Agent S1-04: Knowledge Retrieval Input: The intelligence information entered by the user includes relationship status, ship replenishment base location information, historical navigation trajectory, departure port and time, key sea areas passed through, or specific report generation requirements.

[0078] Output: The corresponding basic information retrieved from the knowledge base based on the input information.

[0079] (5) Agent S1-05: Integration of multi-source heterogeneous information Input: Ship-related intelligence text data (such as relationship situation, ship replenishment base location information, historical navigation trajectory, departure port and time, key sea areas passed through, etc.), document and intelligence parsing results of agent S1-01, image detection and recognition results of agent S1-02, scene description generation results of agent S1-03, and knowledge retrieval output results of agent S1-04.

[0081] Output: Detailed information about a specific fleet group (such as ship technical parameters like displacement, deployment location, weapon configuration, captain information, etc.).

[0082] (6) Intelligent agent S1-06: Comprehensive analysis, understanding and judgment Input: Analysis results of agent S1-05, intelligence text information of aircraft carrier fleet, and military knowledge base data.

[0083] Output: Vessel attributes (e.g., displacement, size, equipment information, possible deployment location, activity areas and historical trajectory information for the past year). Enemy operational intentions (whether to enter our waters or to demonstrate and detour), and the range of sea areas they may reach in the future.

[0085] (7) Agent S1-07: Report Generation

[0086] Input: Ship-related intelligence text data; output information from agents S1-01 to S1-06.

[0087] Output: A preliminary report generated based on comprehensive analysis and judgment of information.

[0088] (8) Agent S1-08: Report Verification

[0089] Input: Information output by agent S1-07.

[0090] Output: The initial report generated, and the suggested modifications based on the report's specifications.

[0091] (9) Agent S1-09: Report Integration

[0092] Input: Agent S1-07 08 Output information.

[0093] Output: A complete report, including the results of logic and common sense checks, and multiple format versions (JSON, XML), with support for downloading in Word format.

[0094] The intelligence information transmitted by the intelligence department includes relationship dynamics, ship replenishment base locations, historical navigation routes, departure ports and times, and key sea areas traversed, which are then input into a large-scale model for analysis. Satellite imagery and AIS data documents are also uploaded.

[0095] The system takes a visible light satellite image of the ship as input, calls a large model to generate an image description, and uses YOLO for image classification, object detection, and semantic segmentation. The image classification, detection, and segmentation results are then fed into the information integration node.

[0096] The system takes in a visible light satellite image of the ship, calls a large model to generate an image description, and then uses a visual language model to pass the image description content to the information integration node.

[0097] The document and intelligence parsing module is used to extract AIS data to obtain specific situational information about the target.

[0098] The knowledge base recalls basic fleet information to aid in decision-making.

[0099] The multi-source heterogeneous information integration node integrates intelligence text, image recognition results, and fleet-related indicators to output detailed fleet information.

[0100] The knowledge base retrieves relevant military / political knowledge to aid in analysis, understanding, and judgment.

[0101] The large model combines information integration results, fleet intelligence information, and military / political knowledge to comprehensively analyze fleet attributes, local operational intentions, and possible future actions.

[0102] The report generation node compiles the target detection, identification, analysis, judgment, and reasoning processes into a report.

[0103] The report verification node verifies the generated report to determine whether there are any logical conflicts and whether it conforms to common sense, and attaches the verification results as an attachment.

[0104] The report integration node combines all content into a complete report, which is then displayed on the front end and supports downloading.

[0105] This invention utilizes large-scale model capabilities and knowledge bases to design maritime vessel identification, detection, and segmentation capabilities, providing an excellent alternative to manual operations. It exhibits adaptability to complex environments, effectively overcoming issues such as sea clutter and weather interference, significantly reducing identification and detection error rates. Real-time dynamic monitoring supports adaptive model updates, and combined with edge computing, it achieves near real-time analysis, shortening emergency response time to minutes.

[0106] Optionally, multimodal information from visible light images and textual intelligence information are integrated to generate multi-source heterogeneous information, including:

[0107] Based on multi-source heterogeneous input data, multi-source heterogeneous information is generated. The multi-source heterogeneous input data includes at least: Automatic Identification System (AIS) data files, visible light satellite image data, and intelligence text information input by the user.

[0108] The first intelligent agent parses the Automatic Identification System (AIS) data file of the ship and outputs the ship's number, category, time, latitude and longitude, speed, and direction information;

[0109] The second intelligent agent performs target detection, image classification, and semantic segmentation on visible light satellite images, and outputs ship feature description information.

[0110] Visual language analysis of visible light satellite images is performed by a third-party intelligent agent to generate a comprehensive scene description that includes entity information and environmental status.

[0111] The fourth intelligent agent retrieves matching basic information from the knowledge base based on intelligence information input by the user.

[0112] Optionally, based on the multi-source heterogeneous information, specific fleet group information is generated, including:

[0113] Based on the multi-source heterogeneous information, ship technical parameters and ship captain information are generated. The ship technical parameters include: displacement, deployment location and weapon configuration.

[0114] Optionally, the current environmental situation is combined with the aforementioned multi-source heterogeneous information to analyze and assess the results. These results include ship attributes, enemy operational intentions, and the potential future sea area reached, including:

[0115] The current environmental situation is combined with the aforementioned multi-source heterogeneous information to analyze and judge the results, which include ship attributes, enemy combat intentions, and the range of sea areas that may be reached in the future.

[0116] The vessel attributes include displacement, size, equipment information, possible deployment location, activity area and historical trajectory information for the past year;

[0117] The enemy's operational intent is whether to enter our territorial waters or to demonstrate by circumventing the rules.

[0118] Optionally, based on the specific fleet group information and the analysis results, an intelligence report is generated, including:

[0119] Based on the detailed information and analysis results of the specific fleet group, a preliminary intelligence report is generated;

[0120] Conduct a standardization check on the preliminary intelligence report and provide suggestions for revision;

[0121] Integrate preliminary intelligence reports and revision suggestions, conduct logic and common sense checks, and output a complete report that supports downloading in multiple formats.

[0122] Therefore, leveraging the large-scale model generation and reasoning capabilities, combined with knowledge from a professional knowledge base, and guided by evaluation dimension prompts, a large-scale model-based image detection and recognition application solution for maritime vessels is provided. The provided basic multimodal model possesses capabilities such as image classification, object detection, image segmentation, and image description. This basic multimodal model can integrate heterogeneous information from multiple sources, including text and images, to comprehensively analyze, understand, and judge the current environmental situation, outputting structured analysis results or reports. The analysis reports should be logically consistent, conform to common sense, and support output in formats such as JSON and XML. This provides a large-scale model-based image detection and recognition application solution for maritime vessels.

[0123] According to another aspect of the invention, a system 200 for detecting and recognizing ships based on large-scale images is also provided, with reference to... Figure 2 As shown, the system 200 includes:

[0124] The multi-source heterogeneous information generation module 210 is used to integrate multimodal information of visible light images and text intelligence information to generate multi-source heterogeneous information;

[0125] A module 220 for generating specific fleet group information is used to generate specific fleet group information based on the multi-source heterogeneous information.

[0126] The analysis and judgment result module 230 is used to combine the current environmental situation with the multi-source heterogeneous information to analyze and judge the results, which include ship attributes, enemy combat intentions and the range of sea areas that may be reached in the future.

[0127] The intelligence report generation module 240 is used to generate an intelligence report based on the detailed information and analysis results of the specific fleet group.

[0128] Optionally, a multi-source heterogeneous information module is generated, including:

[0129] A multi-source heterogeneous information generation submodule is used to generate multi-source heterogeneous information based on multi-source heterogeneous input data. The multi-source heterogeneous input data includes at least: Automatic Identification System (AIS) data files, visible light satellite image data, and intelligence text information input by the user.

[0130] The AIS data file identification submodule is used to parse the ship's Automatic Identification System (AIS) data file through the first intelligent agent and output the ship's number, category, time, latitude and longitude, speed, and direction information;

[0131] The target detection image submodule is used to perform target detection, image classification and semantic segmentation on visible light satellite images through a second intelligent agent, and output ship feature description information;

[0132] The visual analysis image submodule is used to perform visual language analysis on visible light satellite images through a third-party intelligent agent to generate a comprehensive scene description that includes entity information and environmental status.

[0133] The matching information knowledge submodule is used by the fourth intelligent agent to retrieve basic matching information from the knowledge base based on intelligence information input by the user.

[0134] Optionally, a specific fleet group information module is generated, including:

[0135] A submodule for generating specific fleet group information is used to generate ship technical parameters and ship captain information based on the multi-source heterogeneous information. The ship technical parameters include: displacement, deployment location and weapon configuration.

[0136] Optionally, the analysis and judgment results module includes:

[0137] The analysis and judgment results submodule is used to combine the current environmental situation with the multi-source heterogeneous information to analyze and judge the results, which include ship attributes, enemy combat intentions, and the range of sea areas that may be reached in the future.

[0138] The vessel attributes include displacement, size, equipment information, possible deployment location, activity area and historical trajectory information for the past year;

[0139] The enemy's operational intent is whether to enter our territorial waters or to demonstrate by circumventing the rules.

[0140] Optionally, an intelligence report generation module includes:

[0141] The module for generating a preliminary intelligence report is used to generate a preliminary intelligence report based on the detailed information and analysis results of the specific fleet group.

[0142] The Preliminary Intelligence Report Submodule is used to perform a standardization check on the preliminary intelligence report and output suggested revisions.

[0143] The Output Full Report submodule integrates preliminary intelligence reports and modification suggestions, performs logic and common sense checks, and outputs a full report that supports downloading in multiple formats.

[0144] The system 200 for detecting and recognizing ships based on images of a large model, according to an embodiment of the present invention, corresponds to the method 100 for detecting and recognizing ships based on images of a large model, according to another embodiment of the present invention, and will not be described again here.

[0145] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. The solutions in the embodiments of this application can be implemented in various computer languages, such as the object-oriented programming language Java and the interpreted scripting language JavaScript.

[0146] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0147] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0148] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0149] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0150] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A method for detecting and recognizing ships based on large-scale models, characterized in that, include: Multimodal information from visible light images and textual intelligence information are integrated to generate multi-source heterogeneous information; Based on the aforementioned multi-source heterogeneous information, specific fleet group information is generated; The current environmental situation is combined with the aforementioned multi-source heterogeneous information to analyze and judge the results, which include ship attributes, enemy combat intentions, and the range of sea areas that may be reached in the future. An intelligence report is generated based on the information and analysis results of the specific fleet group.

2. The method according to claim 1, characterized in that, Integrating multimodal information from visible light images and textual intelligence information generates multi-source heterogeneous information, including: Based on multi-source heterogeneous input data, multi-source heterogeneous information is generated. The multi-source heterogeneous input data includes at least: Automatic Identification System (AIS) data files, visible light satellite image data, and intelligence text information input by the user. The first intelligent agent parses the Automatic Identification System (AIS) data file of the ship and outputs the ship's number, category, time, latitude and longitude, speed, and direction information; The second intelligent agent performs target detection, image classification, and semantic segmentation on visible light satellite images, and outputs ship feature description information. Visual language analysis of visible light satellite images is performed by a third-party intelligent agent to generate a comprehensive scene description that includes entity information and environmental status. The fourth intelligent agent retrieves matching basic information from the knowledge base based on intelligence information input by the user.

3. The method according to claim 1, characterized in that, Based on the aforementioned multi-source heterogeneous information, specific fleet group information is generated, including: Based on the multi-source heterogeneous information, ship technical parameters and ship captain information are generated. The ship technical parameters include: displacement, deployment location and weapon configuration.

4. The method according to claim 1, characterized in that, The analysis and judgment results, combining the current environmental situation with the aforementioned multi-source heterogeneous information, include ship attributes, enemy operational intentions, and the potential future sea area reached, including: The current environmental situation is combined with the aforementioned multi-source heterogeneous information to analyze and judge the results, which include ship attributes, enemy combat intentions, and the range of sea areas that may be reached in the future. The vessel attributes include displacement, size, equipment information, possible deployment location, activity area and historical trajectory information for the past year; The enemy's operational intent is whether to enter our territorial waters or to demonstrate by circumventing the rules.

5. The method according to claim 1, characterized in that, Based on the specific fleet group information and analysis results, an intelligence report is generated, including: Based on the detailed information and analysis results of the specific fleet group, a preliminary intelligence report is generated; Conduct a standardization check on the preliminary intelligence report and provide suggestions for revision; Integrate preliminary intelligence reports and revision suggestions, conduct logic and common sense checks, and output a complete report that supports downloading in multiple formats.

6. A system for detecting and recognizing ships based on large-scale models, characterized in that, include: A multi-source heterogeneous information generation module is used to integrate multimodal information from visible light images and text intelligence information to generate multi-source heterogeneous information; A module for generating specific fleet group information is used to generate specific fleet group information based on the multi-source heterogeneous information. The analysis and judgment result module is used to combine the current environmental situation with the multi-source heterogeneous information to analyze and judge the results, which include ship attributes, enemy combat intentions, and the range of sea areas that may be reached in the future. The intelligence report generation module is used to generate intelligence reports based on the detailed information and analysis results of the specific fleet group.

7. The system according to claim 6, characterized in that, Generate multi-source heterogeneous information modules, including: A multi-source heterogeneous information generation submodule is used to generate multi-source heterogeneous information based on multi-source heterogeneous input data. The multi-source heterogeneous input data includes at least: Automatic Identification System (AIS) data files, visible light satellite image data, and intelligence text information input by the user. The AIS data file identification submodule is used to parse the ship's Automatic Identification System (AIS) data file through the first intelligent agent and output the ship's number, category, time, latitude and longitude, speed, and direction information; The target detection image submodule is used to perform target detection, image classification and semantic segmentation on visible light satellite images through a second intelligent agent, and output ship feature description information; The visual analysis image submodule is used to perform visual language analysis on visible light satellite images through a third-party intelligent agent to generate a comprehensive scene description that includes entity information and environmental status. The matching information knowledge submodule is used by the fourth intelligent agent to retrieve basic matching information from the knowledge base based on intelligence information input by the user.

8. The system according to claim 6, characterized in that, Generate a specific fleet group information module, including: A submodule for generating specific fleet group information is used to generate ship technical parameters and ship captain information based on the multi-source heterogeneous information. The ship technical parameters include: displacement, deployment location and weapon configuration.

9. The system according to claim 6, characterized in that, The analysis and judgment results module includes: The analysis and judgment results submodule is used to combine the current environmental situation with the multi-source heterogeneous information to analyze and judge the results, which include ship attributes, enemy combat intentions, and the range of sea areas that may be reached in the future. The vessel attributes include displacement, size, equipment information, possible deployment location, activity area and historical trajectory information for the past year; The enemy's operational intent is whether to enter our territorial waters or to demonstrate by circumventing the rules.

10. The system according to claim 6, characterized in that, The intelligence report generation module includes: The module for generating a preliminary intelligence report is used to generate a preliminary intelligence report based on the detailed information and analysis results of the specific fleet group. The Preliminary Intelligence Report Submodule is used to perform a standardization check on the preliminary intelligence report and output suggested revisions. The Output Full Report submodule integrates preliminary intelligence reports and modification suggestions, performs logic and common sense checks, and outputs a full report that supports downloading in multiple formats.

Images