A syntax tree-based structured remote sensing report generation system, method, device and medium

Abstract

The application discloses a kind of structured remote sensing report generation systems, methods, equipment and medium based on syntax tree, wherein the system includes user input layer, core processing layer and output and verification layer, system receives user demand parameter and input data, automatically parses demand parameter and standardizes input data;Generate syntax tree structure in line with industry specifications, drive big model to obtain report draft under the constraint of syntax tree;Automatic verification and manual intervention correction are carried out on report draft, and professional remote sensing report is formed.The application realizes the semantic association and professional term accurate mapping of multi-modal remote sensing data by constructing multi-dimensional syntax tree architecture, combining hierarchical syntax analysis algorithm and domain knowledge graph, supports the automatic generation of long document report in line with remote sensing geographic industry specifications, improves the structured degree and logical coherence of text generation, and can be widely applied in geological monitoring analysis, environmental assessment report generation and other fields.

Description

Technical Field

[0001] This invention relates to the field of remote sensing report generation technology, and in particular to a structured remote sensing report generation system, method, device and medium based on syntax tree. Background Technology

[0002] Large language models can significantly improve office efficiency in generating long documents. However, remote sensing reports, as a type of report with high requirements for professionalism and structure, still face limitations in practical applications compared to traditional long document generation technologies. Current automated methods for remote sensing ecological survey reports have significant shortcomings in terms of the accuracy of professional knowledge and the rigor of structure.

[0003] 1) Structure-content decoupling deficiency: The lack of a deep correlation mechanism between document syntax and professional knowledge leads to a mismatch between chapter hierarchy and ecological analysis granularity, as well as a lack of automatic verification between spatial description statements and GIS layer attributes. 2) Lack of dynamic adaptation capability: The existing system struggles to handle data-driven structural variations in remote sensing ecological research, and cannot autonomously generate fusion analysis chapters when inputting multi-source heterogeneous data. 3) Incomplete verification system: The lack of automated verification mechanisms for professional documents manifests in several ways: terminology consistency verification relies solely on keyword matching, ignoring contextual semantics; spatiotemporal logic error detection is lacking; multimodal feature fusion is insufficient; and the generation processes of text, charts, spatial data, and remote sensing image features are independent, lacking a cross-modal consistency verification mechanism.

[0004] These shortcomings have limited the large-scale application of remote sensing technology in ecological assessment. Therefore, a structured remote sensing report generation system is needed in practice. This system should construct a syntax tree architecture, combine hierarchical syntax parsing algorithms and domain knowledge graphs, to achieve semantic association of multimodal remote sensing data and accurate mapping of professional terms, and automatically generate remote sensing image report documents that conform to remote sensing geographic industry standards. Summary of the Invention

[0005] To address the aforementioned problems, the present invention aims to provide a structured remote sensing report generation system, method, device, and medium based on a syntax tree, which automatically generates report documents that conform to remote sensing geographic industry standards by constructing a multi-layer syntax tree architecture.

[0006] This invention provides a structured remote sensing report generation system, method, device, and medium based on syntax trees.

[0007] First aspect: A structured remote sensing report generation system based on syntax trees, including:

[0008] The user input layer is used to receive and integrate user request parameters and input data;

[0009] The core processing layer is used to associate input data with the knowledge base, generate a syntax tree structure that conforms to industry standards according to the required parameters, drive the large model to complete the report content filling and chart generation under the constraints of the syntax tree, and obtain the first draft of the report.

[0010] The output and verification layer is used to automatically verify the initial draft of the report and to manually correct it, resulting in a professional remote sensing report.

[0011] Optionally, the core processing layer includes:

[0012] The syntax tree building module is used to build report chapter structures, expand chapter nodes, perform semantic parsing, and perform multimodal validation;

[0013] The dynamic parsing engine module is used for contextual reasoning and syntax conflict resolution;

[0014] The intelligent generation module is used for generating prompts for large models and generating multimodal content.

[0015] Optionally, the core framework of the professional remote sensing report includes:

[0016] Metadata constraint units are used to define basic report information and data boundaries to ensure content consistency.

[0017] The core chapter unit is used to organize the main content of the report, covering the entire process from background to analysis;

[0018] The Conclusions and Recommendations section is used to extract core findings and guide practice, enhancing the report's practicality.

[0019] The second aspect: A structured remote sensing report generation method based on syntax trees, including:

[0020] Optionally, the structured remote sensing report generation method of the structured remote sensing report generation system includes the following steps:

[0021] S1. Receive user requirement parameters and input data, automatically parse the requirement parameters and standardize the input data;

[0022] S2. Associate the input data with the knowledge base, generate a syntax tree structure that conforms to industry standards according to the required parameters, drive the large model to complete the report content filling and chart generation under the constraints of the syntax tree, and obtain the first draft of the report.

[0023] S3. Automated verification and manual correction of the initial draft report to generate a professional remote sensing report.

[0024] Optionally, S1 includes:

[0025] S11. Receive user input data and perform standardized processing;

[0026] S12. Receive user-submitted requirements; the system automatically parses the requirement parameters.

[0027] The input data includes remote sensing images, vector data and / or auxiliary data, and the submitted requirements include the study area, ecological indicators and / or analysis methods.

[0028] Optionally, the syntax tree adopts a four-layer nested structure of discourse layer, chapter layer, paragraph layer and sentence layer, supporting dynamic expansion of layer nodes and multi-scale nested analysis.

[0029] Optionally, S2 includes:

[0030] S21. Associate the input data with the knowledge base through a dynamic parsing engine;

[0031] S22. Match the report framework according to the requirements parameters and dynamically expand the layer nodes;

[0032] S23. Analyze the correlation of requirement parameter data and verify the spatiotemporal logic;

[0033] S24. Perform structured analysis on the requirement parameters and intelligently generate a syntax tree structure that conforms to industry standards.

[0034] S25. Drive the large model to automatically complete the report content filling and chart generation under the syntax tree constraints, and obtain the first draft of the report.

[0035] Optionally, S3 includes:

[0036] S31. The system automatically verifies the initial draft of the report and outputs report documents in multiple formats.

[0037] S32. Manually intervene to correct the report document, forming a professional remote sensing report that can be delivered directly;

[0038] The multi-format report documents include standardized documents and / or interactive reports; automated verification includes structural integrity checks and data consistency checks; manual corrections include online revisions and automatic synchronization to the syntax tree.

[0039] Third aspect: An electronic device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the program, performs the steps of the method provided in the second aspect.

[0040] Fourth aspect: A non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method provided in the second aspect.

[0041] The beneficial effects of this invention are:

[0042] 1. This invention constructs a syntax tree architecture, combines a hierarchical syntax parsing algorithm and a domain knowledge graph to achieve semantic association and accurate mapping of professional terms for multimodal remote sensing data. It supports the automatic generation of long document reports that conform to remote sensing and geographic industry standards, improves the structure and logical coherence of text generation, and can be widely used in fields such as geological monitoring and analysis and environmental assessment report generation.

[0043] 2. The syntax tree designed in this invention covers all levels from the overall report chapters to specific sentences, while maintaining flexibility to adapt to different analysis needs, enabling multi-scale analysis, supporting dynamic expansion and nested structures, and combining professional knowledge in the field of remote sensing (such as NDVI calculation and land use classification) with natural language processing technology to ensure that the generated text is effectively combined with elements such as charts and spatial data, so that chart generation and text description are updated synchronously.

[0044] 3. During the report generation process, this invention automatically adjusts the report content structure based on the input data and generates a corresponding report description. At the same time, it introduces a verification mechanism to check the consistency of terminology, spatiotemporal logic, and data integrity, ensuring that the generated text is effectively combined with elements such as charts and spatial data.

[0045] 4. The method of this invention has been tested and applied in multiple scenarios, which can shorten the remote sensing report preparation cycle by 82%. Relying on the improved efficiency of cross-departmental data collaboration in reports, the time for data fusion from meteorological, land, and ecological departments can be reduced from 6.2 hours to 1.1 hours through the structured data anchors of the syntax tree. Through the structured constraints of the syntax tree and multimodal intelligent generation technology, it significantly solves the industry pain points of loose structure, data disconnect, and high dependence on manual labor in remote sensing professional reports, and provides key technical support for the intelligent transformation of geographic information services. Attached Figure Description

[0046] Figure 1 This is a schematic diagram of the structure of the structured remote sensing report generation system of the present invention;

[0047] Figure 2 This is a schematic diagram of the core framework structure of the remote sensing report of this invention;

[0048] Figure 3 This is a flowchart illustrating the structured remote sensing report generation method of the present invention;

[0049] Figure 4 This is a schematic diagram of the structure of the electronic device of the present invention. Detailed Implementation

[0050] Embodiments of the present invention are described in detail below. Examples of these embodiments are illustrated in the accompanying drawings, wherein the same or similar symbols denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0051] Current methods for generating long document reports of remote sensing images have shortcomings. Existing template methods have fixed hierarchical levels and cannot handle multi-scale analysis. There is insufficient integration of professional knowledge in the field of remote sensing (such as NDVI calculation and land use classification) with natural language processing technology. Furthermore, the content structure cannot be automatically adjusted based on the input data, and it is impossible to check terminology consistency, spatiotemporal logic, and data integrity.

[0052] To address the above problems, this invention provides a structured remote sensing report generation system based on syntax trees. Figure 1 This is a schematic diagram of the structure of a structured remote sensing report generation system provided in an embodiment of the present invention. The system includes: a user input layer, a core processing layer, and an output and verification layer.

[0053] The user input layer includes a requirement parameter module and a raw data module. The requirement parameter module is used to receive and integrate user requirement parameters, which include the study area and ecological indicators, etc. The raw data module is used to receive and integrate user input data, which includes remote sensing images and vector data, etc.

[0054] The core processing layer is used to associate input data with the knowledge base, generate a syntax tree structure that conforms to industry standards according to the required parameters, drive the large model to complete the report content filling and chart generation under the constraints of the syntax tree, and obtain the first draft of the report.

[0055] like Figure 1 As shown, the core processing layer includes a syntax tree building module, a dynamic parsing engine module, and an intelligent generation module.

[0056] The syntax tree building module is used to build report chapter structures, expand chapter nodes, perform semantic parsing, and perform multimodal validation.

[0057] The syntax tree can adopt a four-level nested syntax tree structure of document level, chapter level, paragraph level and sentence level, and supports dynamic expansion of layer nodes and multi-scale nested analysis (such as the hybrid analysis of regional ecological assessment and plot-level land use).

[0058] When constructing the syntax tree, industry standards are used to form a constraint-based generation mechanism. Remote sensing industry standards (such as "GB / T35645-2017 Geospatial Analysis Report Compilation Standard") are transformed into syntax tree layer node attributes, and then the report generation process is carried out through structured prompts and engineering constraints of large language models (such as LLM).

[0059] The generated report implements a multimodal verification closed loop, designs cross-modal consistency verification rules, and realizes the collaborative generation and cross-validation of text, statistical tables, and spatial layers.

[0060] The dynamic parsing engine module is used for contextual reasoning and grammatical conflict resolution; the intelligent generation module is used for generating prompt words for large models and generating multimodal content; the output and verification layer is used for automated verification and manual correction of the initial report draft to form a professional remote sensing report.

[0061] The constraints imposed by the syntax tree on large models include: overall report chapters accepting metadata constraints; synchronized updates of generated figures and text analysis (corresponding figures should be automatically re-rendered based on the time range in the text); dynamic table generation; statistical table column dimensions defined by the "Statistical Table" structure node attributes in the syntax tree results and analysis (e.g., the "Land Use Change Matrix" must include three columns: transfer area, change rate, and confidence interval); and validation of table cell values ​​(e.g., NDVI value range is limited to [-1,1], and exceeding the limit triggers an alarm). Spatial description-layer binding, for example, the "Spatial Distribution" constraint in the results analysis requires all directional terms in the text (e.g., "southeast of the study area") to be associated with GIS layer spatial partitions.

[0062] like Figure 3 As shown, the syntax tree structure defines the core framework of a professional remote sensing report. The framework is divided into three main modules: metadata constraint units, core chapter units, and conclusion / recommendation units. Through a hierarchical, nested design, it ensures that the report content conforms to industry standards and is logically rigorous, while also supporting dynamic expansion and multimodal content binding.

[0063] The metadata constraint unit is used to define the basic information and data boundaries of the report, ensuring content consistency.

[0064] The core chapter unit serves to organize the main content of the report, covering the complete process from background to analysis; the introduction includes the research background and technical route, automatically generates flowcharts and explanatory tables, and links the analysis methods; the data and methods include data sources and analysis models, dynamically inserting formulas and algorithm descriptions; the results analysis includes spatial distribution and statistical tables, with table column value ranges automatically verified according to industry standards.

[0065] The conclusion and recommendation unit serves to extract core findings and guide practice, enhancing the report's practicality; core findings automatically extract key indicators and trigger logical checks; the countermeasures library matches and relates to policy databases and generates a targeted terminology manual to confirm mandatory terminology standardization and cite industry standard documents.

[0066] The output and validation layer includes a validation module, a document format module, and a manual correction interface module. The validation module validates the report structure and report text data; the document format module is used to output reports in PDF, Word, and PPT formats; and the manual correction interface module is used for manual revision of the report.

[0067] The remote sensing report structure generation system of this invention achieves fully automated generation of professional remote sensing reports by constructing a composite architecture that integrates domain knowledge, multimodal constraints, and generative AI.

[0068] like Figure 2 As shown, based on the above system, this invention also discloses a method for generating structured remote sensing reports based on syntax trees, including the following steps:

[0069] S1. Receive user requirement parameters and input data, automatically parse the requirement parameters and standardize the input data.

[0070] This step mainly receives user input data and performs standardization processing. It also receives user-submitted requirements, and the system automatically parses the requirement parameters. After the user submits the requirements, the system automatically parses the requirement parameters and standardizes the input data (such as unifying the coordinate system and aligning the time series) to provide structured input for subsequent processing.

[0071] Input data includes remote sensing imagery (such as satellite / UAV data), vector data (such as administrative divisions, boundaries of ecological protection zones) and / or auxiliary data (such as meteorological data, topographic data), etc. Submission requirements include the study area (such as geographical boundaries, coordinate range), ecological indicators (such as vegetation cover, land use type) and / or analysis methods (such as trend analysis, classification model), etc.

[0072] S2. Associate the input data with the knowledge base, generate a syntax tree structure that conforms to industry standards according to the required parameters, drive the large model to complete the report content filling and chart generation under the constraints of the syntax tree, and obtain the first draft of the report.

[0073] This step uses a dynamic parsing engine to associate input data with a knowledge base, then matches a report framework (such as an ecological assessment template) based on the required parameters, and dynamically expands layer nodes (such as adding a "multi-source data fusion analysis" module), parses the data correlation of the required parameters, verifies the spatiotemporal logic, and constructs and expands the syntax tree.

[0074] Then, the demand parameters are analyzed in a structured manner, including semantic parsing and reasoning, parsing data correlation (such as the relationship between vegetation index and climate factors), verifying spatiotemporal logic (avoiding time inversion or spatial range conflicts), and intelligently generating a syntax tree structure that conforms to industry standards.

[0075] Finally, the large model is driven to call the associated knowledge base under the constraints of the syntax tree, automatically complete the report content filling and chart generation, and intelligently generate the first draft of the report, including the large model generating text content (generating professional descriptions under the constraints of the syntax tree), multimodal collaborative generation (automatically inserting matching charts and statistical tables), etc.

[0076] S3. Automated verification and manual correction of the initial draft report to generate a professional remote sensing report.

[0077] This step mainly involves generating standardized documents and ensuring quality. After the initial draft is generated, the system automatically performs logical and data verification, outputs documents in multiple formats for user review, and allows for manual intervention and correction, ultimately forming a professional remote sensing report that can be delivered directly.

[0078] The report documents include standardized documents (PDF, Word, PPT) and / or interactive reports (supporting map interaction and data crawling); automated verification includes structural integrity checks (chapter title hierarchy, numbering rules) and data consistency checks (matching text descriptions with chart values); manual correction includes support for online expert revisions (marking errors, adding supplementary explanations) and automatic synchronization of revised content to the syntax tree (ensuring consistency in subsequent versions).

[0079] This invention achieves a closed loop from the demand input end to the verification output end, covering the entire report generation process from demand input to verification output; it is driven by industry standards, with the syntax tree deeply bound to industry standards to ensure professionalism; at the same time, it realizes human-machine collaboration, with seamless connection between automated generation and manual correction, balancing efficiency and accuracy, significantly improving the standardization and reliability of remote sensing report generation, and is applicable to various scenarios such as ecological monitoring and disaster assessment.

[0080] Application example: Generation of the "Yangtze River Delta Vegetation Change Monitoring Report".

[0081] 1. Input parameters:

[0082] {

[0083] Study Area: "Yangtze River Delta Urban Agglomeration"

[0084] "Data source":["Landsat8 2020-2023NDVI","MODIS surface temperature"],

[0085] "Analytical Methods": ["Sen Slope Estimation", "MK Trend Test"]

[0086] }

[0087] 2. Implementation steps:

[0088] Syntax tree initialization:

[0089] Load the "Vegetation Change Monitoring" template, and automatically bind the root node to the Yangtze River Delta GeoJSON boundary;

[0090] The subsection "Surface Temperature-NDVI Correlation Analysis" has been dynamically expanded (due to the detection of MODIS data);

[0091] Constraint-based generation:

[0092] When LLM generates the "Results Analysis" paragraph, the system intercepts the following error:

[0093] [Error] The text "2022 NDVI peak value 0.85" conflicts with the calculated result 0.812 and should be automatically replaced with "0.81(±0.02)".

[0094] Insert a statistical table TB_001 containing the following columns: year / NDVI mean / slope of change / p-value;

[0095] Multimodal output:

[0096] Generate a heatmap HT_003 (spatial distribution of NDVI variation), and automatically match the color scheme to the Jungle color system;

[0097] The phrase "significant degradation in the Hangzhou Bay area" in the text is associated with the ROI region of HT_003 (spatial matching degree 99.3%).

[0098] Verification and correction:

[0099] The system detected that "2020 data coverage < 95%" and should trigger a yellow warning.

[0100] Automatic supplementary explanation: "Due to cloud coverage, the percentage of valid data in 2020 was 93.2%";

[0101] Output:

[0102] A 62-page PDF report (including 12 statistical tables and 9 thematic maps) was generated and passed the review of the "HJ 1152-2020" standard.

[0103] Manual correction takes only 8 minutes (compared to 2.5 hours using traditional methods).

[0104] This invention significantly addresses the industry pain points of loose structure, data disconnect, and high reliance on manual labor in remote sensing professional reports by using structured constraints of syntax trees and multimodal intelligent generation technology, providing key technical support for the intelligent transformation of geographic information services.

[0105] The present invention also provides an electronic device, Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention, such as... Figure 4As shown, the electronic device may include a processor, a communications interface, memory, and a communication bus, wherein the processor, communications interface, and memory communicate with each other via the communication bus. The processor can invoke logical instructions from the memory, for example, to execute the following method:

[0106] S1. Receive user requirement parameters and input data, automatically parse the requirement parameters and standardize the input data;

[0107] S2. Associate the input data with the knowledge base, generate a syntax tree structure that conforms to industry standards according to the required parameters, drive the large model to complete the report content filling and chart generation under the constraints of the syntax tree, and obtain the first draft of the report.

[0108] S3. Automated verification and manual correction of the initial draft report to generate a professional remote sensing report.

[0109] Furthermore, the logical instructions in the aforementioned memory can be implemented as software functional units and sold or used as independent products, and can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0110] This invention also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, is implemented to perform the methods provided in the above embodiments, including, for example:

[0111] S1. Receive user requirement parameters and input data, automatically parse the requirement parameters and standardize the input data;

[0112] S2. Associate the input data with the knowledge base, generate a syntax tree structure that conforms to industry standards according to the required parameters, drive the large model to complete the report content filling and chart generation under the constraints of the syntax tree, and obtain the first draft of the report.

[0113] S3. Automated verification and manual correction of the initial draft report to generate a professional remote sensing report.

[0114] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0115] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0116] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A structured remote sensing report generation system based on syntax trees, characterized in that, include: The user input layer is used to receive and integrate user request parameters and input data; The core processing layer is used to associate input data with the knowledge base, generate a syntax tree structure that conforms to industry standards according to the required parameters, drive the large model to complete the report content filling and chart generation under the constraints of the syntax tree, and obtain the first draft of the report. The output and verification layer is used to automatically verify the initial draft of the report and manually correct it to form a professional remote sensing report. The constraints that syntax trees impose on large models include: the overall report chapters accept metadata constraints, and the generation of illustrations is updated synchronously with text analysis; The table is dynamically generated. The column dimensions of the statistical table are defined by the "Statistical Table" structure node attributes in the syntax tree results and analysis. The table cell value range is validated. Spatial description - layer binding. The metadata constraint unit is used to define the basic information and data boundaries of the report to ensure content consistency.

2. The structured remote sensing report generation system according to claim 1, characterized in that, The core processing layer includes: The syntax tree building module is used to build report chapter structures, expand chapter nodes, perform semantic parsing, and perform multimodal validation; The dynamic parsing engine module is used for contextual reasoning and syntax conflict resolution; The intelligent generation module is used for generating prompts for large models and generating multimodal content.

3. The structured remote sensing report generation system according to claim 1, characterized in that, The core framework of the professional remote sensing report includes: Metadata constraint units are used to define basic report information and data boundaries to ensure content consistency. The core chapter unit is used to organize the main content of the report, covering the entire process from background to analysis; The Conclusions and Recommendations section is used to extract core findings and guide practice, enhancing the report's practicality.

4. The method for generating structured remote sensing reports according to any one of claims 1 to 3, characterized in that, Including the following steps: S1. Receive user requirement parameters and input data, automatically parse the requirement parameters and standardize the input data; S2. Associate the input data with the knowledge base, generate a syntax tree structure that conforms to industry standards according to the required parameters, drive the large model to complete the report content filling and chart generation under the constraints of the syntax tree, and obtain the first draft of the report. S3. Automated verification and manual correction of the initial draft report to generate a professional remote sensing report.

5. The structured remote sensing report generation method according to claim 4, characterized in that, S1 includes: S11. Receive user input data and perform standardized processing; S12. Receive user-submitted requirements; the system automatically parses the requirement parameters. The input data includes remote sensing images, vector data and / or auxiliary data, and the submitted requirements include the study area, ecological indicators and / or analysis methods.

6. The structured remote sensing report generation method according to claim 4, characterized in that, The syntax tree adopts a four-layer nested structure of discourse layer, chapter layer, paragraph layer and sentence layer, and supports dynamic expansion of layer nodes and multi-scale nested analysis.

7. The structured remote sensing report generation method according to claim 6, characterized in that, S2 includes: S21. Associate the input data with the knowledge base through a dynamic parsing engine; S22. Match the report framework according to the requirements parameters and dynamically expand the layer nodes; S23. Analyze the correlation of requirement parameter data and verify the spatiotemporal logic; S24. Perform structured analysis on the requirement parameters and intelligently generate a syntax tree structure that conforms to industry standards. S25. Drive the large model to automatically complete the report content filling and chart generation under the syntax tree constraints, and obtain the first draft of the report.

8. The structured remote sensing report generation method according to claim 4, characterized in that, S3 includes: S31. The system automatically verifies the initial draft of the report and outputs report documents in multiple formats. S32. Manually intervene to correct the report document, forming a professional remote sensing report that can be delivered directly; The multi-format report documents include standardized documents and / or interactive reports; automated verification includes structural integrity checks and data consistency checks; manual corrections include online revisions and automatic synchronization to the syntax tree.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps of the structured remote sensing report generation method as described in any one of claims 4 to 8.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the steps of the structured remote sensing report generation method as described in any one of claims 4 to 8.

Images