Land space land use refinement method, device, equipment and storage medium
By combining multi-source data fusion and multi-threshold judgment mechanisms with spatial neighborhood smoothing algorithms, the problems of incomplete data and coarse identification results in land use identification have been solved, achieving high-precision and spatially continuous land use identification.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- 杭州市规划和自然资源调查监测中心(杭州市地理信息中心)
- Filing Date
- 2026-04-17
- Publication Date
- 2026-07-21
Smart Images
Figure CN122065090B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to a method, apparatus, equipment, and storage medium for refining land use in national land space, belonging to the fields of geographic information systems and national land space planning technology. Background Technology
[0002] With the deepening of urbanization, the refined governance and planning of national land space places higher demands on the accuracy and timeliness of land use function information. Accurately identifying urban land use is the foundation for optimizing spatial layout, implementing land use control, and assessing urban operational efficiency.
[0003] Traditional land use identification mainly relies on manual field surveys and map interpretation, which suffers from problems such as long cycles, high costs, and slow updates. In recent years, with the development of geographic information big data, automatic identification methods based on single or limited data sources have emerged. For example, patent application number 2018108982603, entitled "A Method for Identifying and Detecting Changes in Urban Land Use Functions," attempts to integrate remote sensing imagery and point of interest (POI) data, identifying land cover through remote sensing and combining it with POI statistical inference functions. However, these methods still have significant limitations: (1) Incomplete data representation: Remote sensing images are good at identifying physical cover (such as buildings and green spaces), but it is difficult to distinguish their socio-economic functions (such as commercial buildings and office buildings). Although POI data can reflect specific facility points, it is point data and cannot fully represent the scope of area functional zones. Moreover, there is insufficient evidence in POI sparse areas such as parks and newly built areas.
[0004] (2) Simple evidence fusion: Existing methods mostly adopt "hard overlay" or "simple voting" fusion, which fail to fully consider the differences in spatial form (point vs. surface), authority (legal data vs. Internet data), coverage and integrity of data from different sources, and lack quantitative evidence credibility assessment and dynamic balancing mechanism.
[0005] (3) The identification results are rough: the ability to identify mixed-use land (such as mixed commercial and residential land) that is common in the block is weak. At the same time, the classification results are easily affected by data noise, and often show "islands" or "jumps" in space that do not conform to the real laws, resulting in poor spatial continuity.
[0006] Therefore, there is an urgent need for an automated identification method that can deeply integrate multi-source heterogeneous spatial data, effectively handle evidence conflicts and sparsity issues, accurately identify single and mixed land use, and ensure the spatial rationality of the results. Summary of the Invention
[0007] To address the aforementioned issues, this invention proposes a method, apparatus, equipment, and storage medium for refining land use in national land space, which can solve the problems of low land use identification accuracy, difficulty in identifying mixed functions, and poor spatial continuity in the prior art.
[0008] The technical solution adopted by this invention to solve its technical problem is as follows: In a first aspect, an embodiment of the present invention provides a method for refining the use of national land space, comprising the following steps: Step S1: Obtain multi-source spatial data, which includes one or more of the following business data: Point of Interest (POI) data, Area of Interest (AOI) data, land survey results data, cadastral survey data, land ownership confirmation data, land use control approval data, real estate registration data, and study area boundary data; Step S2: Reclassify the original POI data and establish a mapping relationship between the original POI categories and the classification standards for land and sea use in national land space. Step S3: Perform coordinate system transformation and standardization preprocessing on the multi-source spatial data; Step S4: Generate basic units for land and space planning based on the business data; Step S5: Integrate the AOI area evidence and the reclassified POI point evidence to calculate the comprehensive evidence strength of different land use types in each basic planning unit; Step S6: Determine the land use type of each planning basic unit by combining a multi-threshold determination mechanism with a spatial neighborhood smoothing algorithm; the multi-threshold determination mechanism includes setting a first threshold for determining AOI mandatory coverage, a second threshold for determining the dominant function, and a third threshold for determining the degree of functional mixing; the spatial neighborhood smoothing algorithm iteratively optimizes the functional category probability vector based on the spatial adjacency relationship of each planning basic unit. Step S7: Output the land use identification results and generate a thematic map of land use in national spatial planning.
[0009] The multi-threshold determination mechanism achieves hierarchical decision-making by setting different thresholds. In particular, the third threshold τ gap Specifically designed for quantifying functional mixing, this method effectively identifies mixed-use land. The spatial neighborhood smoothing algorithm iteratively optimizes the spatial adjacency relationships of basic planning units, significantly improving the spatial continuity and rationality of the identification results. The basic planning unit referred to in this invention is the smallest spatial analysis unit for identifying land use in national land space. In this invention, it can be an independent polygon, a grid unit of a street block, or other suitable spatial unit formed based on national land survey data.
[0010] Secondly, an embodiment of the present invention provides an apparatus for refining the use of national land space, comprising: The data acquisition module is used to acquire multi-source spatial data, which includes one or more of the following business data: point of interest (POI) data, area of interest (AOI) data, land survey results data, cadastral survey data, land ownership confirmation data, land use control approval data, real estate registration data, and study area boundary data; The data reclassification module is used to reclassify the original POI data and establish a mapping relationship between the original POI categories and the classification standards for land and sea use in national land space. The data preprocessing module is used to perform coordinate system transformation and standardization preprocessing on the multi-source spatial data; The basic planning unit generation module is used to generate basic land spatial planning units based on the business data. The evidence fusion module is used to fuse AOI area evidence and reclassified POI point evidence to calculate the comprehensive evidence strength of different land use types in each basic planning unit. The land use determination module is used to determine the land use type of each basic planning unit through a multi-threshold determination mechanism combined with a spatial neighborhood smoothing algorithm. The multi-threshold determination mechanism includes setting a first threshold for determining mandatory AOI coverage, a second threshold for determining the dominant function, and a third threshold for determining the degree of functional mixing. The spatial neighborhood smoothing algorithm iteratively optimizes the functional category probability vector of each basic planning unit based on the spatial adjacency relationship of each unit. The results output module is used to output the land use identification results and generate a thematic map of land use in national spatial planning.
[0011] Thirdly, an electronic device provided by an embodiment of the present invention includes a processor, a memory, and a bus. The memory stores machine-readable instructions executable by the processor. When the electronic device is running, the processor communicates with the memory via the bus, and the processor executes the machine-readable instructions to perform the steps of the method for refining the use of any national land space as described above.
[0012] Fourthly, embodiments of the present invention provide a storage medium storing a computer program, which, when run by a processor, executes the steps of the method for refining any land use in the above-described land use classification.
[0013] One of the above technical solutions has the following advantages or beneficial effects: 1. By establishing a multi-level classification mapping system and keyword mapping mechanism, Internet POI data is accurately converted into land use classification standards, thus solving the problem of inconsistent data standards.
[0014] 2. By weighted fusion of AOI (area-like) evidence and POI (point-like) evidence, and taking into account the spatial morphology and authority differences of different data sources, a three-dimensional evidence system was constructed, which improved the reliability of the evidence.
[0015] 3. By setting including τ aoi τ major The multi-threshold decision-making mechanism of τgap enables hierarchical decision-making, especially the third threshold τ. gap Specifically designed for quantifying the degree of functional mixing, it effectively identifies mixed-use land and solves the problem of weak ability of traditional methods to identify mixed functions.
[0016] 4. By introducing an iterative spatial smoothing algorithm based on spatial adjacency, the preliminary judgment results are optimized, effectively suppressing the "island" misjudgment caused by data noise, and significantly improving the spatial continuity and rationality of the recognition results.
[0017] 5. Through the synergy of the above-mentioned technical means, the method of the present invention has significant improvements in terms of logical consistency, spatial rationality, and mixed land use identification capability compared with traditional single data source or simple fusion methods. Attached Figure Description
[0018] Figure 1 This is a flowchart illustrating a method for refining land use in national land space according to an exemplary embodiment; Figure 2 This is a schematic diagram of the structure of an apparatus for refining the use of national land space, according to an exemplary embodiment; Figure 3 This is a flowchart illustrating a specific implementation of a method for identifying land use in national spatial planning using the present invention, according to an exemplary embodiment. Figure 4 This is a schematic diagram illustrating the land use identification result of a national spatial planning system according to an exemplary embodiment of the present invention. Detailed Implementation
[0019] To more clearly illustrate the technical features of the present invention, the present invention will be described in detail below through specific embodiments and in conjunction with the accompanying drawings.
[0020] like Figure 1 As shown in the figure, an embodiment of the present invention provides a method for refining the land use of national land space, which includes the following steps: Step S1: Obtain multi-source spatial data, which includes one or more of the following business data: Point of Interest (POI) data, Area of Interest (AOI) data, land survey results data, cadastral survey data, land ownership confirmation data, land use control approval data, real estate registration data, and study area boundary data; Step S2: Reclassify the original POI data and establish a mapping relationship between the original POI categories and the classification standards for land and sea use in national land space. Step S3: Perform coordinate system transformation and standardization preprocessing on the multi-source spatial data; Step S4: Generate basic units for territorial spatial planning based on the business data; Step S5: Integrate the AOI area evidence and the reclassified POI point evidence to calculate the comprehensive evidence strength of different land use types in each basic planning unit; Step S6: Determine the land use type of each planning basic unit by combining a multi-threshold determination mechanism with a spatial neighborhood smoothing algorithm; the multi-threshold determination mechanism includes setting a first threshold for determining AOI mandatory coverage, a second threshold for determining the dominant function, and a third threshold for determining the degree of functional mixing; the spatial neighborhood smoothing algorithm iteratively optimizes the functional category probability vector based on the spatial adjacency relationship of each planning basic unit. Step S7: Output the land use identification results and generate a thematic map of land use in national spatial planning.
[0021] As one possible implementation of this embodiment, step S1 includes the following sub-steps: Step S11: Obtain Points of Interest (POI) data from a commercial geospatial platform via an API interface. The commercial geospatial platform includes a first platform and / or a second platform. The POI data of the first platform uses the GCJ-02 coordinate system, and the POI data of the second platform uses the BD-09 coordinate system. Step S12: Obtain Area of Interest (AOI) data from the Open Geospatial Platform via API interface or data file. The AOI data includes areal features obtained from the Open Street Map and contains land use type attributes. Step S13: Obtain land survey data from the land and natural resources authorities, and integrate cadastral survey, land ownership confirmation, land use control approval, and real estate registration data to form land space vector surface data; Step S14: Obtain the boundary vector surface data of the study area to define the scope of the study area; Step S15: Perform format unification and preliminary quality checks on the acquired multi-source data.
[0022] In step S11, POI data within the study area are acquired in batches through the API interface. The acquired POI data includes at least the following attributes: location name, latitude and longitude coordinates, address information, category label, and business status.
[0023] In step S12, the acquired AOI data includes at least the following element types: commercial area, residential area, industrial area, school, hospital, park, sports venue, and transportation hub.
[0024] In step S13, the land space vector surface data includes the following attribute fields: land parcel number, land use type code, land use type name, ownership nature, right of use nature, approval document number, registration time, and area.
[0025] In step S15, the preliminary quality check of the acquired multi-source data includes the following: checking data integrity, checking coordinate validity, checking attribute field integrity, checking logical consistency, and identifying and marking abnormal data.
[0026] As one possible implementation of this embodiment, the land and sea use classification standard for national land space adopts the land and sea use classification standard for national land space survey, planning, and land use control in the "Guidelines for Land and Sea Use Classification in National Land Space Survey, Planning, and Land Use Control". Step S2 includes the following sub-steps: Step S21: Establish a multi-level classification mapping system between the original POI category and the land and sea use classification standards of land and sea use survey, planning, and land use control. The mapping system includes three levels: functional classification FUNC, sub-functional classification SUBFUNC, and category-to-function mapping CATEGORY_TO_FUNC. Step S22: Construct a keyword mapping mechanism CATEGORY_KEYWORDS_TO_FUNC to achieve intelligent conversion from the original POI category to the land use classification; Step S23: Establish an area weight database, including the average area database of different subclass POIs SUBCATEGORY_AVG_AREA_M2 and the backup average area database of the first-level category L1_FALLBACK_AVG_AREA_M2. Step S24: Perform unified reclassification processing on POI data from different sources and output standardized land use classification codes.
[0027] In step S21, the multi-level classification mapping system maps the original POI category to the first-level, second-level, or third-level category in the "Guidelines for Classification of Land and Sea Use in Territorial Spatial Survey, Planning, and Land Use Control".
[0028] In step S22, the keyword mapping mechanism extracts keywords from the POI name using natural language processing technology and matches the keywords with the functional categories in the land use classification standard, including the following processing: Chinese word segmentation, stop word filtering, synonym expansion, and industry terminology recognition.
[0029] In step S23, the area weight database is constructed in the following way: collecting sample data of various POIs, measuring or estimating their actual occupied area, calculating the average occupied area of POIs of the same type, and using the average occupied area of the next higher level category as a backup value for categories with insufficient sample size.
[0030] In step S24, the unified reclassification process for POI data includes the following rules: when the POI data itself contains category labels, the category mapping table is used for conversion first; when the category labels are missing or unreliable, the keyword mapping mechanism is used for conversion; when neither method can determine the category label, it is marked as pending manual verification.
[0031] As one possible implementation of this embodiment, step S3 includes the following sub-steps: Step S31: The coordinate system from different data sources is converted to the WGS84 standard coordinate system through a coordinate transformation algorithm. Specifically, the GCJ-02 coordinate system, BD-09 coordinate system, and CGCS2000 coordinate system are converted to the WGS84 coordinate system. Step S32: When performing spatial calculations, the WGS84 coordinate system is converted to the planar projected coordinate system EPSG:3857 in order to perform area calculations, distance measurements, and spatial overlay analysis. Step S33: Remove duplicate data, abnormal coordinate data, and data with missing attributes; Step S34: Perform integrity, accuracy and consistency checks on the preprocessed data and generate a data quality report.
[0032] In step S31, the coordinate transformation algorithm adopts one of the following methods: seven-parameter transformation method, four-parameter transformation method, transformation based on a public transformation parameter library, or coordinate transformation based on a neural network model.
[0033] In step S33, the specific methods for data cleaning include: For duplicate data, a dual deduplication strategy using coordinates and attributes is adopted, with coordinate precision retained to 5 decimal places; Abnormal coordinate data is identified and removed through spatial range checks and spatial distribution cluster analysis. For data with missing attributes, handle it in one of the following ways: fill with the mean or median of the same type of data, estimate using spatial interpolation, or mark it as missing but do not delete it.
[0034] In step S34, the data quality assessment includes the following indicators: data integrity rate, coordinate accuracy pass rate, attribute field fill rate, logical consistency check pass rate, and consistency with reference data.
[0035] As one possible implementation of this embodiment, step S4 includes the following sub-steps: Step S41: Read the land space vector surface data formed by integrating business data such as land survey results; Step S42: Identify the feature layers in the vector surface data used to represent land parcel units, and filter out polygon features with independent spatial extents; Step S43: Take each independent polygon as a basic unit of spatial planning, and assign a unique identifier code block_id to each unit as the spatial carrier for subsequent analysis; Step S44: Generate a block polygon grid covering the study area, ensuring that there are no gaps or overlaps within the study area; Step S45: Perform coordinate system one on the geometric objects of all basic units and convert them into the EPSG:3857 planar projection coordinate system.
[0036] In step S42, land parcel unit elements are identified in the following ways: the attribute table structure of the vector data is analyzed to identify layers containing key fields such as land use type code, parcel number, and ownership information; the geometric type is checked to filter out polygonal elements; and the independence and integrity of the elements are verified through spatial overlay analysis.
[0037] In step S43, the encoding rule of block_id includes the following information: administrative division code, land parcel type code, and serial number. The encoding format is: AAAA-BB-CCCCCC, where AAAA is the administrative division code, BB is the land parcel type code, and CCCCCC is the six-digit serial number.
[0038] In step S44, when there are gaps or overlaps in the polygonal grid formed by the land business data, the following processing methods are adopted: for small gaps, they are filled by buffer analysis and fusion operation; for overlapping areas, the dominant polygon is determined and the boundary is corrected by overlay analysis and area ratio judgment.
[0039] As one possible implementation of this embodiment, step S5 includes the following sub-steps: Step S51, AOI evidence calculation: Calculate the intersection area between each AOI isal feature and each planning basic unit, which is used as the isal evidence strength to form an AOI evidence matrix A, where A(i,j) represents the AOI coverage area of the j-th land use type in the i-th planning basic unit. Step S52, POI evidence calculation: The reclassified POI point data is spatially allocated to the corresponding planning basic unit. The point evidence strength is calculated based on the category area weight of POI to form a POI evidence matrix P, where P(i,j) represents the weighted POI evidence of the j-th land use type in the i-th planning basic unit. Step S53: Weighted fusion of AOI and POI evidence strengths to calculate the total evidence strength for different types of functional land use in each basic planning unit. The calculation formula is: Where A is the AOI evidence matrix and P is the POI evidence matrix. These are the weighting coefficients for AOI evidence and POI evidence, respectively. Step S54: Perform row normalization on the total evidence strength matrix E to obtain the distribution matrix S of the proportion of different land use types in each basic planning unit; for the element in the i-th row and j-th column, the calculation formula is: ,in This represents the percentage of the j-th type of function in the i-th basic planning unit. It represents the total evidence strength for the j-th function of the i-th basic planning unit.
[0040] In step S51, the method for calculating the AOI evidence strength is as follows: for each basic planning unit i and each land use type j, calculate the sum of the intersection areas of all AOI elements of type j with unit i, and use it as the value of A(i,j).
[0041] In step S52, the method for calculating the strength of POI evidence is as follows: for each basic planning unit i and each land use type j, count the number of all POIs of type j within unit i, multiply by the average area weight of that type of POI, and use this as the value of P(i,j).
[0042] In step S53, the weighting coefficients and The method for determining this parameter is as follows: based on the data quality assessment results and historical validation data, it is calculated using the analytic hierarchy process (AHP) or the entropy weight method; or it can be set as an adjustable parameter with a default value. , .
[0043] In step S54, the normalization process also includes smoothing the proportion distribution matrix S to prevent division by zero errors and the influence of extreme values. Specifically, a very small smoothing coefficient ε is added to the denominator, and the formula is modified as follows: Where n is the total number of land use types, and ε takes the value of .
[0044] As one possible implementation of this embodiment, step S6 includes the following sub-steps: Step S61: Set up a multi-threshold determination mechanism, including a first threshold τ. aoi Second threshold τ major and the third threshold τ gap ; Step S62: For each planning basic unit, check whether the AOI coverage ratio exceeds the first threshold τ. aoi If so, the land use type corresponding to that AOI shall be used as the primary land use. Step S63: If no AOI meets the mandatory conditions, check the evidence ratio of each land use type. If the maximum evidence ratio exceeds the second threshold τ... major If so, then this type will be adopted as the primary land use. Step S64, if the proportion of the largest piece of evidence does not exceed τ major Furthermore, the difference in the proportion of the top two pieces of evidence is less than the third threshold τ. gap If so, the unit will be marked as a mixed-use land use; Step S65: If none of the above conditions are met, then the preset default type shall be used. Step S66: Based on neighborhood relationships, perform spatial smoothing on the preliminary judgment results to improve the spatial continuity of the results.
[0045] In step S61, the threshold setting principle is as follows: the first threshold τ aoi The value range is 0.5-0.8, and the default value is 0.6; the second threshold τ major The value range is 0.3-0.5, and the default value is 0.4; the third threshold τ gap The value range is 0.05-0.15, and the default value is 0.1.
[0046] In step S64, the mixed land use is marked by combining the codes of the two largest land use types, separated by "|", such as "09|07" which represents a mixture of commercial service facilities land and residential land.
[0047] As one possible implementation of this embodiment, in step S66, the spatial neighborhood smoothing algorithm includes the following steps: Step S661: Construct a neighborhood network for each planning basic unit based on the Queen adjacency relationship. If two planning basic units are connected at the boundary or corner, they are considered to be neighbors. Step S662: The functional category probability vector of each unit is updated using a multi-round iterative spatial smoothing update formula. In the t-th iteration, for any basic planning unit i, its functional vector update formula is: , in, This represents the functional category proportion vector of the i-th basic planning unit in the t-th iteration; This represents the set of cells adjacent to cell i; Number of neighbors; The smoothing intensity parameter is used to control the weight ratio between the original value and the neighborhood average, and the preferred value range is 0.1-0.5. Step S663, set the iteration termination condition: when the vector change between two consecutive iterations is less than the threshold δ or the maximum number of iterations T is reached. max The iteration stops when T is reached, with a default δ=0.001. max =10. After the iteration stops, based on the updated scaling vector. Then, re-execute the determination logic of steps S62 to S65 to obtain the smoothed land use type.
[0048] In step S662, the smoothing intensity parameter λ is dynamically adjusted according to the spatial heterogeneity of the planning basic unit: a larger value is taken in areas with gentle spatial changes, and a smaller value is taken in areas with drastic spatial changes. Specifically, it is determined by calculating the variance of the function vector in the neighborhood.
[0049] As one possible implementation of this embodiment, step S7 includes the following sub-steps: Step S71, Standardize data output: Organize the land use identification results into structured data, including the following fields: planning basic unit ID, land use type code, land use type name, mixed use identifier, confidence level, and the proportion of the top two functions; Step S72, Thematic Map Design: Based on the national land spatial planning mapping specifications, design the visual style of the land use thematic map, including: color scheme, symbol system, labeling rules, and legend design; Step S73, Map Element Configuration: Add necessary auxiliary elements to the thematic map, including: legend, scale bar, north arrow, latitude and longitude grid, map title, compiling unit, and compilation date; Step S74, Map Output and Publishing: Output the thematic map as a standard format file, supporting printing and digital publishing.
[0050] In step S71, the standardized data output supports the following formats: Shapefile, GeoJSON, GeoPackage, KML, and CSV; the confidence level is calculated as the proportion of evidence for the main land use type, and for units subject to mandatory AOI determination, the confidence level is fixed at 1.0.
[0051] In step S72, the thematic map design follows these principles: different land use types use distinct color schemes; residential land uses warm colors, commercial land uses bright colors, industrial land uses cool colors, green spaces and plazas use green tones, and water bodies use blue tones; mixed land uses gradients or checkerboard patterns of two main colors.
[0052] In step S74, the map output supports the following formats: PDF, PNG, JPEG, SVG, and Web map service; the output map file contains metadata information, including: coordinate reference system, scale, data source, processing date, and version information.
[0053] like Figure 2 As shown in the figure, an embodiment of the present invention provides a device for refining the use of national land space, comprising: The data acquisition module is used to acquire multi-source spatial data, which includes one or more of the following business data: point of interest (POI) data, area of interest (AOI) data, land survey results data, cadastral survey data, land ownership confirmation data, land use control approval data, real estate registration data, and study area boundary data; The data reclassification module is used to reclassify the original POI data and establish a mapping relationship between the original POI categories and the classification standards for land and sea use in national land space. The data preprocessing module is used to perform coordinate system transformation and standardization preprocessing on the multi-source spatial data; The basic planning unit generation module is used to generate basic land spatial planning units based on the business data. The evidence fusion module is used to fuse AOI area evidence and reclassified POI point evidence to calculate the comprehensive evidence strength of different land use types in each basic planning unit. The land use determination module is used to determine the land use type of each basic planning unit through a multi-threshold determination mechanism combined with a spatial neighborhood smoothing algorithm. The multi-threshold determination mechanism includes setting a first threshold for determining mandatory AOI coverage, a second threshold for determining the dominant function, and a third threshold for determining the degree of functional mixing. The spatial neighborhood smoothing algorithm iteratively optimizes the functional category probability vector of each basic planning unit based on the spatial adjacency relationship of each unit. The results output module is used to output the land use identification results and generate a thematic map of land use in national spatial planning.
[0054] As one possible implementation of this embodiment, the multi-threshold determination mechanism includes: Set the first threshold τ for AOI coverage in a certain planning unit. aoi Second threshold τ major and the third threshold τ gap Construct hierarchical judgment rules; AOI mandatory determination: When the coverage of AOI in a certain planning basic unit exceeds the first threshold τ aoi When this occurs, the land use type corresponding to the AOI shall be used as the primary land use. Majority determination: When the majority of evidence exceeds the second threshold τmajor When using the AOI, the land use type corresponding to that AOI shall be adopted as the primary land use. Mixed-use identification: The difference in the proportion of the current two pieces of evidence is less than the third threshold τ. gap At that time, it was marked as mixed land use; No evidence required: When there is no evidence, the default type is used.
[0055] As one possible implementation of this embodiment, the spatial neighborhood smoothing algorithm includes: The neighborhood network of each planning basic unit is constructed based on the Queen adjacency relationship. If two planning basic units are connected at the boundary or corner, they are considered to be neighbors. A multi-round iterative spatial smoothing update formula is used to update the function category probability vector of each unit. In the t-th iteration, for any basic planning unit i, the function vector update formula is: , in, This represents the functional category proportion vector of the i-th basic planning unit in the t-th iteration; This represents the set of cells adjacent to cell i; Number of neighbors; This is a smoothing intensity parameter used to control the weight ratio between the original value and the neighborhood average.
[0056] Taking a city as an example, this invention utilizes multi-source spatial data, including business data such as cadastral survey results, land title confirmation, land use control approval, and real estate registration results, as well as Point of Interest (POI) data and Area of Interest (AOI) data, to conduct an experiment on a land use identification method based on multi-source data fusion for land spatial planning. In this embodiment, the basic planning unit is specifically represented as a street block, hereinafter referred to as a street block. Figure 3 As shown, the specific steps for identifying land use in the city's territorial spatial planning using the device described in this invention are as follows: Step 1: Data Acquisition and Preprocessing The multi-source geospatial data within the study area were uniformly organized and standardized. The specific steps are as follows: (1) Data collection: Collect the following data within the study area: Based on the results of the land survey, integrate business data such as cadastral survey, land ownership confirmation, land use control approval and real estate registration results to form land space vector surface data, which serves as the basic unit for functional zone division; Point of Interest (POI) data, including category labels (such as commercial, residential, educational, industrial, etc.), latitude and longitude coordinates and other attribute information; Area of Interest (AOI) data, used to provide high-confidence samples of known functional categories; Optional inputs include multi-source POI data from different commercial geospatial platforms or open geospatial platforms such as OpenStreetMap (OSM); (2) Coordinate and format standardization: Convert all input data to a projected coordinate system (e.g., EPSG:3857) and check the integrity of attribute fields. For missing or abnormal data, fill in or remove it using default values; (3) Data cleaning and deduplication: The coordinate accuracy of the POI data is constrained to deduplicate the data. The deduplication latitude and longitude can be set to 5 decimal places (about 1 meter accuracy) to ensure that duplicate points at the same location are merged. (4) Generate basic planning units: Read the land space vector surface data formed by integrating business data such as land survey results, land rights confirmation, land use control approval and real estate registration results, based on the land survey results. Take each polygon with an independent spatial range as a basic planning unit (i.e. the smallest unit of spatial planning analysis), generate a basic planning unit grid (in this embodiment, it is implemented in the form of a block polygon grid), and assign a unique identifier code block_id to each unit as the spatial carrier for subsequent analysis.
[0057] Step 2: Calculation of evidence for neighborhood function: The multi-source data is overlaid with the street blocks to extract functional evidence within each block. The specific steps are as follows: (1) AOI area evidence calculation: through spatial overlay operation, calculate the coverage area of each functional category of AOI in each block; record the coverage area results as matrix A, with rows corresponding to blocks and columns corresponding to functional categories; if a block has no AOI coverage, its AOI evidence value is set to 0. (2) POI density evidence calculation: spatial matching of POIs with blocks, and statistical analysis of the number or area contribution of POIs for each function; for POIs not in a block, the nearest neighbor algorithm is used to assign them to the nearest block; the weight values of POIs in each block are summed according to the function category to obtain matrix P; (3) Multi-source evidence fusion, the AOI area matrix A and the POI weight matrix P are fused according to the set weights ( The weighted sum is calculated using the following formula: (1), In the formula, A is the AOI evidence matrix and P is the POI evidence matrix; Row normalization is performed on the evidence matrix for each block to obtain the distribution matrix S of the proportion of each functional category, as shown in the following formula: (2), in, This represents the percentage of the j-th function in the i-th block. It represents the total evidence strength for the j-th function of the i-th basic planning unit.
[0058] Step 3: Spatial neighborhood smoothing: To eliminate isolated misjudgments and improve the spatial continuity of functional areas, the proportion distribution matrix is smoothed over a neighborhood. The specific steps are as follows: (1) Adjacency relationship construction: Based on the block geometry, the Queen adjacency rule is used to construct the block adjacency matrix and record the neighbor list of each block; (2) The smoothing algorithm is implemented by using a spatial weighted average algorithm to smooth the functional proportion of the blocks. The smoothing intensity is controlled by λ, and the formula is as follows: (3), in, This represents the functional category proportion vector of the i-th basic planning unit in the t-th iteration; This represents the set of cells adjacent to cell i; Number of neighbors; This is a smoothing intensity parameter used to control the weight ratio between the original value and the neighborhood average; (3) Multi-round iterative smoothing: the smoothing rounds can be set to 1 to 5 rounds to enhance the spatial continuity of the classification results and reduce the "island effect".
[0059] Step 4: Functional Area Determination and Classification: After smoothing, the functional categories of each block are classified and labeled. The specific steps are as follows: (1) AOI forced classification: if the AOI coverage ratio of a certain block exceeds the threshold τ aoi If the confidence level is 0.5, then its main function category is directly specified as the AOI type, and the confidence level is set to 1. (2) Identification of main functions: Sort the function proportion vector of each block in descending order and select the function category f1 corresponding to the largest proportion; if the proportion of f1 is ≥ τ major If the value is 0.4, then it is determined to be the main category of that functional area; (3) Mixed function recognition: if the difference between the largest proportion f1 and the second largest proportion f2 is less than the threshold τ gapIf the value is 0.1, then the block is marked as a mixed-use zone (f1|f2) to reflect its multi-functional and complex characteristics; (4) Confidence calculation and classification result output: The confidence score is calculated based on the proportion of the main categories of each block. The higher the confidence score, the clearer the functional category. The confidence score of blocks for AOI forced classification is fixed at 1.
[0060] Step 5: Results Output and Visualization: The specific steps for storing and graphically representing the classification results are as follows: (1) Output the results file. The classification results of each block are output in GeoPackage (GPKG) format, including the following fields: block_id: unique identifier of the block; label: functional category code; label_name: Chinese name of the functional category; confidence: classification confidence; is_mixed: mixed functional area marker; top2: the top two functional categories; share_*: the proportion of each functional category; (2) Visualize the map, assign color palettes according to functional categories, and draw a functional distribution map of the blocks; display the boundaries of the study area, legend, latitude and longitude grid and other elements; export PNG image format as the result map of the distribution of urban functional areas; (3) Application of results: The classification results can be applied to scenarios such as urban planning, land use management, and functional layout analysis to achieve quantitative assessment and dynamic monitoring of urban spatial structure.
[0061] like Figure 4 As shown, this is a schematic diagram of the land use identification results of a city's territorial spatial planning generated using an embodiment of the present invention. Figure 4 The diagram uses street blocks as the basic spatial unit, with different blocks distinguished by different colors according to their dominant functional types. The legend on the right shows the corresponding functional category numbers and their meanings, including various urban construction and natural space types such as residential land, commercial and service facilities land, education and scientific research land, public management and public service land, road and transportation facilities land, park green space, and water area land. Figure 4 The boundaries of each block are represented by gray lines, and the study area as a whole exhibits an irregular, ribbon-like spatial distribution. Road-type land uses run through the blocks in a ribbon-like or linear structure, while commercial and public service blocks are mostly distributed along major roads in a ribbon-like or patchy pattern. Residential land uses are distributed in large clusters. This figure visually illustrates the spatial pattern of the land use identification results obtained by this invention based on multi-source data fusion for national land spatial planning.
[0062] An electronic device provided in this invention includes a processor, a memory, and a bus. The memory stores machine-readable instructions executable by the processor. When the electronic device is running, the processor communicates with the memory via the bus, and the processor executes the machine-readable instructions to perform the steps of the method for refining the use of any national land space as described above.
[0063] Specifically, the aforementioned memory and processor can be general-purpose memory and processor, without any specific limitations. When the processor runs the computer program stored in the memory, it can execute the aforementioned method for refining the use of national land space.
[0064] Corresponding to the above application startup method, this embodiment of the invention also provides a storage medium storing a computer program, which, when run by a processor, executes the steps of the above-described method for refining the use of any national land space.
[0065] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the protection scope of the claims of the present invention.
Claims
1. A method for refining land use in national land space, characterized in that, Includes the following steps: Step S1: Obtain multi-source spatial data, which includes one or more of the following business data: Point of Interest (POI) data, Area of Interest (AOI) data, land survey results data, cadastral survey data, land ownership confirmation data, land use control approval data, real estate registration data, and study area boundary data; Step S2: Reclassify the original POI data and establish a mapping relationship between the original POI categories and the classification standards for land and sea use in national land space. Step S3: Perform coordinate system transformation and standardization preprocessing on the multi-source spatial data; Step S4: Generate basic units for land and space planning based on business data; Step S5: Integrate the AOI area evidence and the reclassified POI point evidence to calculate the comprehensive evidence strength of different land use types in each basic planning unit; Step S6: Determine the land use type of each planning basic unit by combining a multi-threshold determination mechanism with a spatial neighborhood smoothing algorithm; the multi-threshold determination mechanism includes setting a first threshold for determining AOI mandatory coverage, a second threshold for determining the dominant function, and a third threshold for determining the functional mixing degree; the spatial neighborhood smoothing algorithm iteratively optimizes the functional category probability vector based on the spatial adjacency relationship of each planning basic unit. Step S7: Output the land use identification results and generate a thematic map of land use in the national land spatial planning. Step S2 includes: Step S21: Establish a multi-level classification mapping system between the original POI category and the land and sea use classification standards for land and sea use in land and space surveys, planning, and land use control. Step S22: Construct a keyword mapping mechanism to achieve intelligent conversion from the original POI category to the land use classification; Step S23: Establish an area weight database, including an average area database of POIs in different subcategories and a backup average area database of the first-level categories. The area weight database is constructed in the following way: collect sample data of various types of POIs, measure or estimate their actual area, calculate the average area of POIs of the same category, and for categories with insufficient sample size, use the average area of the next higher level category as a backup value. Step S24: Perform unified reclassification processing on POI data from different sources and output standardized land use classification codes; Step S5 includes: Step S51: Calculate the intersection area between each AOI isometric feature and each planning basic unit, which is used as the isometric evidence strength to form the AOI evidence matrix A; Step S52: The reclassified POI point data space is spatially allocated to the corresponding planning basic units, and the point evidence strength is calculated based on the category area weight of POI to form the POI evidence matrix P. Step S53: Weighted fusion of AOI and POI evidence strengths is performed to calculate the total evidence strength E for different types of functional land use in each basic planning unit: , These are the weighting coefficients for AOI evidence and POI evidence, respectively. Step S54: Perform row normalization on the total evidence strength matrix E to obtain the proportion distribution matrix S of different land use types in each planning basic unit.
2. The method for refining land use in national land space according to claim 1, characterized in that, Step S1 includes the following sub-steps: Step S11: Obtain Points of Interest (POI) data from a commercial geospatial platform via an API interface; Step S12: Obtain Area of Interest (AOI) data from the Open Geospatial Platform via API interface or data file; Step S13: Obtain land survey data from the land and natural resources authorities, and integrate cadastral survey, land ownership confirmation, land use control approval, and real estate registration data to form land space vector surface data; Step S14: Obtain the boundary vector surface data of the study area; Step S15: Perform format unification and preliminary quality checks on the acquired multi-source data.
3. The method for refining land use in national land space according to claim 1, characterized in that, Step S3 includes the following sub-steps: Step S31: The coordinate system from different data sources is converted to the WGS84 standard coordinate system using a coordinate transformation algorithm. Step S32: When performing spatial calculations, convert the WGS84 coordinate system to a planar projected coordinate system; Step S33: Remove duplicate data, abnormal coordinate data, and data with missing attributes; Step S34: Perform integrity, accuracy and consistency checks on the preprocessed data and generate a data quality report.
4. The method for refining land use in national land space according to claim 1, characterized in that, Step S4 includes the following sub-steps: Step S41: Read the land space vector surface data formed by integrating business data; Step S42: Identify the feature layers in the vector surface data used to represent land parcel units, and filter out polygon features with independent spatial extents; Step S43: Take each independent polygon as a basic unit of spatial planning and assign a unique identification code to each unit. Step S44: Generate a block polygon grid covering the study area; Step S45: Perform coordinate system one on all the geometric objects of the basic units and convert them into a planar projected coordinate system.
5. The method for refining land use in national land space according to any one of claims 1-4, characterized in that, Step S6 includes the following sub-steps: Step S61: Set up a multi-threshold determination mechanism, including a first threshold τ. aoi Second threshold τ major and the third threshold τ gap ; Step S62: For each planning basic unit, check whether the AOI coverage ratio exceeds the first threshold τ. aoi If so, the land use type corresponding to that AOI shall be used as the primary land use. Step S63: If no AOI meets the mandatory conditions, check the evidence ratio of each land use type. If the maximum evidence ratio exceeds the second threshold τ... major If so, then this type will be adopted as the primary land use. Step S64, if the proportion of the largest piece of evidence does not exceed τ major Furthermore, the difference in the proportion of the top two pieces of evidence is less than the third threshold τ. gap If so, the unit will be marked as a mixed-use land use; Step S65: If none of the above conditions are met, then the preset default type shall be used. Step S66: Perform spatial smoothing on the preliminary judgment results based on neighborhood relationships.
6. An apparatus for refining land use in national spatial planning, used to implement the method for refining land use in national spatial planning as described in any one of claims 1 to 5, characterized in that, include: The data acquisition module is used to acquire multi-source spatial data, which includes one or more of the following business data: point of interest (POI) data, area of interest (AOI) data, land survey results data, cadastral survey data, land ownership confirmation data, land use control approval data, real estate registration data, and study area boundary data; The data reclassification module is used to reclassify the original POI data and establish a mapping relationship between the original POI categories and the classification standards for land and sea use in national land space. The data preprocessing module is used to perform coordinate system transformation and standardization preprocessing on the multi-source spatial data; The basic planning unit generation module is used to generate basic land spatial planning units based on the business data. The evidence fusion module is used to fuse AOI area evidence and reclassified POI point evidence to calculate the comprehensive evidence strength of different land use types in each basic planning unit. The land use determination module is used to determine the land use type of each basic planning unit through a multi-threshold determination mechanism combined with a spatial neighborhood smoothing algorithm. The multi-threshold determination mechanism includes setting a first threshold for determining mandatory AOI coverage, a second threshold for determining the dominant function, and a third threshold for determining the degree of functional mixing. The spatial neighborhood smoothing algorithm iteratively optimizes the functional category probability vector of each basic planning unit based on the spatial adjacency relationship of each unit. The results output module is used to output the land use identification results and generate a thematic map of land use in national spatial planning.
7. An electronic device, characterized in that, The device includes a processor, a memory, and a bus. The memory stores machine-readable instructions that the processor can execute. When the electronic device is running, the processor communicates with the memory via the bus, and the processor executes the machine-readable instructions to perform the steps of the method for refining land use as described in any one of claims 1 to 5.
8. A storage medium, characterized in that, The storage medium stores a computer program that, when executed by a processor, performs the steps of the method for refining the land use of national spatial space as described in any one of claims 1 to 5.