A multi-source data fusion complex terrain ecological region boundary demarcation method and system

Abstract

This invention discloses a method and system for delineating ecological zone boundaries in complex terrain using multi-source data fusion, relating to the fields of geographic information technology and ecological protection planning. The method includes: using the boundary line in the ecological protection plan as the initial boundary line and collecting multi-source heterogeneous data to construct a unified benchmark dataset; based on the initial boundary line, sequentially overlaying natural, economic, geomorphological, national land survey, and industry management boundaries on both sides of the boundary line for spatial analysis to generate a refined boundary line for internal use; conducting on-site verification of the refined boundary line through field reconnaissance or UAV aerial photography to obtain the verified boundary line; organizing experts from multiple fields to review the verified boundary line, and establishing markers for the expert-approved boundary lines, constructing a traceable spatial database containing boundary lines, boundary markers, and attribute information to achieve the physical and information-based management of the boundary line. This invention achieves high-precision, high-compliance, and high-operability delineation of ecological protection boundaries under complex terrain conditions.

Description

Technical Field

[0001] This invention relates to the fields of geographic information technology and ecological protection planning technology, and in particular to a method and system for delineating the boundaries of complex terrain ecological zones through multi-source data fusion. Background Technology

[0002] The delineation of ecological protection zone boundaries is a crucial foundation for implementing ecological space management and ensuring the integrity of ecosystems. Currently, research on ecological zone boundary delineation often relies on single data sources, such as remote sensing imagery, digital elevation models, or thematic data. In areas with complex terrain, this leads to problems such as blurred boundaries, mismatch with actual topography, and conflicts with land use and industry management boundaries. Consequently, the boundaries lack scientific rigor, are difficult to implement, and fail to meet the needs of refined ecological protection and law enforcement supervision.

[0003] In recent years, multi-source data fusion technology has demonstrated its advantages in areas such as administrative boundary demarcation and ecological boundary identification by integrating various types of data, including remote sensing, geographic information, land surveys, and industry-specific data. However, its systematic application in delineating ecological boundaries in complex terrain remains relatively lacking. Particularly in areas with significant vertical differentiation, such as mountains and hills, where ecologically sensitive areas intertwine with human activity zones, there is still a lack of operational technical processes and methodologies for coordinating multi-source data to effectively harmonize boundary alignment with natural, economic, and geographical features, current land use, and industry management scope. Summary of the Invention

[0004] The technical problem to be solved by this invention is to disclose a method and system for delineating the boundaries of ecological zones in complex terrain using multi-source data fusion. This method achieves high precision, high compliance and high operability in delineating ecological protection boundaries under complex terrain conditions, effectively supporting the scientific delineation and long-term supervision of ecological space.

[0005] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows: Firstly, a method for delineating the boundaries of complex terrain ecological zones through multi-source data fusion, the method comprising: Step 1: Obtain the protection scope boundary in the ecological protection plan as the preliminary boundary, collect multi-source heterogeneous data, and construct a dataset with unified benchmark; Step 2: Based on the preliminary boundary, extract the natural boundary, economic boundary, and geomorphic boundary for overlay analysis, identify the difference areas, and perform manual interpretation and spatial conflict resolution to obtain the preliminary verification boundary. Step 3: Based on the preliminary verification boundary, the land use map patches from the third national land survey data are overlaid on it, and the boundary is adjusted in accordance with the principle of avoiding or minimizing the cutting of construction land map patches to obtain the optimized land use boundary. Step 4: Optimize the boundaries according to land type, load the management boundaries from the industry-specific data, perform spatial overlay and conflict detection, and obtain the refined boundaries for internal processing; Step 5: Import the detailed boundary lines from the office into the mobile GIS device for field surveys, and use drone aerial photography to construct a real-world 3D model for disputed or inaccessible areas to obtain field verification boundary lines; Step 6: Based on the field verification of the boundary line, organize experts from multiple fields to conduct a joint review to obtain the expert-approved boundary line. Further construct a traceable spatial database containing boundary lines, boundary markers, and attribute information through the establishment of markers to realize the physical and information-based management of the boundary line.

[0006] Furthermore, the boundaries of protected areas in ecological protection plans are obtained as preliminary boundaries, multi-source heterogeneous data are collected, and a unified benchmark dataset is constructed, including: Step 1.1: The cloud processor receives and loads the vector data of the protection area in the ecological protection plan and determines it as the preliminary boundary. Step 1.2: Collect multi-source heterogeneous data within the spatial range defined by the preliminary boundaries; Step 1.3: Based on the collected multi-source heterogeneous data, perform coordinate transformation and standardization processing to obtain a dataset with a unified benchmark; Furthermore, based on the preliminary boundary, natural boundaries, economic boundaries, and geomorphic boundaries are extracted and overlaid to identify discrepancies. These discrepancies are then manually interpreted and spatial conflict resolved to obtain the preliminary verification boundary, including: Step 2.1: Based on the preliminary boundaries, extract vegetation boundaries from remote sensing images and water system and road boundaries from digital line maps, perform spatial overlay analysis, and obtain the difference areas. Step 2.2: Based on the differences in the regions, manual interpretation is performed to assist in the process, and spatial conflict resolution algorithms are used to generate natural and economic boundary verification lines. Step 2.3: Based on the natural and economic boundary verification lines, perform terrain factor analysis on the digital elevation model to identify steep slopes, valleys, and ridges as key terrain units; Step 2.4 involves constructing the boundaries of key terrain units and overlaying them with the natural and economic boundary verification lines to adjust the direction of the boundaries and obtain the preliminary verification boundaries.

[0007] Furthermore, based on the preliminary verification boundary, land use patches from the Third National Land Survey are overlaid with it, and boundary adjustments are made based on the principle of avoiding or minimizing the cutting of construction land patches to obtain optimized land use boundaries, including: Step 3.1: Receive the preliminary verification boundary, load the data from the Third National Land Survey, and extract land use patch data containing land category attributes; Step 3.2: Spatial overlay analysis is performed on the land use map data and the preliminary verification boundary to identify the different land use types traversed by the preliminary verification boundary. Step 3.3: Based on the areas traversing different land types, and in accordance with the principle of prioritizing ecological land use, automatically perform a spatial adjustment algorithm on the boundary orientation of the traversing construction land use patches to obtain the spatial adjustment result; Step 3.4: Based on the spatial adjustment results, update and reconstruct the boundary vector data to generate optimized land use boundaries.

[0008] Furthermore, based on land use category optimization, management boundaries from industry-specific data are loaded, and spatial overlay and conflict detection are performed to obtain refined internal boundaries, including: Step 4.1: Receive land category optimization boundaries, load industry-specific data, and extract various management boundaries. Step 4.2: Perform spatial overlay analysis on the management boundary and the land category optimization boundary to identify the overlapping areas and spatial conflict areas between them; Step 4.3: Based on the spatial conflict areas and the principle of prioritizing industry data, automatically adjust the direction of the land category optimization boundary in the conflict area according to the corresponding industry management boundary to obtain the adjustment result; Step 4.4: Based on the adjustment results, integrate the optimized land use boundaries and management boundaries to generate detailed internal boundaries.

[0009] Furthermore, the detailed boundary lines from the office are imported into mobile GIS devices for field surveys. For disputed or inaccessible areas, drone aerial photography is used to construct realistic 3D models, resulting in field-verified boundary lines, including: Step 5.1: Receive the detailed boundary lines from the office, convert them into a format specific to field reconnaissance, generate and synchronize the field reconnaissance thematic data package to the mobile GIS device; Step 5.2: By synchronizing the field survey thematic data package to the mobile GIS device, field personnel conduct on-site surveys, record the coordinates and on-site attribute information of the survey points, form field verification result data, and transmit the result data back to the cloud processor; Step 5.3: Based on the returned field verification results data, identify the disputed areas and inaccessible areas that require further verification; Step 5.4: Based on the identified disputed and inaccessible areas, plan and control the UAV to conduct aerial photography, acquire high-resolution oblique image data of the corresponding areas, and generate a real-world 3D model; Step 5.5: Based on the real-scene 3D model, perform spatial overlay and fusion analysis with the refined boundary lines in the office, and finely adjust the boundary vectors to obtain the field verification boundary lines.

[0010] Furthermore, based on the field verification of the boundary lines, a joint review was conducted by experts from multiple fields to obtain expert-approved boundary lines. Through the establishment of markers, a traceable spatial database containing boundary lines, boundary markers, and attribute information was further constructed to achieve the physical and information-based management of the boundary lines, including: Step 6.1: Based on the field verification boundary, generate an expert review data package, organize experts from multiple fields to conduct a joint review through an online review system, and automatically revise the boundary data according to the review comments to obtain the expert demonstration boundary. Step 6.2: Submit the expert-demonstrated boundary line to the competent authority's review system for compliance review, receive review feedback, adjust the boundary line results based on the feedback, and generate the approved boundary line; Step 6.3: Based on the approved boundary lines, automatically plan the boundary marker layout scheme, and generate boundary marker layout data after the scheme is approved; Step 6.4: Based on the boundary marker deployment data, establish markers, spatially associate and attribute-link the boundary lines and boundary markers, construct and store them in the database to form a traceable and dynamically updated spatial database of ecological protection zone boundaries, and realize the physical and information-based management of the boundaries.

[0011] Secondly, a system for delineating the boundaries of complex terrain ecological zones through multi-source data fusion includes: The acquisition module is used to acquire the protection scope boundaries in the ecological protection plan as preliminary boundaries, collect multi-source heterogeneous data, and construct a dataset with unified benchmarks. The adjustment module is used to extract natural boundaries, economic boundaries, and geomorphic boundaries from the preliminary boundaries, perform overlay analysis, identify discrepancies, and perform manual interpretation and spatial conflict resolution to obtain preliminary verification boundaries. Based on the preliminary verification boundaries, land use patches from the Third National Land Survey are overlaid with them, and boundary adjustments are made to avoid or minimize the cutting of construction land patches to obtain optimized land use boundaries. Based on the optimized land use boundaries, management boundaries from industry-specific data are loaded, and spatial overlay and conflict detection are performed to obtain refined boundaries for internal processing. The verification module is used to import the detailed boundary lines from the office into mobile GIS devices for field surveys, and to use drone aerial photography to construct real-scene 3D models of disputed or inaccessible areas to obtain field verification boundaries. The processing module is used to organize experts from multiple fields to conduct joint reviews of the field-verified boundaries, obtain expert-approved boundaries, and further construct a traceable spatial database containing boundary lines, boundary markers, and attribute information through the establishment of markers, thereby realizing the physical and information-based management of the boundaries.

[0012] Thirdly, a computing device includes: One or more processors; A storage device for storing one or more programs that, when executed by one or more processors, cause the one or more processors to implement the method.

[0013] Fourthly, a computer-readable storage medium storing a program that, when executed by a processor, implements the method.

[0014] The above-described solution of the present invention has at least the following beneficial effects: This invention constructs a unified multi-source heterogeneous dataset and employs a combination of multi-level progressive boundary intelligent verification, internal and external collaborative verification, and expert review mechanisms to achieve a closed-loop technology process from data integration and boundary optimization to boundary entity management. A multi-source fusion base map under a unified spatial framework was constructed. Through a progressive verification process involving natural and economic boundary extraction and conflict resolution, topographic factor analysis and optimization, land use consistency adjustment, industry management boundary fusion, field reconnaissance and UAV 3D model-assisted correction, joint review by multi-domain experts, and compliance audit, this approach effectively overcomes a series of technical problems in traditional ecological zone boundary delineation, such as boundary ambiguity due to scattered data sources and inconsistent benchmarks, boundary mismatch with the actual situation due to complex terrain and overlapping land types, insufficient compliance due to industry management boundary conflicts, and low boundary accuracy due to limited field verification methods. This approach achieves a comprehensive technical effect, including high consistency between ecological protection boundaries and complex terrain, coordination with legal land types and industry management scope, accurate field implementation, reliable compliance audit, and dynamic management through both physical and informational means. This significantly improves the scientific rigor, accuracy, and operability of ecological protection zone delineation. Attached Figure Description

[0015] Figure 1 This is a flowchart illustrating a method for delineating the boundaries of complex terrain ecological zones using multi-source data fusion, as provided in an embodiment of the present invention.

[0016] Figure 2 This is a schematic diagram of a complex terrain ecological zone boundary delineation system provided by an embodiment of the present invention, which integrates multi-source data. Detailed Implementation

[0017] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0018] like Figure 1As shown, an embodiment of the present invention proposes a method for delineating the boundaries of complex terrain ecological zones through multi-source data fusion. The method includes the following steps: Step 1: Obtain the protection scope boundary in the ecological protection plan as the preliminary boundary, collect multi-source heterogeneous data, and construct a dataset with unified benchmark; Step 2: Based on the preliminary boundary, extract the natural boundary, economic boundary, and geomorphic boundary for overlay analysis, identify the difference areas, and perform manual interpretation and spatial conflict resolution to obtain the preliminary verification boundary. Step 3: Based on the preliminary verification boundary, the land use map patches from the third national land survey data are overlaid on it, and the boundary is adjusted in accordance with the principle of avoiding or minimizing the cutting of construction land map patches to obtain the optimized land use boundary. Step 4: Optimize the boundaries according to land type, load the management boundaries from the industry-specific data, perform spatial overlay and conflict detection, and obtain the refined boundaries for internal processing; Step 5: Import the detailed boundary lines from the office into the mobile GIS device for field surveys, and use drone aerial photography to construct a real-world 3D model for disputed or inaccessible areas to obtain field verification boundary lines; Step 6: Based on the field verification of the boundary line, organize experts from multiple fields to conduct a joint review to obtain the expert-approved boundary line. Further construct a traceable spatial database containing boundary lines, boundary markers, and attribute information through the establishment of markers to realize the physical and information-based management of the boundary line.

[0019] In this embodiment of the invention, a technical approach combining multi-source heterogeneous data fusion and progressive spatial verification is adopted to construct a complete technical chain from data unification, matching of natural boundaries (vegetation, rivers), economic boundaries (roads, ditches), terrain optimization, land use coordination, industry integration to field verification, expert review and physical management. This systematically overcomes the technical problems in traditional ecological delineation, such as inaccurate boundaries, disconnect from reality, poor compliance and difficulty in implementation, caused by single data sources, inconsistent benchmarks, complex terrain, conflicting management boundaries and insufficient field verification. As a result, high-precision, high-compliance and high-operability delineation of ecological protection boundaries under complex terrain conditions is achieved, effectively supporting the scientific delineation and long-term supervision of ecological space.

[0020] In a preferred embodiment of the present invention, step 1 above may include: Step 1.1: The cloud processor receives and loads the protection area vector data from the ecological protection plan, identifying it as the preliminary boundary. Specifically, the cloud processor first establishes a dedicated network connection with the official data storage server of the ecological protection planning authority. It then initiates a request to obtain the ecological protection area vector data through a compliant data exchange protocol, clearly stating that the data is intended for ecological protection zone boundary delineation. After authentication, the server transmits the protection area vector data from the planning document to the cloud processor. Upon receiving the data, the cloud processor first verifies the integrity of the data transmission, confirming that the received file size and number of data entries match the metadata provided by the server, and that there is no packet loss or data corruption. Next, it parses the structure of the vector data, checking the continuity and logic of the inflection point coordinates, and investigating for issues such as duplicate coordinates or disordered order. After completing all verification tasks, the cloud processor officially identifies the vector data as the preliminary boundary, categorizes and saves it according to the project's preset file naming rules and storage paths, generates a data reception receipt, and sends it back to the planning authority's server. Simultaneously, it records the data source, reception time, verification results, and other traceability information in its local database.

[0021] Step 1.2 involves collecting multi-source heterogeneous data within the spatial range defined by the preliminary boundaries. Specifically, this includes: delineating the data collection area based on the preliminary boundaries, ensuring the collected data comprehensively covers this area and a surrounding buffer zone; obtaining basic geographic information data through the government's data sharing platform, including a 1:10,000 scale digital line map and a 5-meter resolution digital elevation model; the digital line map must include complete elements such as water systems, roads, residential areas, and administrative boundaries, and the digital elevation model must reflect the regional topographic relief characteristics; obtaining 1-meter spatial resolution multispectral remote sensing imagery from the natural resources department, requiring the imagery to undergo prior radiometric and atmospheric correction processing to clearly distinguish different land types and vegetation cover; and extracting land parcel data for the corresponding region from the Third National Land Survey Results Management Platform, ensuring the data includes complete primary and secondary land type information, as well as the boundary coordinates, land type code, area, and other attributes of each parcel. We coordinate with relevant government departments in the fields of ecology and environment, forestry, water resources, and culture and tourism to collect thematic data on nature reserves, drinking water source protection areas, state-owned natural forests, important wetlands, large and medium-sized reservoirs, and cultural relics protection units, clearly defining the boundaries and management attributes of each type of data. During the collection process, each type of data is registered, with detailed records of key information such as the data provider, collection time, data format, coordinate system, and accuracy indicators, establishing a complete data collection ledger.

[0022] Step 1.3: Based on the collected heterogeneous data from multiple sources, perform consistency processing, and obtain a dataset with a unified benchmark through coordinate transformation and format standardization. First, organize technical personnel to study the original coordinate system and data format of various types of collected data, clarify the differences between different data, determine the unified target coordinate system and projection method, and ensure compliance with relevant national surveying and mapping standards. A multi-source data intelligent fusion algorithm is introduced to construct a semantic rule model based on a knowledge graph. This model automatically matches the spatial reference and attribute field associations of different data sources. For remote sensing image data, professional remote sensing image processing software is used to perform geometric precision correction. First, ground control point data with accurate geographic coordinates and evenly distributed distribution within the study area are collected. These control points must be clearly identifiable landmarks on the image, such as road intersections, bridge endpoints, and corner points of landmark buildings. Then, using the software's control point selection tool, the pixel position corresponding to each ground control point is found one by one on the original distorted image, and its pixel coordinates are recorded. Next, an appropriate mathematical model is selected according to the image distortion type. The pixel coordinates of the control points and their corresponding real geographic coordinates are substituted into the model, and the unknown parameters of the model are solved using the least squares method to construct the mapping relationship between image pixels and geographic coordinates. Finally, based on the solved mapping model, each pixel on the original image is processed. Coordinate transformation is performed to calculate the accurate geographical location in the target coordinate system. Then, a bilinear interpolation algorithm is initiated for resampling. First, the corresponding floating-point coordinates in the original image are determined for each blank pixel to be filled in the target image. The four nearest neighbor pixels around these coordinates are found and their grayscale values ​​are recorded. Next, linear interpolation is performed on the pixels along the upper and lower horizontal lines respectively. Weights are calculated based on the offset of the point to be filled, and the horizontal interpolation result is obtained. Subsequently, linear interpolation is performed in the vertical direction based on these two horizontal interpolation results. Weights are calculated using the offset, and a weighted sum is obtained to obtain the grayscale value of the pixel to be filled, assigning a reasonable grayscale value to the blank pixel position after transformation. Finally, some checkpoints that did not participate in the model solution are selected, and their pixel coordinates on the corrected image are compared with the actual geographic coordinates. If the deviation exceeds the allowable range, the control point selection is readjusted or the correction model is optimized until the accuracy meets the requirements, completing the coordinate unification and distortion correction of the remote sensing image.For vector data such as digital line maps and industry-specific data, geographic information data processing software is used to perform specific format conversion operations. First, the software is opened and a new vector data project is created. The original vector data file is imported, and the software's data inspection function is used to check for topological errors, missing features, and other issues. If any are found, they are repaired. Next, in the software's output settings module, a preset standardized vector format is selected as the target format, and output parameters are configured. Then, the format conversion program is started. The software automatically reads the feature information, attribute fields, and spatial coordinates of the original data and converts them according to the syntax rules and data structure of the target format, maintaining the topological relationships between features throughout the process. After the conversion is complete, the converted vector file is exported and its integrity is verified. The file is opened to check whether the number of features, attribute fields, and spatial locations are consistent with the original data. After confirming that no data is lost or incorrect, the vector data format standardization is complete. For land parcel data from the Third National Land Survey, the focus is on standardizing its attribute fields. Field names, data types, and encoding rules are unified according to preset field specifications, and missing necessary attribute information is supplemented. After all data processing is completed, the various types of data are overlaid and compared in the same coordinate system to check for spatial location deviations or data conflicts. Any problems found are corrected in a timely manner, ultimately forming a multi-source dataset with unified spatial coordinates, standardized data format, and standardized attribute information. At the same time, a data processing report is prepared, which records in detail the processing flow, parameter settings, and processing results for each type of data.

[0023] In this embodiment of the invention, by employing a series of collaborative data processing and analysis techniques, such as compliant collection and ledger registration of multi-source heterogeneous data, coordinate transformation and standardization, the technical problems of single data source, inconsistent coordinate benchmarks leading to difficulties in overlay analysis, and the difficulty in systematically discovering and locating spatial conflicts between different data sources in traditional ecological protection boundary delineation are overcome. This achieves the technical effect of realizing accurate spatial integration of multi-source data, providing a high-precision, traceable, and complete fusion base map and problem list for boundary scientific verification, and laying a solid data foundation for subsequent boundary optimization.

[0024] In a preferred embodiment of the present invention, step 2 above may include: Step 2.1: Based on the preliminary boundaries, extract vegetation boundaries from the remote sensing imagery and water system and road boundaries from the digital line map. Perform spatial overlay analysis to obtain the differential areas. Specifically, this includes: first, loading the generated preliminary boundary layer into the geographic information system software, confirming that the spatial reference of this layer is consistent with the data used in subsequent analysis, and completing the reception and loading of the preliminary boundaries. For the high-resolution multispectral remote sensing imagery that has undergone geometrical fine correction, call the software's band calculation tool and input the normalized vegetation index formula as follows: NDVI = (NIR - Red) / (NIR + Red) ; in, NIR For near-infrared reflectivity, Red For red light band reflectivity, NDVI To calculate the normalized vegetation index, the image is calculated pixel by pixel. NDVI Values ​​are used to generate coverage analysis areas. NDVI Raster layer; select at least 50 typical vegetation sample areas using a stratified random sampling method, covering different vegetation types such as trees, shrubs, and herbs, with each sample area no smaller than 10×10 pixels; and at least 30 non-vegetation sample areas, covering types such as water bodies, built-up land, and bare soil, with each sample area no smaller than 10×10 pixels. Extract all sample areas. NDVI The minimum, maximum, and mean values ​​of vegetation sample areas and non-vegetation sample areas are statistically calculated. The median of the minimum value of vegetation samples and the maximum value of non-vegetation samples is used as the initial threshold. 100 verification points (50 vegetation points and 50 non-vegetation points) are randomly selected. The actual land cover type at the verification points is compared with the initial threshold segmentation results to calculate the classification accuracy. If the accuracy is lower than 90%, the threshold is adjusted in steps of 0.01 and the verification is repeated until the accuracy is ≥90% to determine the final vegetation index threshold. The final threshold is applied to the NDVI raster layer for threshold segmentation. Pixels with NDVI greater than the threshold are identified as vegetation-covered areas. The Canny edge detection algorithm is then used to extract pixel-level edges of the vegetation-covered areas. The specific steps are as follows: First, the binarized vegetation cover raster layer after threshold segmentation is processed with Gaussian filtering using a 5×5 Gaussian kernel and a standard deviation of 1.4 to smooth the image and remove noise interference. Second, the gradient magnitude and direction of the filtered image are calculated using... x direction Sobel Operator ; y direction Sobel Operator Convolve the filtered feature map and calculate the gradient strength. ; G Represents the gradient intensity of a pixel, where G x for x directional convolution results G y for y The result of directional convolution also includes the gradient direction θ = arctan2( G y , G xThe first step involves quantizing the gradient into four main directions (0°, 45°, 90°, 135°). The second step involves performing non-maximum suppression, iterating through each pixel and comparing the gradient magnitude of the current pixel with the magnitudes of two adjacent pixels along the gradient direction. Only pixels with local maximum values ​​are retained, non-edge pixels are suppressed, and the edge contours are refined. The third step involves setting a dual threshold to filter edges, with the low threshold set to 0.1 and the high threshold set to 0.3. Pixels with gradient magnitudes greater than the high threshold are marked as strong edges, those between the high and low thresholds are marked as weak edges, and those below the low threshold are discarded. The fourth step involves edge connection, using strong edges as the core, incorporating the weak edges connected to them into the final edge, discarding isolated weak edges, and generating complete pixel-level edges of the vegetation cover area. Subsequently, the edge pixels are converted to the standardized vector format determined in step 1.3, and the converted vector boundaries are smoothed to remove burrs and breakpoints. After correcting topological errors, a vegetation cover boundary layer is generated. Open the formatted digital line drawing layer and extract water system elements (including rivers, lakes, reservoirs, ditches, etc.) and transportation elements (including national, provincial, county, township, village roads and railways, etc.) using the feature filtering tool to form natural and economic boundary layers. Perform pairwise overlay analysis on the natural and economic boundary layers and the preliminary boundary layer, recording the boundary and its relative position, distance from the shoreline, and span length. Integrate all overlay analysis results to obtain a complete dataset of difference regions containing location coordinates, size, and difference type. Each difference region is labeled with the corresponding contrast data layer and specific difference characteristics.

[0025] Step 2.2: Based on the discrepancies in the regions, manual interpretation is assisted, and spatial conflict resolution algorithms are used to generate natural and economic boundary verification lines. Specifically, this includes: receiving the dataset of discrepancies in the regions; associating each region with corresponding high-resolution remote sensing imagery, digital line maps, and other raw data in the geographic information system software; generating a visual analysis base map of the discrepancies, which must clearly present the geographical location, surrounding landform features, and raw data information of the discrepancies. Technical personnel with an ecological geography background are organized to conduct manual interpretation. These personnel examine each discrepancy region one by one using the visual analysis base map, combining the landform texture features of the remote sensing imagery and the topographic element information of the digital line map to determine the causes of the discrepancies. For example, are the discrepancies due to data accuracy issues, discrepancies due to the actual distribution of natural geographical features not matching the boundaries, or discrepancies due to human activities altering landforms? The interpretation conclusions and preliminary adjustment suggestions for each discrepancy region are recorded in detail. The interpretation record must include the discrepancy region number, geographical location, discrepancy type, causal analysis, and adjustment direction. After manual interpretation, the spatial conflict resolution algorithm is initiated. First, a conflict determination rule base is constructed, covering thresholds for the fit between the boundary and vegetation cover boundaries (e.g., the deviation distance between the boundary and vegetation boundary should not exceed 5 meters), and thresholds for the distance between the boundary and water system shorelines (e.g., the boundary should maintain a distance of at least 3 meters from the river shoreline), as well as spatial matching rules between the boundary and road boundaries. The manual interpretation conclusions and data on discrepancies are input into the algorithm. Based on the rule base, the algorithm determines the conflict level for each discrepancy area, distinguishing between three levels: minor deviation (deviation ≤ 5 meters), moderate conflict (5 meters < deviation ≤ 20 meters), and severe conflict (deviation > 20 meters). For areas with minor deviations, the basic boundary orientation is maintained, with only minor adjustments to the inflection point coordinates to ensure the deviation meets the threshold requirements. For moderate conflict areas, the boundary segments are corrected according to the actual orientation of natural and economic boundaries, retaining the original inflection point coordinates as a reference during the correction process. For severe conflict areas, the boundary segment orientation is replanned based on the manual interpretation suggestions, prioritizing alignment with natural and economic boundaries. After processing, the corrected boundary vector data is output. A topological check is performed on the corrected boundary to fix issues such as broken features, non-overlapping nodes, and self-intersections, ensuring the topological integrity of the boundary. Finally, natural and economic boundary verification boundary layers are generated, and a complete record of conflict resolution is saved, including rule base parameters, conflict level determination results, and coordinate comparisons before and after boundary adjustment.

[0026] Step 2.3: Based on the natural and economic boundary verification lines, perform terrain factor analysis on the digital elevation model to identify steep slopes, valleys, and ridges as key terrain units. Specifically, this includes: loading the digital elevation model data that has undergone unified spatial benchmark processing into the geographic information system software, confirming that the data resolution reaches 5 meters, the elevation accuracy meets the requirements of terrain analysis, and there are no missing values, abnormal elevation points, or other data quality issues; simultaneously loading the natural and economic boundary verification line layer, using the area defined by the boundary as the analysis area, calling the software's clipping tool, setting the clipping range to the bounding rectangle of the natural and economic boundary verification lines, and retaining the terrain data inside the boundary. The clipped terrain data must completely match the boundary range, with no areas exceeding or missing. The software's terrain factor calculation tool is invoked, and the third-order inverse distance weighted square algorithm is initiated to perform calculations for various terrain factors sequentially. For slope factor calculation: First, the core parameters for slope calculation are determined, and the grid neighborhood window is set to 3×3. Centered on the target grid, eight neighboring grids are selected as calculation samples. The third-order inverse distance weighted square algorithm is used to interpolate and optimize the elevation value of each target grid. The algorithm uses the reciprocal of the squared distance between the target grid and its neighboring grids as the weight. The weight calculation formula is as follows: w=1 / d 2 , w Let d be the distance between the target grid and its adjacent grids. The elevation values ​​of adjacent grids are weighted and summed to obtain the optimized elevation value of the target grid. Using differential geometry principles, the slope value is calculated by the angle between the normal vector of the terrain surface and the plumb line, in degrees. The calculation formula is: ; in, for x Rate of change of elevation in direction for y Rate of change of elevation in direction The slope value is calculated grid-by-grid, generating a slope factor raster layer where each grid pixel corresponds to a unique slope value. Aspect factor calculation: Based on the elevation change rates in the x and y directions obtained from the slope calculation, it is calculated using the formula... Aspect=arctan2( , ) ,in AspectThe azimuth angles range from 0° to 360°, where 0° represents true north, 90° represents true east, 180° represents true south, and 270° represents true west. The calculated azimuth angles are categorized and organized to determine the slope aspect corresponding to each grid cell, generating a slope aspect factor raster layer. Topographic relief calculation: The analysis window is set to 5×5, and the entire cropped digital elevation model data is traversed with each target grid cell as the center. Within each window, the elevation values ​​of all grid cells are weighted and interpolated using a third-order inverse distance weighted square algorithm to optimize the continuity of the elevation data. Then, the maximum and minimum interpolated elevation values ​​within the window are found, and the difference between them is calculated and used as the topographic relief value of the target grid cell. After completing the calculation window by window, a topographic relief raster layer is generated. Assisted in calculating terrain factors: When calculating valley depth, valley lines are first identified through terrain curvature analysis. Areas with negative curvature and large absolute values ​​are identified as valleys. Using the elevation value of the grid cell containing the valley line as a reference, the elevation difference between the valley line and the surrounding non-valley grid cells within a 5×5 window is calculated. The average of the differences is taken as the valley depth value of the valley grid cell, generating a valley depth raster layer. When calculating ridgeline density, ridgelines are first identified through terrain curvature analysis. Areas with positive curvature and large absolute values ​​are identified as ridges. The total length of ridgelines within each 1km×1km grid is counted, generating a ridgeline density raster layer to improve the terrain factor dataset. A topographic-ecological coupling analysis model was introduced, and a machine learning-based model was constructed to correlate topographic features with ecological sensitivity. Topographic factors such as slope, aspect, topographic relief, valley depth, and ridge density were used as input features, and ecological sensitivity levels were used as labels to train the model for adaptive identification of key topographic units. Simultaneously, an ecological connectivity index was introduced to quantify the ecological correlation strength between different topographic units. Based on the ecological protection requirements of the study area, thresholds for topographic factor determination were set: areas with a slope greater than 25 degrees were directly identified as steep slope areas; areas with a difference greater than 50 meters between aspect and topographic relief were jointly analyzed, and areas on shady slopes with a topographic relief difference exceeding 50 meters were identified as valley areas; ridge lines were extracted from the digital elevation model using a ridge line extraction algorithm, and the boundary of the ridge area was delineated by extending 50 meters to both sides of the ridge line. The identified steep slopes, valleys, and ridges are converted into vectors by using the software's raster-to-vector tool and setting the vector output format to a standardized vector format. The raster's attribute information is preserved during the conversion process. Attribute fields are added to each terrain unit, including detailed information such as the average slope, elevation range, average terrain relief, average valley depth, and ridge line density, ultimately forming the boundary layer of the key terrain units.

[0027] Step 2.4 involves overlaying the key terrain unit boundaries with the natural and economic boundary verification lines, adjusting the boundary alignment, and obtaining preliminary verification boundaries. Specifically, this includes loading the generated key terrain unit boundary layer and the generated natural and economic boundary verification line layer into the same geographic information system (GIS) to ensure complete spatial consistency, with coordinate precision retained to six decimal places. The pairwise overlay analysis tool is then activated. First, the natural and economic boundary verification lines are overlaid with steep slope unit boundaries, recording the location, length, and average slope of the sections crossing steep slopes, and marking the start and end coordinates of each section. Next, the natural and economic boundary verification lines are overlaid with valley unit boundaries, marking the locations where the boundaries cross valleys, their deviation from the valley centerline, and recording the valley depth and width. Finally, the natural and economic boundary verification lines are overlaid with ridge unit boundaries, identifying sections where the boundary lines do not coincide with the ridgeline, and measuring the length and deviation of these sections. Based on the overlay analysis results, the following rules for adjusting the boundary alignment are established: For boundary segments traversing steep slope areas, if the boundary divides the steep slope area into scattered patches, the boundary alignment is adjusted to completely encompass the steep slope area, with the minimum area of ​​each patch not less than 0.1 hectares. If the boundary does not cover the core area of ​​the steep slope area, the boundary is extended to the edge of the steep slope area, with the extension distance ensuring complete coverage. For boundary segments crossing valleys, the boundary is adjusted to align with the valley's orientation, using the valley's centerline as a reference. The parallelism error between the adjusted boundary and the valley's centerline should not exceed 5°. Using the ridgeline as a natural boundary, the coordinates of the boundary's inflection points are corrected to extend the boundary along the ridgeline, with the corrected inflection point coordinates no more than 10 meters from the ridgeline. The natural and economic boundary verification lines are corrected segment by segment according to the adjustment rules. During the correction process, the original coordinates and the coordinates after adjustment of each adjusted segment are retained, and the basis for adjustment, the adjustment range, and the terrain matching degree after adjustment are recorded. After all segments are adjusted, the adjusted boundary lines are checked for topological consistency to ensure that the boundary lines have no self-intersections, no missing elements, and no topological errors. At the same time, the boundary range before and after adjustment is compared, and the area, number of inflection points, total length of the boundary lines, and other indicators are statistically analyzed to verify the rationality of the adjusted boundary lines. Finally, a preliminary verification boundary line layer is generated, and a complete record of the boundary line adjustment is saved, including the overlay analysis report, adjustment rules, coordinate comparison table, statistical indicators, etc.

[0028] In this embodiment of the invention, a series of technical means are employed, including NDVI-based adaptive threshold segmentation for accurate extraction of vegetation boundaries, Canny edge detection and multi-source natural and economic boundary overlay analysis, quantitative calculation of terrain factors and intelligent identification of key terrain units, and rule-driven spatial conflict resolution and adaptive boundary adjustment, to achieve deep integration and high-precision fitting of natural and economic boundaries and terrain features. By analyzing vegetation indexes and edge detection from high-resolution remote sensing imagery, combined with water system and road data from digital line maps, the automatic and refined extraction of natural and economic boundaries was achieved. Furthermore, by overlaying these boundary lines with the preliminary boundary lines, discrepancy areas were accurately identified. Digital elevation models were then used to calculate multiple factors such as slope, aspect, and topographic relief, scientifically identifying key topographic units such as steep slopes, valleys, and ridges. Finally, the boundary alignment was intelligently adjusted according to pre-defined ecological topographic matching rules. This effectively overcame core technical challenges in complex terrain areas, such as low efficiency of manual interpretation, inaccurate boundary extraction, severe disconnect between the boundary lines and actual topography, and difficulty in reflecting real ecological and geographical units due to large variations in vegetation cover and complex water systems and topographic features. This resulted in a high degree of spatial alignment between the ecological protection zone boundary and natural, economic entities, and topographic features, achieving a harmonious and unified ecological function. This significantly improved the scientific rigor of boundary delineation, the accuracy of topographic matching, and the interpretability of the results.

[0029] In a preferred embodiment of the present invention, step 3 above may include: Step 3.1: Receive the preliminary verification boundary, load the Third National Land Survey data, and extract land use patch data containing land category attributes. Specifically, this includes: First, loading the preliminary verification boundary layer in the project engineering of the Geographic Information System (GIS) software. Using the software's layer attribute check function, confirm that the spatial reference of this layer is the CGCS 2000 coordinate system, the projection method is consistent with the Third National Land Survey data, and the coordinate precision is retained to six decimal places. This completes the acquisition and loading confirmation of the preliminary verification boundary. Then, through the software's database connection module, load the Third National Land Survey vector dataset, which has undergone unified spatial benchmark conversion. This dataset must contain core attribute fields such as land category patches covering all county-level and above administrative regions, land category codes, land use rights, patch area, and location unit. The software's feature filtering tool was used, and filtering conditions were set in the attribute filtering interface to retain only vector features of land use type, while non-plot auxiliary features such as annotation points, administrative boundaries, and land use boundaries were removed. Simultaneously, attribute field filtering was used to retain core attribute fields such as land use code, land use name, plot number, area, and ownership code, while redundant survey record numbers and unnecessary fields such as remarks were deleted, completing the initial extraction of land use plot data. The extracted land use plot data underwent full-process preprocessing: First, the software's topology check tool was run to check for topological errors such as overlapping plots, gaps, hanging nodes, and broken lines. For overlapping plots, main plots with clear ownership and large area proportions were retained, and the overlapping parts were deleted. For gaps, if the gap area was less than 0.001 hectares, it was automatically merged into the adjacent main plot; if it was greater than 0.001 hectares, it was marked as pending verification. For hanging nodes and broken lines, node snapping and line connection tools were used to repair them, ensuring that all plot features had topological closure. Secondly, land use attribute verification is carried out. Finally, the preprocessed patch data is saved as a standardized vector format to generate a land use patch layer containing complete land use attributes. Each patch in the layer is bound with a unique identifier code and complete land use attribute information.

[0030] Step 3.2 involves spatially overlaying the land use patch data with the preliminary verification boundary to identify the different land use types traversed by the preliminary verification boundary. Specifically, this includes loading the pre-processed land use patch layer and the preliminary verification boundary layer into the same geographic information system (GIS) project. Using the software's spatial reference matching tool, the coordinate systems and projection parameters of the two layers are confirmed to be completely consistent, with no spatial offset issues. The software's line-surface overlay analysis tool is then launched, selecting the overlay modes of line-segmented surface and surface attribute assignment. Using the preliminary verification boundary as input and land use patches as overlay elements, a pairwise overlay analysis is performed. The software automatically segments the preliminary verification boundary according to its intersections with different patches, ensuring that each segment corresponds to only one land use patch's land use attribute. Simultaneously, the land use code, land use name, and primary land use type attributes of the patch are automatically assigned to the corresponding boundary segments. After the overlay analysis is completed, a boundary crossing land use analysis result layer and attribute statistics table are generated. In the analysis result layer, boundary segments corresponding to different land use types are rendered with different colors for easy visualization and identification. The attribute statistics table includes information such as the unique ID of the boundary segment, the corresponding map patch number, the land use code, the land use name, the starting coordinates of the boundary segment, the ending coordinates of the boundary segment, the segment length, the area of ​​the map patch, and the ownership code. Based on this result, the system sorts out all land use type areas crossed by the boundary for preliminary verification: First, it counts the number of primary land use types crossed and the total length of the boundary corresponding to each primary land use type. Second, for key land use types such as construction land, it further subdivides the secondary land use types crossed, such as urban residential land, industrial and mining land, and transportation facility land, and records the number of boundary segments, total length, and specific geographical location corresponding to each secondary land use type. Finally, it classifies and labels the cases where the boundary crosses a single land use type map patch and crosses multiple land use type map patches, forming a complete boundary crossing land use analysis dataset. All data are associated with accurate spatial coordinates and attribute information.

[0031] Step 3.3: Based on the regions traversing different land types, and in accordance with the principle of ecological land priority, automatically execute a spatial adjustment algorithm on the boundary direction of the construction land patches to obtain the spatial adjustment results. Specifically, based on the land type analysis dataset of the boundary crossing, firstly, formulate quantitative implementation rules for the principle of ecological land priority: clarify that ecological land includes land types with ecological protection functions such as forest land, grassland, water area and water conservancy facilities land, and wetlands, while construction land includes land types for artificial construction such as urban and rural land, industrial and mining land, and transportation land; that is, the boundary should extend along the ecological land boundary first, avoiding crossing the construction land. If it is necessary to be adjacent, it should maintain a distance of not less than 5 meters from the construction land boundary, and the adjusted boundary should not reduce the protection scope of the core ecological land. An embedded conflict automatic identification and intelligent adjustment algorithm is used as the core to launch a spatial adjustment algorithm. The algorithm execution flow is as follows: First, extract all boundary segments that cross the construction land area as the core processing object. At the same time, extract spatial data within a 500-meter radius of these segments, including the boundary coordinates of the construction land, the distribution range of surrounding ecological land, and the boundary nodes of adjacent non-construction land patches. Second, perform spatial analysis on each boundary segment crossing the construction land, calculating the crossing length of the boundary within the construction land, the coordinates of the crossing midpoint, the shortest distance to the construction land boundary, and the coordinates of the connection point between the surrounding ecological land boundary and the construction land boundary. Third, select the optimal adjustment path. If there are continuous ecological land boundaries adjacent to construction land boundaries around the boundary line, the algorithm automatically uses the ecological land boundary as the adjustment benchmark, corrects the boundary line inflection point coordinates, and makes the boundary line extend along the outer side of the boundary between ecological land and construction land, completely avoiding the construction land area; if the construction land it crosses is a scattered small patch with an area of ​​less than 0.1 hectares, the algorithm calculates the minimum bounding rectangle boundary of the patch, adjusts the boundary line inflection point to bypass the patch, and the adjusted boundary line maintains a distance of not less than 5 meters from the patch boundary; if there are no continuous ecological land boundaries around the construction land, the algorithm prioritizes the path with the shortest crossing length and the least impact on the ecological land protection area to fine-tune the boundary line, ensuring that the adjustment range is controlled within 10% of the original boundary line direction.

[0032] Step 3.4: Based on the spatial adjustment results, update and reconstruct the boundary vector data to generate optimized land use boundaries. This includes: First, based on the spatial adjustment results, integrate the boundary vector data. Load the adjusted construction land crossing boundary and the original boundary that does not cross construction land and does not require adjustment into the same layer. Call the software's feature merging tool, match the starting node with the ending node of the adjacent segment according to the spatial connection relationship of the boundary segments, and stitch all segments sequentially. During the stitching process, enable the node snapping function with a snapping tolerance of 0.5 meters to ensure that the node coordinates of adjacent segments are completely coincident, without breakpoints or overlaps, forming complete preliminary integrated boundary vector data. Then, reconstruct and optimize the preliminary integrated boundary, using the core integration of the Douglas-Pock algorithm and the Bézier curve smoothing algorithm. The Douglas-Pock algorithm is used for node simplification. The specific operation process is as follows: First, set the core parameters of the algorithm, and accurately set the simplification tolerance to 2 meters.

[0033] Treating a single boundary segment as an independent processing unit, extract the coordinate sequence of all nodes of that boundary segment, denoted as... P 0 ( x 0 , y 0 ), P 1 ( x 1 , y 1 ),..., P n ( x n ,y n ),in, P 0 The first node of the segment boundary. P 1 , P 2 ,..., P n-1 This is the middle node of the boundary line. P n The end node of the segment boundary. x 0 First node P 0 x-coordinate y 0 First node P 0 The ordinate, x 1 The first intermediate node P 1 x-coordinate y 1 The first intermediate nodeP 1 The ordinate, x n Tail node P n x-coordinate, y n Tail node P n The ordinate of , n This represents the total number of nodes for this boundary segment minus 1, connecting the first node. P 0 Tail node P n A baseline line L is formed.

[0034] Traverse all intermediate nodes of the boundary segment. P 1 to P n-1 Using the formula for the distance from a point to a line Where d represents the perpendicular distance from a point to a line, and A, B, and C are the coefficients of the line. The general formula for line L is Ax + By + C = 0. Calculate the perpendicular distance from each intermediate node to the baseline line L point by point; filter out nodes with a perpendicular distance greater than 2 meters; delete redundant nodes with a perpendicular distance less than 2 meters; reassemble the retained nodes into a new coordinate sequence, and repeat the operation of connecting the first and last nodes, calculating the perpendicular distance of intermediate nodes, and filtering the retained nodes. Iterate until the perpendicular distance from all intermediate nodes to the line connecting the first and last nodes is ≤ 2 meters, thus completing the node simplification of a single boundary segment. Then, process all boundary segments segment by segment according to this process to ensure that the simplified boundary not only eliminates meaningless redundant nodes but also fully retains the key geographical feature connection points without changing the core direction of the boundary.

[0035] The Bézier curve smoothing algorithm is used for boundary fitting. The specific operation process is as follows: Set the core parameters of the algorithm, select a third-order Bézier curve, and set the control point spacing to 5 meters; use the simplified boundary node as the anchor point, and denote adjacent anchor points as... P 0 ( x 0 , y 0 )and P 1 ( x 1 , y 1 Two third-order Bézier curve control points are generated between the two anchor points. Q 0 ( x q0 , y q0 )and Q1 ( x q1 , y q1 The coordinates of the control points are calculated using the spatial orientation of the anchor points. Q 0 x-coordinate x q0 = x 0 +( x 1 - x 0 ) / 3, ordinate y q1 = y 0 +( y 1 - y 0 ) / 3+ k Where k is the offset in the direction of the perpendicular line of the line connecting adjacent anchor points, and its value is 1 / 3 of the anchor point spacing; Q 1 x-coordinate x q1 = x 1 -( x 1 - x 0 ) / 3, ordinate y q1 = y 1 -( y 1 - y 0 ) / 3+ k ; using the formula for third-order Bézier curves Where t∈[0,1], values ​​are taken in steps of 0.01, and the coordinate sequence of the smooth curve is calculated point by point to replace the inflection points of the original boundary line; the curvature of the curve is constrained during the fitting process, and the curvature formula is used to determine the curve curvature. ,in This represents the curvature value of the curve at a certain point. These are the parameter variables of the curve. ( ) represents the curve in the parameter The x-coordinate function at the location, ( ) represents the curve in the parameter The vertical coordinate function at the location is controlled with a curvature radius ≥ 3 meters. The fourth step is to convert the fitted smooth curve coordinate sequence into vector line elements, precisely connecting them to the key anchor points of the original boundary line, with a coordinate error ≤ 0.1 meters at the connection point. After completing node simplification and smoothing, the topology inspection tool is activated to perform precise repairs for different topology problems: one is self-intersection repair, where the algorithm automatically traverses all node coordinates of the boundary line and identifies self-intersecting coordinate points through coordinate comparison. x i , y i The algorithm performs three main steps: First, it calculates the circumscribed rectangle of the intersecting region, extracts all boundary segments within that region, deletes duplicate intersecting segments, and recalculates the inflection point coordinates of the intersection point, ensuring that the boundary retains only one continuous path at that location, and that the distance between the new inflection point and the surrounding land use boundaries is ≥5 meters. Second, it repairs suspended nodes: for suspended nodes that are not connected to adjacent segments, the algorithm automatically searches for valid nodes within a 5-meter radius of the node, completes node connection through coordinate matching (coordinate error ≤0.5 meters), eliminates breakpoints, and ensures boundary continuity. Third, it repairs topological gaps: for tiny gaps within the boundary area (area <0.001 hectares), the algorithm extracts the coordinates of the boundary nodes around the gap. Add new nodes and close gaps; for gaps with an area > 0.001 hectares, mark the gap's geographical location, area, surrounding land types, etc., and list them as items to be manually verified. After reconstruction, update the boundary attribute information, and assign the land type analysis results, adjustment parameters from step 3.3, such as adjusting distance and direction, reconstruction algorithm parameters such as Douglas-Puck tolerance of 2 meters, third-order Bézier curve, control point spacing of 5 meters, radius of curvature ≥ 3 meters, and adjustment basis to the corresponding boundary segments one by one; at the same time, use attribute verification tools to check the completeness and accuracy of the information field by field to ensure that each boundary segment is bound to traceable analysis, adjustment, and reconstruction records. Finally, the updated and reconstructed boundaries undergo final verification using a three-layer verification logic: The first layer is spatial extent verification, which uses area calculation tools to calculate the area of ​​the ecological protection zone before and after adjustment based on the polygonal range enclosed by the boundary, ensuring that the adjusted area is greater than or equal to the area before adjustment, and that the coverage ratio of core ecological land has not decreased or been reduced; the second layer is land use matching verification, which performs pairwise overlay analysis again, overlaying the reconstructed boundary with land use patch data, and checking the spatial relationship between the boundary and construction land segment by segment, confirming that there are no boundary segments that cross construction land, and that all boundary segments adjacent to construction land are at least 5 meters away from the construction land boundary, in accordance with the ecological land priority rule; the third layer is format standardization verification, which checks the file format, coordinate system, attribute field encoding, and coordinate precision of the boundary vector data. After the verification is passed, the land use optimized boundary layer is officially generated, and a complete process record is saved.

[0036] In this embodiment of the invention, a series of technical means are employed, including precise extraction of land use patches based on attribute filtering and topology preprocessing, intelligent identification with line-surface overlay analysis and automatic attribute assignment, spatial adjustment algorithms driven by ecological land priority rules, and vector reconstruction optimization integrating Douglas-Puk node simplification and Bézier curve smoothing, to standardize and clean the land survey data and perform topological repair, ensuring the integrity and accuracy of land use patch data. Furthermore, through high-precision line-surface overlay analysis, the spatial crossing relationship between ecological protection boundaries and various land use patches is automatically identified and quantified. Automated spatial path optimization and adjustment are performed on boundary segments crossing construction land. This effectively overcomes a series of technical difficulties in traditional demarcation methods, such as low efficiency of manual interpretation, complex land use boundaries leading to serious conflicts between boundaries and the legal land use status, fragmentation of protection areas caused by construction land crossings, and the potential for simple adjustments to damage the integrity of ecological land. As a result, the ecological protection zone boundaries and the legal land use patches of the third national land survey are highly coordinated, completely avoiding or minimizing crossings of construction land areas, significantly improving the compliance, manageability, and operability of the boundary results and spatial planning.

[0037] In a preferred embodiment of the present invention, step 4 above may include: Step 4.1: Receive land use optimization boundaries, load industry-specific data, and extract various management boundaries. Specifically, this includes: First, loading the land use optimization boundary layer into the project module of the GIS software. Confirming the spatial reference of this layer using the layer attribute check function is the CGCS 2000 coordinate system and Gauss-Kruger projection, with coordinate precision retained to six decimal places, completes the receipt and loading confirmation of the land use optimization boundaries. Then, loading the full amount of industry-specific data covering the analysis area through the software's database connection module. This data must include vector datasets of management boundaries for core industries such as forestry, agriculture, water conservancy, natural resources, and ecological environment, specifically covering forestry protected area boundaries, basic farmland protection red lines, river and lake management scope lines, ecological protection red lines, and urban development boundaries. Introducing a multi-source data intelligent fusion algorithm, a unified spatial benchmark transformation is performed on the loaded industry-specific data. The algorithm automatically identifies the original coordinate system of various industry data, corrects the spatial benchmark deviation of the industry data through a coordinate registration algorithm, and adjusts the coordinate system and projection parameters of all industry data to be completely consistent with the land use optimization boundaries. Finally, calling the software's feature filtering tool, the thematic data is classified and extracted according to industry type. For forestry sector data, forestry management boundary elements such as national-level public welfare forest boundaries, nature reserve boundaries, and forest land protection and utilization planning boundaries were selected; for agricultural sector data, agricultural management boundary elements such as basic farmland protection red lines, permanent basic farmland reserve area boundaries, and arable land protection boundaries were selected; for water conservancy sector data, water conservancy management boundary elements such as river and lake management scope lines, dike engineering management boundaries, and water source protection boundaries were selected; for natural resource sector data, natural resource management boundary elements such as ecological protection red lines, urban development boundaries, and mineral resource prohibition mining area boundaries were selected; for ecological environment sector data, ecological environment management boundary elements such as environmental control unit boundaries and key ecological function zone boundaries were selected. Preprocessing is performed on the extracted industry management boundary data: First, a topology check tool is run to identify and repair topology errors such as overlapping, gaps, hanging nodes, and broken lines in boundary features. For overlapping boundaries, the boundary features corresponding to the latest version of the land rights confirmation data are retained. Gaps smaller than 0.001 hectares are automatically merged into adjacent main boundaries. Hanging nodes and broken lines are repaired using node capture and line connection tools. Second, attribute verification is performed to check the core attribute fields such as industry code, boundary name, delineation basis, and validity period for each type of boundary. Issues such as misaligned codes and incorrect name labeling are corrected. Industry type and boundary level fields are added and automatically assigned values. Finally, the preprocessed industry management boundary data is saved as standardized vector layers according to industry type. Each layer is bound with complete attribute information, forming a management boundary dataset covering the entire industry.

[0038] Step 4.2 involves spatial overlay analysis of the management boundaries and land use optimization boundaries to identify overlapping and conflicting areas. Specifically, this includes loading the entire industry's management boundary dataset into the same geographic information system (GIS) and using spatial reference matching tools to reconfirm that all data have identical coordinate systems and projection parameters, with no spatial offset or coordinate deviation issues. The software's multi-element overlay analysis tool is then activated, selecting either line-to-area or line-to-line mixed overlay mode. Using the land use optimization boundary as the baseline, pairwise overlay analysis is performed on various industry management boundaries as overlay layers. For areal industry management boundaries, such as basic farmland protection red lines and ecological protection red lines, the software automatically identifies the spatial relationship between the land use optimization boundary and the areal boundary, including three cases: the boundary is completely within the areal boundary, the boundary partially crosses the areal boundary, and the boundary is completely outside the areal boundary. For linear industry management boundaries, such as river and lake management boundaries and dike engineering management boundaries, the software automatically calculates the spatial distance, angle, and intersection coordinates between the land use optimization boundary and the linear boundary, identifying overlapping segments, parallel segments, intersecting segments, and deviating segments. The system initiates a conflict detection algorithm, using spatial distance thresholds (e.g., a distance of less than 5 meters between the boundary line and the industry boundary is considered a conflict) and industry control rules (e.g., the boundary of the core protected area cannot be breached). It accurately identifies spatial conflict areas and labels them according to the severity of the conflict. After overlay analysis, the analysis result layer uses different colors to distinguish the different spatial relationships between the boundary line and the industry management boundary: overlapping areas are marked in green, and spatial conflict areas are marked in red. Based on this result, the system clarifies overlapping and spatial conflict areas: overlapping areas are sections where the land use optimization boundary line completely coincides with the industry management boundary line or is within the boundary compliance range; spatial conflict areas are further subdivided into three categories: first, sections where the boundary line crosses the industry's prohibited boundary; second, sections where the boundary line deviates from the industry control boundary by a distance exceeding the compliance threshold, such as sections deviating more than 10 meters from the river and lake management area line; and third, sections where the boundary line intersects with the industry's confirmed rights boundary but does not extend according to the boundary direction. All conflict areas are labeled with specific conflict types, involved industry types, precise coordinates of the conflict location, and conflict severity, forming a complete spatial conflict area dataset.

[0039] Step 4.3: Based on the spatial conflict areas and adhering to the industry data priority principle, automatically adjust the direction of the land use optimization boundary in the conflict sections according to the corresponding industry management boundary to obtain the adjustment result. Specifically, based on the generated spatial conflict area dataset, firstly, formulate quantitative execution rules for the industry data priority principle, and then launch the intelligent adjustment algorithm with these rules as the core. The algorithm execution process is as follows: First, extract the boundary data of all spatial conflict sections, including the coordinate sequence, length, and industry management boundary type involved in the conflict sections; at the same time, extract the complete coordinate sequence, boundary attributes, compliance control requirements, and other data of the corresponding industry management boundary; Second, perform spatial analysis on each conflict boundary segment. For segments that cross prohibited boundaries, calculate the outer boundary coordinates of the industry management boundary, and generate a new boundary path based on these coordinates to ensure the adjusted boundary. The first step involves ensuring the boundary is completely outside the prohibited boundary and meets distance requirements. For sections deviating from the control boundary, the direction vector of the industry management boundary is calculated, and the boundary inflection points are adjusted through coordinate translation and rotation to make the boundary parallel to the industry management boundary, with the deviation distance controlled within the compliance threshold. For sections intersecting with the established rights boundary, the inflection point coordinates of the industry rights boundary are extracted, and the inflection points of the conflict section's boundary are replaced with the corresponding inflection points of the established rights boundary, extending the boundary along the established rights boundary. The second step involves setting adjustment constraints to ensure that the adjusted boundary has no self-intersections, no topological errors, and a smooth connection with the boundary of non-conflict sections, with the adjustment range not exceeding 15% of the original boundary direction, and without reducing the core ecological protection area. The third step involves the algorithm automatically generating the adjusted coordinate sequence of conflict sections, recording parameters such as the original coordinates, adjusted coordinates, adjusted distance, adjusted direction, and the industry rules applied to each boundary segment. After all conflicting sections have been adjusted, an adjustment result dataset is generated, which includes the adjusted boundary segment vector data, the adjustment result attribute table, and the boundary comparison layer before and after the adjustment. The adjustment results are initially verified to confirm that all conflicting sections have been adjusted according to the industry data priority principle and that the adjusted boundaries meet the control requirements of the corresponding industry. After the verification is passed, the final adjustment result is formed.

[0040] Step 4.4: Based on the adjustment results, integrate the optimized land use boundaries and management boundaries to generate detailed boundaries for internal processing. This includes: First, based on the adjustment results, integrate the boundary vector data: load the adjusted boundary data of conflict zones and the optimized land use boundary data retained in non-conflict zones onto the same layer, call the software's feature merging tool, and stitch all segments sequentially according to the spatial connection relationship of the boundary segments; during the stitching process, enable the high-precision node capture function, with a capture tolerance of 0.1 meters, to ensure that the node coordinates of adjacent segments are completely coincident, without breakpoints or overlaps, forming complete preliminary integrated boundary vector data. The initially merged boundaries underwent reconstruction and optimization: First, node simplification was performed using the Douglas-Puk algorithm with a simplification tolerance of 2 meters, eliminating redundant nodes caused by adjustments and retaining key nodes connecting to the industry management boundary. Second, boundary smoothing was performed using a third-order Bézier curve smoothing algorithm with a control point spacing of 5 meters and a curvature radius ≥ 3 meters, optimizing the inflection points of the boundary lines to ensure a natural and continuous merging of the boundaries while maintaining precise alignment with the industry management boundary. Finally, a topology consistency check was conducted to identify and repair issues such as self-intersections, dangling nodes, and topological gaps. Topological gaps with an area > 0.001 hectares were marked for manual verification to ensure the topological integrity of the boundary vector data. After reconstruction, the attribute information of the land use optimization boundary and the industry management boundary was merged. Simultaneously, the core attributes of the industry management boundary, such as the boundary demarcation basis, validity period, and control requirements, were associated with the merged boundary segments, forming a complete dataset of boundary spatial data and industry attribute data. The final compliance verification of the merged boundaries involves three steps: first, industry rule verification, checking each segment to ensure it complies with the corresponding industry's control requirements and that there are no instances of crossing prohibited boundaries or deviating from controlled boundaries; second, spatial scope verification, comparing the boundary enclosure before and after merging to confirm that core ecological land and compliant protection areas have not been reduced; and third, format standardization verification, checking that the file format, coordinate system, and attribute fields of the boundary vector data are consistent with the data from previous steps. Once verification is passed, a detailed boundary layer is officially generated for internal use, and a complete process record is saved.

[0041] In this embodiment of the invention, a series of technical means are employed, including automatic extraction and benchmark unification of industry boundaries driven by multi-source data intelligent fusion algorithms, high-precision conflict detection and hierarchical labeling under hybrid overlay mode, and intelligent adjustment algorithms under industry data priority rules. These systematically overcome the technical bottlenecks in traditional ecological zoning, which are caused by the diversity of industry data sources, inconsistent spatial benchmarks, and complex management rules, resulting in difficulties in data integration, reliance on manual conflict identification, low efficiency in adjustment and coordination, and susceptibility to subjective bias. This makes it difficult to simultaneously meet the compliance requirements of multiple industries. As a result, the technical effect of seamless spatial connection, strict adherence to rules, and deep integration of attributes between the boundaries of ecological protection zones and the full-caliber industry management boundaries is achieved. This enables the boundary adjustment process to be efficient, objective, and traceable, and significantly improves the industry compliance, management synergy, and law enforcement operability of the results.

[0042] In a preferred embodiment of the present invention, step 5 above may include: Step 5.1: Receive the detailed boundary lines from the office, convert them to a format specific to field reconnaissance, and generate and synchronize the field reconnaissance thematic data package to the mobile GIS device. Specifically, this includes: First, loading the detailed boundary line layer into the project in the GIS software. Then, comprehensively verifying the relevant parameters of the layer using the layer attribute check function to confirm that the spatial reference of the layer is the CGCS 2000 coordinate system, Gauss-Kruger projection, coordinate precision to 6 decimal places, good topological integrity, and no issues such as self-intersections, dangling nodes, or topological gaps. The attribute fields are complete and error-free, thus completing the reception and confirmation of the detailed boundary lines. Next, start the software's format conversion tool to convert the vector format of the detailed boundary lines to a format specific to field reconnaissance. Considering the compatibility requirements of the mobile GIS device, the converted format is determined to be a dual format: shp and mbtile.

[0043] The shp format is used for accurate display and editing of boundary vector data, while the mbtile format is used for rapid loading and offline viewing of boundaries on mobile GIS devices. During the conversion process, all attribute information of the boundaries is strictly preserved, including unique boundary segment IDs, industry fusion identifiers, fusion basis, adjustment parameters, and relevant industry attributes, ensuring that no attribute information is lost or misaligned after format conversion. After format conversion, the thematic data package for field reconnaissance is generated. This spatially correlates the boundary data with high-resolution remote sensing imagery (1-meter resolution) and digital elevation models, generating tile data through a data tiling algorithm. It marks the land cover types, potential disputed areas, and previously adjusted conflict sections within a 500-meter radius of the boundary. The data package also includes recommended routes generated by the field reconnaissance route planning algorithm, field reconnaissance record templates, reconnaissance point collection specifications, and relevant industry control requirements. After the thematic data package is generated, it is compressed using the zip compression format. Subsequently, the compressed thematic data package for field reconnaissance is synchronized to all mobile GIS devices used for field reconnaissance via wired or wireless LAN connections. During the synchronization process, a data integrity verification algorithm is used to verify the integrity of the data. After synchronization is completed, the dedicated reconnaissance software is launched on each mobile GIS device, the synchronized thematic data package is loaded, and all files in the data package are checked one by one to ensure that they are complete, that the boundary display is normal, and that the route planning is reasonable, so as to ensure that the thematic data package in the mobile GIS device can be directly used for field reconnaissance work.

[0044] Step 5.2: Using the field survey thematic data package synchronized to the mobile GIS device, field personnel conduct on-site surveys, recording the coordinates and site attribute information of the survey points to form field verification result data. This result data is then transmitted back to the cloud processor. Specifically, after arriving at the survey site with the mobile GIS device synchronized with the field survey thematic data package, field personnel first launch the dedicated survey software on the device, log in to their personal operating account, confirm that the software has successfully loaded the thematic data package, and verify the display effect of data such as detailed boundaries, basic geographic base maps, and relevant industry management boundaries to ensure there are no data gaps, display misalignments, or lag issues. Subsequently, field personnel conduct on-site surveys according to the preset survey route, utilizing the positioning function of the mobile GIS device. The survey route prioritizes covering identified spatial conflict areas, boundary adjustment areas, and key connection points of various industry management boundaries, while also considering all segments of the boundaries to ensure no omissions or blind spots in the survey. After the survey is completed, the mobile GIS device is connected to the office network, and the data is transmitted back to the cloud processor through a batch transmission algorithm to form the field verification result data.

[0045] Step 5.3: Based on the returned field verification results data, identify disputed areas requiring further verification and inaccessible areas. Specifically, after receiving the field verification results data returned by field personnel, firstly, invalid data is removed using a data integrity verification algorithm, such as coordinates exceeding the research scope, empty attribute fields, or photos that cannot be opened, generating a list of valid data; then, a boundary uncertainty modeling algorithm is introduced, based on a multi-source data error propagation model, to quantify the error of the field verification data: the root mean square error is used to calculate coordinate errors, the confusion matrix is ​​used to calculate attribute labeling errors, the land cover types retrieved from remote sensing images are compared with the on-site labeling types, and combined with the disputed descriptions provided by field personnel, automatically identifying disputed areas requiring further verification and classifying them according to the degree of uncertainty.

[0046] For example, high: error > 3 meters or attribute labeling inconsistency rate > 30%; medium: 1-3 meter error or 10%-30% inconsistency rate; low: < 1 meter error or < 10% inconsistency rate. Simultaneously, digital elevation model data is loaded, combined with access restriction information marked by field personnel, such as mountains, swamps, and military management areas. A terrain accessibility analysis algorithm is used: a cost-distance weighted algorithm is employed to calculate the access cost for field personnel to reach each area. The greater the slope and the higher the terrain undulation, the higher the cost. A access cost threshold is set; areas exceeding the threshold are considered inaccessible. Areas inaccessible by human intervention are identified, such as steep mountains (slope > 45°), swamps and wetlands (terrain undulation < 5 meters and NDVI < 0.2), and cliffs (slope > 60°). Vector boundaries are generated for disputed and inaccessible areas, with attribute information labeled as follows: disputed areas: uncertainty level, dispute type, and involved features; inaccessible areas: reason for inaccessibility, terrain parameters, forming a regional identification result dataset.

[0047] Step 5.4: Based on the identified disputed and inaccessible areas, plan and control the UAV for aerial photography to acquire high-resolution oblique image data of the corresponding areas and generate a real-world 3D model. Specifically, this includes: based on the area identification result dataset, conduct precise aerial photography design through aerial photography path planning algorithms: adopting a zonal planning strategy, firstly, use spatial clustering algorithms to integrate scattered target areas into several independent aerial photography units according to distance thresholds, such as the distance between adjacent areas being <500 meters. For each aerial photography unit, call the terrain complexity assessment sub-algorithm, and automatically match the flight path type by extracting slope and terrain undulation indicators from the regional digital elevation model data. Parallel flight paths are used in gently sloping areas, and circular flight paths are used in complex terrain areas. With the center point of the area as the center, set 3-5 concentric circle flight paths to cover different slope surfaces. The algorithm has a built-in parameter calculation model. Based on the core requirement of 0.1-meter ground resolution, combined with the drone camera focal length (e.g., 24mm) and sensor size (e.g., 1 / 2.3 inch), it automatically calculates the flight altitude using the formula: Flight Altitude = (Focal Length × Ground Resolution) / Sensor Size. It needs to be corrected by overlaying the actual terrain elevation. Simultaneously, the flight speed is set to 5-8 meters per second, the forward overlap is 80%, and the lateral overlap is 60%. Finally, it generates a detailed aerial photography parameter table containing the aerial photography unit number, flight path coordinate sequence, and parameters of each flight segment. The aerial photography plan is imported into the UAV ground control station, and equipment preparation is completed: a fully charged battery is installed, and the power supply is confirmed to be stable through a voltage detection algorithm. The GPS+BeiDou dual-mode positioning calibration algorithm is activated to complete signal calibration, with a positioning deviation of ≤0.5 meters. The ISO, shutter speed (≥1 / 1000 second), and white balance are set through a camera parameter debugging algorithm. When weather conditions meet the preset thresholds (wind force <4, no rain, visibility >5 km), the UAV's automatic aerial photography mode is activated. The ground control station receives the UAV's position, altitude, speed, battery level, and other status data through a real-time data transmission algorithm. At the same time, the image quality real-time evaluation algorithm monitors the shooting effect based on three-dimensional indicators of sharpness, exposure, and color saturation. If image blur or exposure abnormality is detected, a remote control algorithm is automatically triggered to adjust the flight speed by 20%. If the battery level is below 30%, an emergency return-to-home algorithm is activated. After aerial photography is completed, the drone is retrieved and high-resolution oblique image data is exported. The data quality is optimized through image preprocessing algorithms: first, a radiometric correction algorithm is executed to eliminate the influence of atmospheric scattering and sensor noise on image brightness; then, a geometric correction algorithm is used to correct image distortion using ground control point coordinates, with a planar deviation of ≤0.1 meters; finally, a distortion correction algorithm is run to remove image deformation caused by lens optical distortion, while filtering and removing blur and overexposure, retaining the effective image dataset.

[0048] The preprocessed effective image data is imported into the 3D reality and virtual reality-assisted decision-making system, and the core algorithm chain of 3D modeling is activated: Image correlations are constructed using the SFM (Structure of Motion Recovery) algorithm. First, the SIFT (Scale Invariant Feature Transform) algorithm is used to extract key feature points from each image, generating descriptors containing the location, scale, and orientation of these feature points. Then, a feature point matching algorithm, based on a Euclidean distance threshold of <0.6, finds corresponding feature points between different images, establishing spatial constraints between images and ultimately generating a sparse point cloud with a density ≥10 points / square meter. The Poisson reconstruction algorithm is then used for densification of the sparse point cloud: an octree data structure is first constructed to organize the sparse point cloud, and then an implicit function fitting algorithm is used to transform the point cloud information into a continuous 3D surface model, generating a dense point cloud. This ensures a point cloud resolution of at least 0.2 meters, accurately reproducing ground feature details. The Delaunay triangulation algorithm is employed, automatically connecting adjacent points based on the spatial coordinates of the dense point cloud according to the principle of maximizing the minimum angle, generating an irregular triangular 3D mesh model with a mesh side length ≤0.3 meters, ensuring good terrain fit. The texture mapping algorithm is activated. First, the tilted image and the 3D mesh model are precisely aligned using the image registration algorithm, with a texture mapping error of ≤1 pixel. Then, the texture fusion algorithm is used to eliminate the color difference between adjacent images and completely attach the high-resolution image texture to the surface of the mesh model to generate a visually realistic 3D model.

[0049] Step 5.5: Based on the real-scene 3D model, perform spatial overlay and fusion analysis with the refined boundary lines from the office, and finely adjust the boundary vectors to obtain the field verification boundary lines. Specifically, this includes: loading the real-scene 3D model and the refined boundary lines from the office into the geographic information system software, first starting the spatial alignment algorithm, and ensuring accurate data fusion through coordinate matching and feature point registration. Based on the coordinate matching sub-algorithm, the boundary line is forcibly aligned with the spatial reference of the 3D model to correct coordinate offset errors. The feature point registration sub-algorithm is used to extract feature points of land features in the 3D model, such as water system inflection points, road intersections, terrain high points and key inflection points of the boundary line. The registration parameters are calculated by the least squares method to achieve fine alignment between the two, ensuring that the boundary line can be accurately superimposed on the corresponding spatial position of the 3D model without misalignment or offset. A boundary spatial analysis algorithm based on a real-world 3D model is introduced to conduct immersive verification and accurately identify deviation areas: First, a distance measurement algorithm is used to calculate the shortest straight-line distance between each point on the boundary and the physical features in the model based on the spatial Euclidean distance formula, automatically marking potential conflict points with a distance of less than 5 meters; Second, a feature collision detection algorithm is run, using axis-aligned bounding box collision detection logic to determine the intersection of the spatial range of the boundary and the physical features in the model. If there is overlap, it is determined to be a crossing conflict, and the start and end coordinates of the conflict section and the types of features involved are recorded; Third, a terrain fit analysis algorithm is activated to extract the 3D model elevation value corresponding to the boundary inflection point, calculate the height difference between the inflection point elevation and the actual elevation of the terrain surface, and mark sections with a height difference greater than 1 meter as areas with insufficient fit.

[0050] Through joint analysis of three types of algorithms, a complete dataset of deviation areas was formed, including deviation type, coordinate range, and deviation degree. Combined with the data from field verification, fine-tuning was carried out on the deviation areas: For conflict areas where the boundary line crosses physical features, a path avoidance algorithm was used to construct a 5-meter buffer zone based on the feature boundary. The optimal avoidance path was planned outside the buffer zone using the A* path search algorithm, correcting the boundary line inflection point coordinates to ensure that the distance between the adjusted boundary line and the feature boundary is no less than 5 meters; For areas with insufficient terrain fit and an elevation difference greater than 1 meter, an inflection point elevation correction algorithm was run to replace the elevation values ​​of the boundary line inflection points with the actual terrain elevations at the corresponding locations in the 3D model. Simultaneously, linear interpolation was used to correct the elevation transition between adjacent inflection points, ensuring the boundary line perfectly fits the terrain surface; For areas inaccessible by manual intervention, multi-angle visualization using the 3D model was utilized. To leverage its advantages, the boundary alignment was adjusted using a spatial path optimization algorithm, with an auxiliary optimization algorithm chain employed during the adjustment process: First, the Douglas-Puk algorithm was used to simplify nodes, with a simplification tolerance of 1 meter. The specific process involved connecting the first and last nodes of the boundary to form a baseline straight line, calculating the perpendicular distance from all intermediate nodes to the baseline straight line, eliminating redundant nodes with a distance less than 1 meter, and repeating the above operation on the remaining nodes until no redundant nodes were found, ensuring that the boundary was simple and the core alignment remained unchanged. Second, a third-order Bézier curve smoothing algorithm was used for boundary fitting. Using the simplified nodes as anchor points, two control points were generated between adjacent anchor points, with a spacing of 3 meters. These control points were generated based on the perpendicular bisector offset of the anchor point connection line, and the third-order Bézier curve formula was used for smoothing. Where t∈[0,1], by the curvature formula ,in This represents the curvature value of the curve at a certain point. These are the parameter variables of the curve. ( ) represents the curve in the parameter The x-coordinate function at the location, ( ) represents the curve in the parameter The ordinate function at the location is used to control the radius of curvature to be ≥2 meters. Simultaneously, a boundary reliability assessment algorithm is initiated to quantify the reliability of the adjusted boundary. First, an evaluation index system is constructed, covering multi-source data error, adjustment range, and on-site matching degree. The analytic hierarchy process (AHP) is used to determine the weights of each index, and a judgment matrix is ​​built through pairwise comparisons. The weight vector is calculated and consistency is checked. Then, each index is quantitatively scored, and the final credibility score is calculated using a weighted summation formula, generating a credibility evaluation report containing detailed scores, error sources, and improvement suggestions. After each boundary adjustment is completed, the adjusted vector coordinate data is automatically saved, and complete adjustment parameters are recorded to ensure full traceability of the adjustment process. After all sections are adjusted, a topology consistency check algorithm is run to comprehensively identify issues such as self-intersections, loops, topological gaps, and non-overlapping nodes on the boundaries. For self-intersection areas, duplicate segments are deleted and intersection coordinates are recalculated. For topological gaps, nodes around the gaps are extracted to generate closed segments. For non-overlapping nodes, node capture is used to complete topology repair, boundary attribute information is updated, and finally, a three-layer final verification logic is initiated: The first layer is on-site matching verification, which compares the 3D model with field photos to check the matching degree between the boundary and actual features on site segment by segment; the second layer is compliance verification, which verifies whether the boundary complies with industry control requirements and the principle of ecological land priority, and whether it crosses prohibited areas; the third layer is data standardization verification, which checks whether the file format, spatial reference, coordinate accuracy, and attribute fields of the boundary vector data are consistent with previous results. After successful verification, the field verification boundary layer is officially generated.

[0051] In this embodiment of the invention, a series of technical means are employed, including optimal path (uncertainty modeling and terrain accessibility analysis), intelligent aerial photography planning and high-precision 3D modeling, and optimization of 3D spatial analysis fusion algorithms. By using dual-format data packets and optimal paths, precise field visualization navigation is achieved. This overcomes a series of technical bottlenecks in traditional field verification, such as the difficulty in interpretation caused by the disconnect between 2D maps and actual scenes, the inability to cover areas due to complex terrain and control restrictions, the strong subjectivity and insufficient accuracy of adjustments due to human experience, and the inaccurate matching of boundaries and terrain features due to the lack of fusion analysis of multi-source data in 3D space. As a result, the field verification process is intuitive and efficient, with full coverage and quantifiable evaluation of disputed and inaccessible areas, sub-meter-level precise positioning of boundaries in the real 3D environment and high consistency with terrain features, and the final results are objective, credible, and traceable throughout the entire process. This significantly improves the on-site consistency, scientificity, and feasibility of boundary delineation in complex terrain areas.

[0052] In a preferred embodiment of the present invention, step 6 above may include: Step 6.1: Based on the field verification boundary lines, generate an expert review data package and organize experts from multiple fields to conduct a joint review through an online review system. The boundary data is automatically revised based on the review comments to obtain the expert-approved boundary lines. Specifically, this includes: first, loading the field verification boundary layer into the project engineering of the geographic information system software, comprehensively verifying the completeness, accuracy, and standardization of the layer; then, starting the data package generation tool to create the expert review data package, which includes: standardized vector files of the field verification boundary lines in both shp and mbtile formats, attribute data tables, the entire boundary optimization process results, field verification result data, a realistic 3D model, a credibility evaluation report, and aerial photography and modeling related materials. Experts from multiple fields such as ecology, surveying and mapping, land resources, and remote sensing are organized to conduct a joint review through the online review system. The system has a built-in boundary credibility assessment algorithm that automatically displays the credibility level, credibility score, and error sources of each boundary segment to the experts, assisting them in focusing on high-uncertainty sections for key reviews. Invitations to review the boundary were sent to all invited experts via an online review system. The data package access password and an online review system operation guide were also sent to the experts, instructing them on how to log in, download the data package, and complete the review process. After logging into the online review system, experts downloaded the expert review data package and, based on their professional expertise, conducted a comprehensive review of the entire boundary optimization process and the field verification boundaries. They filled in their review comments in real time, marking agreed-upon sections, sections requiring modification, and modification suggestions. The system used an opinion aggregation algorithm to integrate all expert opinions by section, automatically marking sections with disagreements and displaying the different experts' viewpoints. Based on the expert opinions, an intelligent revision algorithm, combined with previous data fusion rules and adjustment logic, automatically revised the boundaries: for sections explicitly agreed upon by experts, the field verification boundaries were retained; for sections with specific modification suggestions from experts, parameters were adjusted according to the suggestions, and the corresponding algorithm was executed; for sections with disagreements among experts, the algorithm automatically generated 2-3 revision schemes, retaining a manual intervention interface, allowing the technical team to select the optimal scheme based on expert opinions, 3D models, and field data. After revision, the credibility of the revised boundaries is verified again using a boundary credibility assessment algorithm to ensure that the proportion of high-credibility segments is not less than 90%. A preliminary version of the expert-deliberated boundaries and a detailed revision report are then generated. Subsequently, the preliminary version of the expert-deliberated boundaries and the detailed revision report are uploaded to the online review system and fed back to all participating experts. Experts are invited to review the revised boundaries to confirm whether the revisions conform to the review comments and whether the original problems have been resolved. After the expert review is completed, if all experts approve the revision results, the review is marked as passed, and the expert-deliberated boundary layer is officially generated.

[0053] Step 6.2 involves submitting the expert-approved boundary lines to the competent authority's review system for compliance review, receiving review feedback, and adjusting the boundary line results based on the feedback to generate approved boundary lines. Specifically, this includes: after the expert-approved boundary lines are officially generated, conducting a comprehensive self-check of the boundary lines and review materials to confirm compliance with the competent authority's review requirements; organizing and submitting materials, including boundary line vector files, attribute tables, review results, boundary line optimization process materials, and compliance explanation reports; standardizing these materials according to the review system requirements before online submission; filling in the submission information; and retaining the receipt. This involves cooperating with the competent authority in conducting compliance reviews, responding promptly to inquiries, and supplementing materials. The competent authority's review system incorporates a preprocessing module with a dynamic boundary update and monitoring mechanism: extracting core boundary information through metadata extraction algorithms, converting review requirements into quantifiable indicators through compliance indicator quantification algorithms, and recording the boundary line's compliance indicator values; focusing on verifying boundary line planning compliance, industry control compliance, implementation of expert opinions, and data standardization; and providing feedback after review, categorized as either approved or returned for modification. After approval, the review receipt and report are retained, and the boundary marker deployment phase begins. For returned applications, feedback issues are analyzed, a modification plan is developed, and the boundary lines are adjusted specifically based on review comments, expert suggestions, and previous results. Attribute information is supplemented and improved, and any unresolved review comments are addressed. After modification, a self-check is conducted to confirm no new issues. Modified materials are then compiled, marked with a resubmission tag, and the status of issue resolution is explained. The application is resubmitted for review until approval is granted. A formally approved boundary line layer and related materials are generated, and the boundary data is linked to the dynamic monitoring interface.

[0054] Step 6.3: Based on the approved boundary lines, automatically plan the boundary marker layout scheme. The scheme generates boundary marker layout data, which specifically includes: loading the boundary lines approved by the competent authority, remote sensing images, and local real-world 3D models; using a multi-source data intelligent fusion algorithm to associate key boundary nodes, terrain conditions, land feature distribution, field verification points, and other data to construct a boundary marker layout decision dataset; the dataset includes the coordinates, terrain parameters, land feature interference degree, accessibility, and other attributes of each potential layout location. An automatic boundary marker placement planning algorithm is activated. Based on preset placement principles, key nodes are required, straight sections are placed at intervals, dangerous areas are avoided, and additional markers are added at industry boundary junctions. The algorithm automatically calculates the precise placement coordinates of each boundary marker: for boundary inflection points and industry boundary junctions, main boundary markers are placed, with the coordinates of the main boundary markers consistent with the coordinates of the boundary inflection points; for straight sections of the boundary, the placement density is set according to the terrain complexity, 500-1000 meters per marker in flat areas and 100-500 meters per marker in complex areas, and the coordinates of auxiliary boundary markers are calculated through an equidistant segmentation algorithm; at the junctions of the boundary with industry-prohibited boundaries such as ecological protection red lines and basic farmland protection red lines, special boundary markers are added, marked with special identification attributes. The algorithm combines 3D model analysis to determine terrain adaptability, eliminating potential locations with slopes greater than 15°, terrain undulations greater than 50 meters, or distances from water systems less than 5 meters. It selects locations that are flat, offer open views, are easily accessible, and easy to maintain. A boundary marker placement plan is generated, including an overview map, detailed boundary marker information, and placement instructions. Technical personnel and field supervisors review the plan, verifying the rationality of each boundary marker placement location using the 3D model, correcting any unreasonable aspects, and finalizing the official boundary marker placement plan.

[0055] Step 6.4 involves establishing boundary markers based on the boundary marker deployment data, spatially associating and attribute-linking the boundary lines and boundary markers, and constructing and storing this data to form a traceable and dynamically updated spatial database of ecological protection zone boundaries. This achieves both physical and information-based management of the boundaries. Specifically, this includes: generating a deployment task list based on the formal plan, dividing tasks among field operation teams, clarifying the deployment scope, number of boundary markers, completion deadline, and quality standards for each team, converting the data to a format recognizable by field terminals, supplementing with operation guidelines, and distributing it to each field terminal through a dedicated channel; after field personnel deploy the boundary markers according to the plan, they upload the deployment results data back to the cloud to form the official boundary marker deployment results. The boundary marker deployment results are pre-processed to verify the completeness of the boundary marker information and the accuracy of the coordinates, ensuring consistency with the approved boundary spatial reference. Any problems found are addressed and rectified until all requirements are met. Spatial association and attribute binding are performed using a multi-source data intelligent fusion algorithm: Spatial association employs a coordinate matching algorithm to bind each boundary marker to its corresponding boundary segment, clearly defining the segment to which the boundary marker belongs and its distance from the boundary inflection point; Attribute binding uses a field association algorithm to bidirectionally associate boundary marker layout information, boundary optimization information, field verification information, and industry data information, ensuring that each boundary segment and each boundary marker is bound with complete attribute records. After the association is completed, data integration and optimization are carried out: Topology consistency check algorithm is used to identify and repair issues such as association errors, boundary self-intersections, and boundary marker coordinate mismatches; attribute field standardization algorithm is used to unify field names, codes, and data formats, delete redundant fields, and supplement missing attributes; Douglas-Pock algorithm is used to simplify the boundary nodes with a tolerance of 1 meter. In accordance with the national geographic information database construction standards and ecological protection zone management requirements, the spatial database structure of the ecological protection zone boundary is designed, dividing it into a basic geographic information layer, a boundary core layer, a boundary marker information layer, a full-process record layer, and an industry association layer, and a database management system is used to build the storage environment.

[0056] Configure database hardware and software parameters, set access permissions, and batch import all integrated and optimized data into the database. The database has a built-in dynamic boundary update and monitoring module: it supports access to time-series remote sensing imagery and IoT sensor data, and monitors changes in the surrounding ecological environment and the status of boundary markers in real time. When changes in land features that may affect boundary compliance or boundary marker anomalies are detected, an early warning is automatically triggered and update suggestions are pushed. The database also includes boundary uncertainty and reliability attribute fields, supporting quantitative assessment results queried by section. A daily database maintenance system is established, with regular data backups, security risk checks, and regular database updates based on the latest data to generate maintenance reports. This achieves dual management of the physical and informational aspects of the boundary, providing data support for ecological supervision and law enforcement.

[0057] In this embodiment of the invention, a series of intelligent technologies are employed, including credibility assessment-driven online multi-expert collaborative review, compliance indicator quantitative review, three-dimensional assisted automatic boundary marker layout planning, marker erection, and the construction of a dynamically traceable spatial database. Built-in credibility algorithms assist experts in focusing their reviews, and opinion summarization and intelligent revision algorithms achieve efficient consensus building. Combined with compliance indicator quantification and dynamic monitoring preprocessing modules, the review is ensured to be objective and measurable. Multi-source data fusion and three-dimensional analysis are used to achieve automatic optimization and precise implementation of boundary marker layout. Marker erection further constructs a dynamically updated, fully traceable boundary spatial database. This overcomes the problems of inefficient and discrete expert reviews, vague review standards relying on manual experience, lack of scientific planning in boundary marker layout leading to implementation deviations, and static and isolated results management that makes continuous monitoring and updating difficult in traditional boundary verification. Thus, a closed-loop management effect is achieved, ensuring scientific and efficient boundary review, objective and credible review conclusions, precise and reasonable boundary marker layout, traceable results data, and dynamic maintenance throughout the entire process. This realizes intelligent collaboration and continuous optimization of ecological protection zone boundaries from demonstration, review, physicalization to information management.

[0058] like Figure 2 As shown, embodiments of the present invention also disclose a system for delineating the boundaries of complex terrain ecological zones using multi-source data fusion, comprising: The acquisition module is used to acquire the protection scope boundaries in the ecological protection plan as preliminary boundaries, collect multi-source heterogeneous data, and construct a dataset with unified benchmarks. The adjustment module is used to extract natural boundaries, economic boundaries, and geomorphic boundaries from the preliminary boundaries, perform overlay analysis, identify discrepancies, and perform manual interpretation and spatial conflict resolution to obtain preliminary verification boundaries. Based on the preliminary verification boundaries, land use patches from the Third National Land Survey are overlaid with them, and boundary adjustments are made to avoid or minimize the cutting of construction land patches to obtain optimized land use boundaries. Based on the optimized land use boundaries, management boundaries from industry-specific data are loaded, and spatial overlay and conflict detection are performed to obtain refined boundaries for internal processing. The verification module is used to import the detailed boundary lines from the office into mobile GIS devices for field surveys, and to use drone aerial photography to construct real-scene 3D models of disputed or inaccessible areas to obtain field verification boundaries. The processing module is used to organize experts from multiple fields to conduct joint reviews of the field-verified boundaries, obtain expert-approved boundaries, and further construct a traceable spatial database containing boundary lines, boundary markers, and attribute information through the establishment of markers, thereby realizing the physical and information-based management of the boundaries.

[0059] It should be noted that this system is a system corresponding to the above method. All implementation methods in the above method embodiments are applicable to this embodiment and can achieve the same technical effect.

[0060] Embodiments of the present invention also disclose a computing device, including: a processor and a memory storing a computer program, wherein the computer program, when executed by the processor, performs the method described above. All implementations in the above method embodiments are applicable to this embodiment and can achieve the same technical effects.

[0061] Embodiments of the present invention also disclose a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform the method described above. All implementations in the above method embodiments are applicable to this embodiment and can achieve the same technical effects.

[0062] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for complex terrain ecological region boundary delineation of multi-source data fusion, characterized in that, The method comprises: Step 1, obtaining the protection range boundary in the ecological protection planning as a preliminary boundary, collecting multi-source heterogeneous data, and constructing a benchmark unified data set; Step 2, according to the preliminary boundary, extracting natural boundary, economic boundary and topographic boundary for superposition analysis, identifying difference areas and performing manual interpretation and spatial conflict resolution to obtain a preliminary checked boundary; Step 3, based on the preliminary checked boundary, superimposing the land use polygons of the third national land survey data with the preliminary checked boundary, and adjusting the boundary according to the principle of avoiding or minimizing cutting construction land polygons to obtain a land class optimized boundary; Step 4, according to the land class optimized boundary, loading the management boundary in the industry thematic data, and performing spatial superposition and conflict detection to obtain an indoor refined boundary; Step 5, importing the indoor refined boundary into a mobile GIS device for outdoor field reconnaissance, and using a unmanned aerial vehicle aerial photography to construct a real scene three-dimensional model for the disputed or unreachable areas to obtain an outdoor verified boundary; Step 6, according to the outdoor verified boundary, organizing multi-field experts to jointly review the outdoor verified boundary to obtain an expert demonstration boundary, and further constructing a traceable spatial database containing the boundary, boundary post and attribute information through a standard to realize the entity and information management of the boundary.

2. The method according to claim 1, wherein, Obtain the protection range boundary in the ecological protection planning as a preliminary boundary, collect multi-source heterogeneous data, and construct a benchmark unified data set, including: Step 1.1, the cloud processor receives and loads the protection range vector data in the ecological protection planning, and determines it as a preliminary boundary; Step 1.2, collect multi-source heterogeneous data through the space range defined by the preliminary boundary; Step 1.3, according to the collected multi-source heterogeneous data, perform coordinate conversion and standardization format processing to obtain a benchmark unified data set.

3. The method according to claim 2, wherein, According to the preliminary boundary, extract natural boundary, economic boundary and topographic boundary for superposition analysis, identify difference areas and perform manual interpretation and spatial conflict resolution to obtain a preliminary checked boundary, including: Step 2.1, based on the preliminary boundary, extract the vegetation boundary from the remote sensing image, extract the water system and road boundary from the digital line graph, and perform spatial superposition analysis to obtain the difference area; Step 2.2, according to the difference area, assist in manual interpretation, and process through a spatial conflict resolution algorithm to generate a natural and economic boundary checked boundary; Step 2.3, based on the natural and economic boundary checked boundary, perform terrain factor analysis on the digital elevation model to identify steep slopes, valleys and ridges as key terrain units; Step 2.4, by constructing the boundary of the key terrain unit and superimposing it with the natural and economic boundary checked boundary, adjust the boundary direction to obtain a preliminary checked boundary.

4. The method according to claim 3, wherein, Based on the preliminary checked boundary, superimpose the land use polygons of the third national land survey data with the preliminary checked boundary, and adjust the boundary according to the principle of avoiding or minimizing cutting construction land polygons to obtain a land class optimized boundary, including: Step 3.1, receive the preliminary checked boundary, load the third national land survey data, and extract the land use polygon data containing the land class attribute; Step 3.2, perform spatial superposition analysis on the land use polygon data and the preliminary checked boundary to identify different land use type areas crossed by the preliminary checked boundary; Step 3.3: Based on the areas traversing different land types, and in accordance with the principle of prioritizing ecological land use, automatically perform a spatial adjustment algorithm on the boundary orientation of the traversing construction land use patches to obtain the spatial adjustment result; Step 3.4: Based on the spatial adjustment results, update and reconstruct the boundary vector data to generate optimized land use boundaries.

5. The method according to claim 4, wherein, Based on the optimized land use boundaries, the management boundaries from industry-specific data are loaded, and spatial overlay and conflict detection are performed to obtain the refined boundaries for internal processing, including: Step 4.1: Receive land category optimization boundaries, load industry-specific data, and extract various management boundaries. Step 4.2: Perform spatial overlay analysis on the management boundary and the land category optimization boundary to identify the overlapping areas and spatial conflict areas between them; Step 4.3: Based on the spatial conflict areas and the principle of prioritizing industry data, automatically adjust the direction of the land category optimization boundary in the conflict area according to the corresponding industry management boundary to obtain the adjustment result; Step 4.4: Based on the adjustment results, integrate the optimized land use boundaries and management boundaries to generate detailed internal boundaries.

6. The method according to claim 5, wherein, The detailed boundary lines from the office are imported into mobile GIS devices for field reconnaissance. For disputed or inaccessible areas, drone aerial photography is used to construct realistic 3D models, resulting in field-verified boundary lines, including: Step 5.1: Receive the detailed boundary lines from the office, convert them into a format specific to field reconnaissance, generate and synchronize the field reconnaissance thematic data package to the mobile GIS device; Step 5.2: By synchronizing the field survey thematic data package to the mobile GIS device, field personnel conduct on-site surveys, record the coordinates and on-site attribute information of the survey points, form field verification result data, and transmit the result data back to the cloud processor; Step 5.3: Based on the returned field verification results data, identify the disputed areas and inaccessible areas that require further verification; Step 5.4: Based on the identified disputed and inaccessible areas, plan and control the UAV to conduct aerial photography, acquire high-resolution oblique image data of the corresponding areas, and generate a real-world 3D model; Step 5.5: Based on the real-scene 3D model, perform spatial overlay and fusion analysis with the refined boundary lines in the office, and finely adjust the boundary vectors to obtain the field verification boundary lines.

7. The method according to claim 6, wherein, Based on the field verification of the boundary lines, a joint review was conducted by experts from multiple fields to obtain expert-approved boundary lines. Further, a traceable spatial database containing boundary lines, boundary markers, and attribute information was constructed through the establishment of markers, thereby achieving the physical and information-based management of the boundary lines, including: Step 6.1: Based on the field verification boundary, generate an expert review data package, organize experts from multiple fields to conduct a joint review through an online review system, and automatically revise the boundary data according to the review comments to obtain the expert demonstration boundary. Step 6.2: Submit the expert-demonstrated boundary line to the competent authority's review system for compliance review, receive review feedback, adjust the boundary line results based on the feedback, and generate the approved boundary line; Step 6.3: Based on the approved boundary lines, automatically plan the boundary marker layout scheme, and generate boundary marker layout data after the scheme is approved; Step 6.4: Based on the boundary marker deployment data, establish markers, spatially associate and attribute-link the boundary lines and boundary markers, construct and store them in the database to form a traceable and dynamically updated spatial database of ecological protection zone boundaries, and realize the physical and information-based management of the boundaries.

8. A multi-source data fusion complex terrain eco-region boundary delineation system, the system implements the method of any one of claims 1 to 7, characterized in that, include: The acquisition module is used to acquire the protection scope boundaries in the ecological protection plan as preliminary boundaries, collect multi-source heterogeneous data, and construct a dataset with unified benchmarks. The adjustment module is used to extract natural boundaries, economic boundaries, and geomorphic boundaries from the preliminary boundary line, perform overlay analysis, identify areas of difference, and perform manual interpretation and spatial conflict resolution to obtain the preliminary verification boundary line. Based on the preliminary verification boundary, the land use map patches from the third national land survey data are overlaid with the boundary, and the boundary is adjusted according to the principle of avoiding or minimizing the cutting of construction land map patches to obtain the optimized land use boundary. Based on the optimized land use boundary, the management boundary in the industry thematic data is loaded, and spatial overlay and conflict detection are performed to obtain the refined boundary for internal processing. The verification module is used to import the detailed boundary lines from the office into mobile GIS devices for field surveys, and to use drone aerial photography to construct real-scene 3D models of disputed or inaccessible areas to obtain field verification boundaries. The processing module is used to organize experts from multiple fields to conduct joint reviews of the field-verified boundaries, obtain expert-approved boundaries, and further construct a traceable spatial database containing boundary lines, boundary markers, and attribute information through the establishment of markers, thereby realizing the physical and information-based management of the boundaries.

9. A computing device, comprising: include: One or more processors; A storage device for storing one or more programs, which, when executed by one or more processors, cause the one or more processors to implement the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a program that, when executed by a processor, implements the method as described in any one of claims 1 to 7.

Images