Spatial matching method and system based on natural resource monitoring and demographic data

Abstract

The invention discloses a spatial matching method and system based on natural resource monitoring and demographic data. The method comprises the following steps: obtaining residential land and house building data in regional natural resource survey and monitoring data, using spatial position and attribute matching technology, combining with a random forest algorithm to spatially match residential land and house buildings, extracting residential house building patches and their attribute information; creating an attribute feature mapping between the residential house building patch scale and the administrative division unit scale, and using a random forest hyperparameter search method to convert administrative division demographic data to residential house building patches; further integrating multi-source data in natural resource survey and monitoring data, constructing a population geographic imprint index with a water area as a mask, and correcting the number of residential house building population; finally, calculating a population distribution fitting coefficient, iteratively correcting a population allocation error, and realizing accurate spatial matching of residential house buildings and population distribution through a zoning adjustment strategy.

Description

Technical Field

[0001] The present invention belongs to the technical field of population spatialization, and in particular relates to a spatial matching method and system based on natural resource monitoring and demographic data. Background Art

[0002] In order to realize the concept of harmonious coexistence between man and nature, facing the increasingly tense relationship between man and land and the huge challenges of ecological environment protection in my country, it is particularly important to efficiently and accurately integrate population data and natural resource data in geographic space. Population data is a key indicator reflecting the characteristics of human activities, and is usually counted according to multi-level administrative divisions; while authoritative and high-precision natural resource survey and monitoring data is the basis for protecting and managing the ecological environment and optimizing resource utilization. Therefore, achieving accurate spatial matching of natural resource survey and monitoring data with demographic data to form high-precision population spatial distribution data will greatly improve the scientificity and effectiveness of decision-making in areas such as resource management, environmental protection and regional planning.

[0003] However, there are few studies on the spatial matching methods of natural resource survey and monitoring data with demographic data, and there are certain limitations. First, existing methods often fail to make full use of the two key information of residential land and housing construction, resulting in blind spots in population matching, and research cases that comprehensively consider and accurately integrate these two types of data are still rare. Second, since demographic data are usually expressed in administrative units, while natural resource data are scaled in patches, this scale difference makes it complicated to accurately match population data to specific residential building patches. In addition, when dealing with population distribution, existing methods fail to fully consider the spatial differences inside and outside the residential building patches and the cross-scale effects of the characteristic relationships of related influencing factors, which affects the accuracy of the estimation. Finally, although the population distribution fitting coefficient is an important tool for evaluating the fitting accuracy, in practical applications, how to effectively calculate and adjust the population distribution error is still an issue that needs to be explored in depth. Summary of the invention

[0004] In view of the above-mentioned deficiencies of the prior art, the present application provides a spatial matching method and system based on natural resource monitoring and demographic data.

[0005] In the first aspect, the present application proposes a spatial matching method based on natural resource monitoring and demographic data, comprising the following steps:

[0006] Step S1: extract residential map data, house building map data, water network, road network, public service facilities, administrative division boundaries and DEM data from natural resource survey and monitoring data, and obtain population data, OSM road network, POI, night light and vegetation coverage data from the Internet;

[0007] Step S2: using spatial position and attribute matching in combination with a random forest algorithm, spatially matching the residential map patch data and the building map patch data, identifying and extracting residential map patch data having building attribute characteristics;

[0008] Step S3: integrating the residential map patch data with housing building attribute characteristics and associating the residential buildings in the data with the administrative division information, creating an attribute feature mapping of the residential map patch data with housing building attribute characteristics and the administrative division after associating, and using the random forest algorithm and the demographic data to allocate initial population numbers for different types of residential buildings to obtain an initial allocation result;

[0009] Step S4: constructing a population geographic footprint index by integrating the water network, road network, public service facilities, DEM, nighttime lights and vegetation coverage data, and correcting the initial allocation result to obtain a population distribution correction result;

[0010] Step S5: performing zoning statistics on the population distribution correction result according to the administrative division boundary lines of the same level in the administrative division boundary vector data, calculating the population distribution fitting coefficient, iteratively correcting the allocation error in the population distribution correction result, and obtaining the population distribution correction result;

[0011] Step S6: Perform zoning statistics on the population distribution correction results according to the secondary administrative division boundaries in the administrative division boundary vector data, and calculate the relative error between the initial value of the administrative division population and the statistical population value of each division. If the absolute value of the relative error value does not exceed the preset value, the spatial matching is completed. If the absolute value of the relative error value exceeds the preset value, the secondary administrative division is recorded as the area to be corrected and step S5 is executed. Further iterative correction is performed until the absolute value of the relative error value does not exceed the preset value, thereby completing the final spatial matching of residential buildings and population distribution.

[0012] In some embodiments, step S2 includes:

[0013] Step S21: pre-processing the residential map patch data and the house building patch data to ensure that the spatial reference systems of the two are consistent and the integrity of the residential house building patch attribute table is guaranteed, and then the spatial matching rules are specifically implemented: if the house building patch is completely located in the residential map patch, the house building patch is extracted and recorded as a residential house building; if the residential map patch is completely located inside the house building patch, the residential map patch is extracted and given a house building attribute feature; if the residential map patch intersects with the house building patch, the house building patch intersecting with the residential land is extracted, and the integrity of the intersecting patch is retained; if the residential building cannot be matched to the house building type feature, the random forest algorithm is used to infer the house building type attribute feature of the residential map patch;

[0014] Step S22: Integrate the residential building patch data. According to the spatial matching rules, merge all the identified and extracted residential building patch data into a unified layer. If attribute information is missing in the layer, supplement the corresponding patch attribute information according to the spatial matching rules.

[0015] In some embodiments, step S3 includes:

[0016] Step S31: Associating residential buildings with administrative division information, spatially superimposing the residential buildings with the administrative division boundary layer, and adding the relevant attribute information of the administrative division boundary to the residential building patch attribute table;

[0017] Step S32: converting the residential building pattern with building attribute characteristics into centroid data, classifying the centroid data according to the building type information, saving each classified centroid data as an independent layer, and retaining the area value of the original residential building pattern in the layer attribute table;

[0018] Step S33: using the spatial position coordinates of the centroid data and combining the corresponding residential building area as weights, perform spatial kernel density estimation to generate kernel density distribution raster data of various types of residential buildings;

[0019] Step S34: for the kernel density distribution grid data of various types of residential buildings, the data with a kernel density value of 0 is classified into the first level, and the data with a kernel density value greater than 0 is classified using the natural breakpoint method;

[0020] Step S35: constructing a feature vector based on the scale of the patch, and constructing an administrative division scale feature vector and a residential building feature vector according to the kernel density distribution level of the classified residential building patch;

[0021] Step S36: inputting the residential building feature vector, administrative division scale feature vector and administrative division demographic data as training samples into the random forest model to select the optimal parameters;

[0022] Step S37: Based on the statistical characteristics of the residential building pattern, the initial value of the residential building population is estimated through a random forest model with optimal parameters to obtain an initial allocation result.

[0023] In some embodiments, step S4 includes:

[0024] Step S41: constructing a population geographic footprint index by integrating DEM, public service facilities, road network, water network, nighttime lighting and vegetation coverage data, and correcting the initial population of the area to be corrected. The correction steps are as follows:

[0025] Step S42: Calculate the single factor weights according to the DEM, public service facilities, road network, water network, night light and vegetation coverage data and generate single factor grid distribution data respectively, wherein the single factor grid distribution data includes DEM single factor grid data, public service facilities single factor grid data, road network single factor grid data, water network single factor grid data, night light single factor grid data and vegetation coverage single factor grid data;

[0026] Step S43: performing equal-weighted addition on the DEM single-factor raster data, the public service facility single-factor raster data, the road network single-factor raster data, the water network single-factor raster data, the night light single-factor raster data and the vegetation coverage single-factor raster data to generate a population geographic imprint index;

[0027] Step S44: Use the population geographic imprint index to correct the initial distribution result extracted in step S37, and the correction formula is as follows:

[0028]

[0029] in, is the corrected population, is the population initially allocated to the classified residential buildings obtained in step S3, is the population geographic footprint index of grid j, It is the largest population geographic imprint index among all calibration grids. is an adjustment parameter that controls the impact of the population distribution indicator weight value on the correction result. A factor that stretches and limits the range of the correction weights.

[0030] In some embodiments, step S5 includes:

[0031] Step S51: Calculating the population distribution fitting coefficient of the administrative division at this level and correcting the grid population number includes: calculating the population distribution fitting coefficient, summing up the cumulative value of all grid population correction values ​​in the administrative division at this level and comparing it with the original population statistics of the administrative division. If the two are not equal, calculating the population distribution fitting coefficient, that is, the ratio of the administrative division population statistics to the grid population cumulative value, the calculation formula is as follows:

[0032]

[0033] Among them, F represents the population distribution fitting coefficient, represents the original population statistics of administrative divisions, It represents the sum of the population values ​​of all grids in the administrative area;

[0034] Correct the grid population, divide the population of each grid by the population distribution fitting coefficient, and get the corrected population of each grid. The formula is as follows:

[0035]

[0036] in, represents the corrected grid population, It represents the sum of the population values ​​of all grids in the administrative area;

[0037] Repeat the correction process for all grids and accumulate the corrected population numbers to check whether they are equal to the original population statistics of the administrative division. If the two are not equal, continue to iterate until the two are equal to obtain the population distribution correction result. The specific calculation process is:

[0038]

[0039] in, Represents the sum of the population numbers after correction for all grids.

[0040] In some embodiments, step S6 includes:

[0041] Calculate the population distribution fitting coefficient of the secondary administrative divisions and combine it with the zoning adjustment strategy to continue to correct the grid population:

[0042] Based on the sub-administrative division boundaries, the grid population values ​​in each sub-administrative division are counted, and the relative error between the grid population cumulative value and the original population statistical value of each sub-administrative division is calculated. The sub-administrative divisions with relative error absolute values ​​greater than the preset value are marked as areas to be adjusted. The calculation formula is as follows:

[0043]

[0044] Where E represents the relative error, It represents the cumulative sum of the population values ​​of the grids within the secondary administrative division. Indicates the original population statistics of administrative divisions;

[0045] For the area to be adjusted, the sub-administrative division boundary is spatially matched with the grid to determine the region to which each grid belongs;

[0046] Calculate the population adjustment coefficient within the area to be adjusted, that is, the ratio of the population correction value of each sub-administrative area to the population statistical value, and count the population adjustment coefficient of each sub-administrative area within the preset value as 1; then record the ratio of the population correction value of each grid to the population adjustment coefficient as the population number after adjustment of each grid, and take the integer value of the population number after adjustment. The calculation formula is as follows:

[0047]

[0048] in, is the population after grid adjustment, is the population modifier of the sub-district, is the corrected grid population, is the original population statistics of administrative divisions;

[0049] Summarize and calculate the difference in population numbers, calculate the difference between the population value of all grids of the secondary administrative divisions and the demographic value, and further correct the population value of each grid. The calculation formula is as follows:

[0050]

[0051] in, represents the sum of the population values ​​of all grids in the administrative area. is the original population statistics of the administrative division. If the difference is greater than 0, all grid population values ​​in the administrative division are sorted from large to small, and 1 is subtracted in sequence until the difference adjustment is 0; if the difference is less than 0, all grid population values ​​in the administrative division are sorted from large to small, and 1 is added in sequence until the difference adjustment is 0;

[0052] The total population after adjustment within the area is further summarized and compared with the actual total population to calculate the relative error of the sub-area summary. The calculation is shown as follows:

[0053]

[0054] in, is the relative error after partition aggregation, is the total population after adjustment, If the relative population error is still greater than the preset value, the iterative operation is repeated until the accuracy requirement is met.

[0055] In some embodiments, for the area to be adjusted, spatially matching the secondary administrative division boundary with the grid to determine the region to which each grid belongs includes:

[0056] Check whether the spatial location of each grid is completely located within a certain secondary administrative division. If the grid crosses the boundaries of multiple administrative divisions, it is necessary to calculate the area occupied by the residential building blocks in the grid within this administrative division. For grids that cross multiple administrative divisions, use the following formula to calculate the proportion of the area of ​​residential building blocks in each administrative division.

[0057]

[0058] in, represents the proportion of the residential building area of ​​the i-th administrative division in the grid, represents the residential building area of ​​the i-th administrative division within the grid, Indicates the area of ​​residential buildings within the grid;

[0059] Determine the administrative division with the largest residential building area within the grid as the administrative division to which the grid belongs.

[0060] In the second aspect, the present application proposes a spatial matching system based on natural resource monitoring and demographic data, including a data extraction module, a patch data matching module, an initial population division module, a verification module, a population correction module and a final matching module;

[0061] The data extraction module is used to execute step S1: extract residential map data, building map data, water network, road network, public service facilities, administrative division boundaries and DEM data from natural resource survey and monitoring data, and obtain demographic data, OSM road network, POI, night light and vegetation coverage data from the Internet;

[0062] The patch data matching module is used to execute step S2: using spatial position and attribute matching, combined with a random forest algorithm, to spatially match the residential map patch data and the building patch data, and identify and extract the residential map patch data having building attribute characteristics;

[0063] The initial population division module is used to execute step S3: integrating the residential map patch data with housing building attribute characteristics and associating the residential buildings in the data with administrative division information, creating an attribute feature mapping of the residential map patch data with housing building attribute characteristics and the administrative division after association, and using the random forest algorithm and the demographic data to allocate initial population numbers for different types of residential buildings to obtain an initial allocation result;

[0064] The population correction module is used to execute step S4: constructing a population geographic imprint index by integrating multi-source data in the natural resource survey and monitoring data, correcting the initial distribution result, and obtaining a population distribution correction result;

[0065] The verification module is used to execute step S5: perform zoning statistics on the population distribution correction result according to the administrative division boundary lines of the same level in the administrative division boundary vector data, calculate the population distribution fitting coefficient, iteratively correct the allocation error in the population distribution correction result, and obtain the population distribution correction result;

[0066] The final matching module is used to execute step S6: perform zoning statistics on the population distribution correction results according to the secondary administrative division boundaries in the administrative division boundary vector data, calculate the relative error between the initial value of the administrative division population of each partition and the statistical population value, if the absolute value of the relative error value does not exceed the preset value, then complete this spatial matching, if the absolute value of the relative error value exceeds the preset value, record the secondary administrative division as a to-be-corrected area and execute step S5, further iterate and correct until the absolute value of the relative error value does not exceed the preset value, thereby completing the final spatial matching of residential buildings and population distribution.

[0067] In a third aspect, the present application proposes an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above method when executing the computer program.

[0068] In a fourth aspect, the present application proposes a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above method are implemented.

[0069] Beneficial effects of the present invention:

[0070] 1. Improved algorithm performance and innovative fusion rules: An advanced algorithm combining spatial location and attribute matching is adopted, especially the introduction of the random forest algorithm, which can more accurately process a large number of complex spatial data sets and improve the spatial matching accuracy of residential land and house building data. The selection and application of this algorithm is uncommon in the existing technology. It can effectively process a large amount of data and reduce the possibility of mismatching. In addition, by establishing fusion rules between natural resource survey and monitoring data from different sources (residential land in land survey data and house building data in geographic national conditions monitoring data). This innovation not only optimizes the extraction process of residential building house spots and their attribute information, but also provides a high-quality basic data source for the subsequent generation of high-precision population distribution data.

[0071] 2. Innovative scale mapping technology: In view of the limitation of traditional methods that directly transplant the correlation between population and various factors based on administrative division scale to the patch level, a method of mapping attribute characteristics between residential building patch level and administrative division unit scale is proposed. This method effectively converts the attribute information of residential building scale to administrative division scale, and establishes a quantitative mapping relationship between the attribute feature vectors of buildings at two scales. The random forest hyperparameter search method is further used to achieve the accurate conversion of demographic data from administrative division unit to residential building patch. This method not only effectively solves the problem of fine expression of attribute characteristics at different spatial scales, but also significantly improves the accuracy and efficiency of spatial matching of cross-scale data.

[0072] 3. Data source improvement and refined fusion processing: This invention is not a simple data source matching, but by establishing specific rules to integrate the urban land space monitoring road network and public service facility data in the natural resource survey and monitoring data with the OSM road network data and the Amap POI data, a more complete, refined and practical road network and public service facility data are constructed. This innovative data fusion strategy not only provides a solid data foundation for the subsequent spatial matching fine modeling, but also significantly improves the accuracy and reliability of spatial matching.

[0073] 4. Construct a correction mechanism for the population geographic footprint index: By integrating multi-source data such as DEM, public service facilities, water networks, and road networks in natural resource survey and monitoring data, and combining other data such as POI, night lights, and vegetation coverage, a population geographic footprint index with water areas as a mask is constructed, which can more accurately correct the population number at the residential building map level. This innovation helps to improve the authenticity and reliability of population distribution data.

[0074] 5. Optimize the consistency of population distribution and natural resource element distribution: By iteratively correcting population allocation errors and combining zoning adjustment strategies, accurate spatial matching of residential buildings and population distribution is achieved. This strategy not only improves the accuracy of spatial matching, but also ensures the consistency and rationality of the population distribution in the entire region. BRIEF DESCRIPTION OF THE DRAWINGS

[0075] Figure 1 It is the overall flow chart of the present invention.

[0076] Figure 2 It is a system principle block diagram of the present invention. DETAILED DESCRIPTION

[0077] The exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments described herein; on the contrary, these embodiments are provided to enable a more thorough understanding of the present invention and to fully convey the scope of the present invention to those skilled in the art.

[0078] In the first aspect, the present application proposes a spatial matching method based on natural resource monitoring and demographic data, comprising the following steps:

[0079] Step S1: extract residential map data, house building map data, water network, road network, public service facilities, administrative division boundaries and DEM data from natural resource survey and monitoring data, and obtain population data, OSM road network, POI, night light and vegetation coverage data from the Internet;

[0080] Among them, the residential map data, house building map data, public service facilities, water network, road network, DEM and other related data of the target area are obtained from the natural resources survey and monitoring data, and other data such as public demographic data, OSM road network, POI, night lights and vegetation coverage are obtained; the OSM road network is spatially integrated with the urban land space monitoring road network data; the POI data is spatially integrated with the urban land space monitoring public service facility data;

[0081] S11. Obtain relevant data such as residential map patches, house building map patches, public service facilities, water network, road network, DEM, etc. in the area from natural resource survey and monitoring data.

[0082] Residential land is extracted from land survey data, and residential land is divided into urban residential land and rural residential land according to land classification codes.

[0083] Extract DEM data from geographic national conditions data; further extract housing building patch data from geographic national conditions data, and divide housing buildings into 10 types according to patch coding, including high-density multi-story and above housing building areas, low-density multi-story and above housing building areas, high-density low-rise housing building areas, low-density low-rise housing building areas, multi-story independent housing buildings, medium- and high-rise independent housing buildings, high-rise independent housing buildings, super-high-rise independent housing buildings, low-rise independent housing buildings, and abandoned housing buildings. The extracted residential land and housing building patch data are saved as two surface feature layers respectively.

[0084] Data on five types of public service facilities, including schools, medical institutions, parks and squares, cultural and sports facilities, and transportation service stations, are extracted from the urban land space monitoring data, and the relevant surface layer data are converted into point features, the surface layer facility area values ​​are retained, and merged with the point layer data, and saved as five point feature layers respectively; further, water network and road network (railway, urban roads, highways, and rural roads) data are extracted from the urban land space monitoring data and saved as two linear feature layers respectively.

[0085] Extract the administrative division boundary vector data of the region from the land survey data, and obtain the corresponding administrative division population statistics through other channels such as the statistical bureau website or statistical bulletin, regional statistical yearbook, etc. Spatially associate the obtained population statistics with the administrative division boundary vector, and append the population count to the attribute table of the administrative division boundary vector data.

[0086] S12. Obtain public demographic data, OSM road network, POI, night lights, vegetation coverage and other data; perform spatial fusion processing on POI data and urban land space monitoring public service facility data.

[0087] Based on the five types of public service facility data extracted in step S11, namely schools, medical institutions, parks and squares, cultural and sports facilities, and transportation service stations, use the "Merge" tool to spatially merge the above five types of public service facility data with the POI data; after the merge is completed, use the "Select Features" tool and the "Delete Features" operation in the "Edit" function to remove duplicate features; then use the "Buffer" tool to create a 200-meter buffer zone, and merge the public service facilities with the same name that fall within the buffer zone, and save the processed public service facility data as a separate layer.

[0088] S13. Obtain the road network data of the area from the OSM data, and perform spatial fusion with the road network data extracted from the urban land space monitoring data. The steps for spatial fusion of different types of road network data are as follows:

[0089] (1) Railway spatial integration

[0090] Extract the rail pavement features with the attribute field "fclass" of "rail" in the OSM road network data and import them into ArcGIS software; use the "Merge" tool to spatially merge the rail pavement features selected in the previous step with the railway line layer (LRRL) in the urban land space monitoring data; after the merge is completed, use the "Select Features" tool and the "Delete Features" operation in the "Edit" function to remove duplicate railway features; use the "Topology" tool to interrupt the intersecting lines of the merged layer.

[0091] (2) Highway spatial integration

[0092] Trackless roads outside the city were screened from the OSM road network data, and according to the types of roads monitored in the urban land space, the trackless roads outside the city in the OSM road network were distinguished as national roads, provincial roads, county roads, township roads, special roads, connecting roads, and other roads. They were imported into ArcGIS software, and the above roads were spatially merged. The screened roads were merged with the highway line layer (LRDL) in the urban land space monitoring data; the duplicate roads after the merger were removed, and the intersecting lines of the merged highway layer were interrupted.

[0093] (3) Urban road space integration

[0094] The trackless roads within the city were screened from the OSM road network data, and based on the urban road types monitored in the urban land space, the trackless roads within the city in the OSM road network were distinguished as expressways, main roads, secondary roads, branch roads, subways, ground rails, and other roads. The roads were then imported into ArcGIS software, and the above urban roads were spatially merged. The screened urban roads of various types were merged with the urban road line layer (LCTL) in the urban land space monitoring data. The duplicate urban roads after the merger were removed, and the "topology" tool was used to interrupt the intersecting lines of the merged urban road layer.

[0095] (4) Spatial integration of rural roads

[0096] The trackless roads outside the city were screened from the OSM road network data, and the trackless roads outside the city in the OSM road network were distinguished as rural hardened roads, mechanized roads, and rural roads based on the types of rural roads monitored in the urban land space. At the same time, the small roads in the OSM road network data were classified to identify rural hardened roads and mechanized roads. The identified rural roads were imported into ArcGIS software, and the above-screened rural roads were spatially merged and merged with the rural road line layer (LVLL) in the urban land space monitoring data; the duplicate rural roads after the merger were removed, and the "topology" tool was used to interrupt the intersecting lines of the merged rural road layer.

[0097] (5) Ramp space integration

[0098] Trackless pavement features with “fclass” marked as “link” were extracted from the OSM road network data and merged with the ramp line layer (LLKL) in the urban land space monitoring data in ArcGIS software. Then, duplicate elements were removed, and the “topology” tool was used to interrupt the intersecting lines of the merged rural road layer.

[0099] The corresponding relationship between OSM road network data and urban land space monitoring road network data is shown in Table 1 below:

[0100] Table 1: Correspondence between OSM road network data and urban land space monitoring road network data

[0101] (6) The road network data obtained by integrating the OSM road network with the urban land space monitoring road network data is used as the final road network data.

[0102] Step S2: using spatial position and attribute matching in combination with a random forest algorithm, spatially matching the residential map patch data and the building map patch data, identifying and extracting residential map patch data having building attribute characteristics;

[0103] In some embodiments, step S2 includes:

[0104] Step S21: pre-processing the residential map patch data and the house building patch data to ensure that the spatial reference systems of the two are consistent and the integrity of the residential house building patch attribute table is guaranteed, and then the spatial matching rules are specifically implemented: if the house building patch is completely located in the residential map patch, the house building patch is extracted and recorded as a residential house building; if the residential map patch is completely located inside the house building patch, the residential map patch is extracted and given a house building attribute feature; if the residential map patch intersects with the house building patch, the house building patch intersecting with the residential land is extracted, and the integrity of the intersecting patch is retained; if the residential building cannot be matched to the house building type feature, the random forest algorithm is used to infer the house building type attribute feature of the residential map patch;

[0105] The specific process is as follows:

[0106] (1) Extract relevant features from the residential map patches of known house building types in the above-mentioned areas, including area, shape index, distance to the nearest road, elevation, slope, distance to the nearest bus stop, etc., and form a feature matrix with the extracted features.

[0107] (2) The feature matrix is ​​combined with the corresponding building type attribute labels to form a training set. The model is further trained using the sklearn.ensemble.RandomForestClassifier library in Python, and the hyperparameter search method is used to set appropriate model parameters such as the number of decision trees, tree depth, sample ratio, and feature ratio.

[0108] (3) Use sklearn.model_selection.StratifiedKFold for stratified cross validation to evaluate the model.

[0109] (4) Extract the corresponding features of the unmatched residential map patches to form a feature matrix. Use the trained random forest model rf_model to make predictions and output the prediction results as an array containing the building type label.

[0110] (5) Conduct a field survey of the prediction results, or compare them with Amap to verify the accuracy of the prediction. Repeat the above steps and verify them to ensure that the model accurately predicts the type of house buildings on residential land.

[0111] Step S22: Integrate the residential building patch data. According to the spatial matching rules, merge all the identified and extracted residential building patch data into a unified layer. If there is missing attribute information in the layer, supplement the corresponding patch attribute information according to the spatial matching rules to ensure that the attribute information of each extracted patch is complete.

[0112] Step S3: integrating the residential map patch data with housing building attribute characteristics and associating the residential buildings in the data with the administrative division information, creating an attribute feature mapping of the residential map patch data with housing building attribute characteristics and the administrative division after associating, and using the random forest algorithm and the demographic data to allocate initial population numbers for different types of residential buildings to obtain an initial allocation result;

[0113] In some embodiments, step S3 includes:

[0114] Step S31: Associating residential buildings with administrative division information, spatially superimposing the residential buildings with the administrative division boundary layer, and adding the relevant attribute information of the administrative division boundary to the residential building patch attribute table;

[0115] The specific steps of integrating residential map patch data with building attribute characteristics and associating residential buildings in the data with administrative division information are as follows:

[0116] (1) Import the residential building layer (the layer attribute field information includes building ID, building type code, building type, and building area) and the administrative division boundary layer (the layer attribute field information includes: administrative division ID, administrative division code, administrative division name, and population) into ArcGIS software.

[0117] (2) Use the “Intersect” tool in ArcGIS software to select the residential building layer and the administrative boundary layer for spatial overlay.

[0118] (3) The newly generated overlay layer will include the intersection of each residential building block and the administrative division in which it is located. The attribute table will include the following fields: building ID, building type code, building type, building area, administrative division code, administrative division name, and population.

[0119] (4) Review the overlay results. For example, the attribute table of map A shows: "The building ID is 123, the building type code is 0511, the building type is a high-density multi-story and above residential building area, the building area is 2,000 square meters, the administrative division code is 110108, the administrative division name is Beijing, and the population is 3,133,469 people."

[0120] Step S32: converting the residential building pattern with building attribute characteristics into centroid data, classifying the centroid data according to the building type information, saving each classified centroid data as an independent layer, and retaining the area value of the original residential building pattern in the layer attribute table;

[0121] Step S33: using the spatial position coordinates of the centroid data and combining the corresponding residential building area as weights, perform spatial kernel density estimation to generate kernel density distribution raster data of various types of residential buildings;

[0122] Step S34: for the kernel density distribution grid data of various types of residential buildings, the data with a kernel density value of 0 is classified into the first level, and the data with a kernel density value greater than 0 is classified using the natural breakpoint method;

[0123] Step S35: constructing a feature vector based on the scale of the patch, and constructing an administrative division scale feature vector and a residential building feature vector according to the kernel density distribution level of the classified residential building patch;

[0124] Among them, the feature vector based on the scale of the patch is constructed. According to the kernel density distribution level of the classified residential building patch, the corresponding feature vector is constructed. If the kernel density of a residential building belongs to the kth level, the kth level of the feature vector is assigned a value of 1, and the rest of the positions are assigned a value of 0.

[0125] For example, if the kernel density data of high-density multi-story and above residential buildings are divided into 6 levels by the natural breakpoint method, namely 0, (0, 0.28), [0.28, 3.45), [3.45, 13.58), [13.58, 55.46], [55.46, 459.46), then the feature vector corresponding to the pixel with a kernel density value of 16.25 is encoded as (0, 0, 0, 0, 1, 0). The feature vectors of various types of residential buildings are shown below.

[0126] Characteristic vector of high-density multi-story and above housing construction area: [[1 0 0 0 0 0][1 0 0 0 0 0][0 0 01 0 0]...;

[0127] Characteristic vector of low-density multi-story and above housing construction area: [0 0 1 0 0 0][0 0 0 0 1 0]...;

[0128] Construct the feature vector of administrative division scale. The feature vector of administrative division scale is the weighted average of the feature vectors of each type of residential building patch. If the patch crosses the administrative division boundary, it will be assigned to the administrative division with the largest patch area. Each administrative division feature vector is shown below. [[0.57079231 0.12758243 0.10922019 0.11159556 0.8808095]; [0.24213938 0.11932289 0.1619665 0.21426256 0.26230866]; [0.83174793 0.0238409 0.03342518 8.13310171 0.77788427]; [8.5336769 010506718 0.11248473 0.10846317 0.14030802]; [0.40440948 011568971 0.12860256 0.15701512 0.19428314];

[0134] Step S36: inputting the residential building feature vector, administrative division scale feature vector and administrative division demographic data as training samples into the random forest model to select the optimal parameters;

[0135] Furthermore, the random forest algorithm is used to make a preliminary distribution of the population. The residential building feature vector, administrative division feature vector, and administrative division demographic data are input into the random forest model as training samples, and the data set is divided into a training set and a test set. Then a hyperparameter search is performed to select the optimal parameters for the number of decision trees (n_estimators), the maximum depth of the tree (max_depth), the minimum number of samples required for splitting (min_samples_split), and the minimum number of samples required for leaf nodes (min_samples_leaf).

[0136] Step S37: Based on the statistical characteristics of the residential building pattern, the initial value of the residential building population is estimated through a random forest model with optimal parameters to obtain an initial allocation result.

[0137] Step S4: constructing a population geographic imprint index by integrating the water network, road network, public service facilities, DEM, nighttime lights and vegetation coverage data, correcting the initial allocation result, and obtaining a population distribution correction result;

[0138] Among them, the preset value in this embodiment is 5%;

[0139] In some embodiments, step S4 includes:

[0140] Step S41: constructing a population geographic footprint index by integrating DEM, public service facilities, road network and water network data, and correcting the initial population of the area to be corrected. The correction steps are as follows:

[0141] Step S42: Calculate the single factor weights according to the DEM, public service facilities, road network and water network data and generate single factor raster distribution data respectively, wherein the single factor raster distribution data includes DEM single factor raster data, public service facilities single factor raster data, road network single factor raster data and water network single factor raster data;

[0142] Among them, 1. Based on the extracted DEM data, the elevation values ​​are assigned from 1 to 10 points from large to small using the quantile method.

[0143] 2. Based on the extracted water network data, calculate the closest distance from each residential building to the water network, and generate raster data of the closest distance to the water network. Using the quantile method, the closest distance value is divided into 10 levels, and the values ​​are assigned from 1 to 10 from large to small.

[0144] 3. Based on the fused public service facility data, the spatial kernel density estimation is performed with the area value of the public service facility as the weight value to generate the kernel density distribution raster data of various types of public service facilities. The raster with a density value of 0 is assigned a value of 0; the raster with a density greater than 0 is divided into 10 levels using the quantile method, and values ​​of 1 to 10 are assigned in descending order.

[0145] 4. Based on the extracted night light raster data, the grid with a pixel value of 0 is assigned a score of 0; the grid with a pixel value greater than 0 is divided into 10 levels using the quantile method, and assigned scores of 1 to 10 from small to large.

[0146] 5. Based on the extracted vegetation coverage data, the grids with pixel values ​​less than or equal to 0 are assigned a score of 0; the grids with pixel values ​​greater than 0 are divided into 10 levels using the quantile method, and are assigned scores of 1 to 10 from large to small.

[0147] 6. Based on the five types of road network data after fusion processing, the road density is calculated respectively and the corresponding road density raster data is generated; the raster with a density of 0 is assigned a value of 0; the raster with a density greater than 0 is divided into 10 levels using the quantile method, and the values ​​are assigned from 1 to 10 from small to large; the normalization processing is performed with railway (2), highway (4), urban road (5), rural road (3), and ramp (1) as weights. For each raster pixel, according to the road type to which it belongs, the corresponding road density value is multiplied by the corresponding weight normalization value. Finally, the five weighted synthesized raster layers are spatially superimposed to obtain the final road density raster layer.

[0148] Step S43: performing equal-weighted addition on the DEM single-factor raster data, the public service facility single-factor raster data, the road network single-factor raster data, and the water network single-factor raster data to generate a population geographic imprint index;

[0149] The six single-factor grid data in step S42 are equally weighted and added to generate the population geographic imprint index grid data. The specific steps are as follows:

[0150] 1. Use the raster calculator in ArcGIS software to perform equal-weighted addition on the above six raster layers and save the result as a new raster layer, which is the initial population geographic imprint index raster data.

[0151] 2. Extract water feature areas such as river surfaces, lake surfaces, coastal tidal flats, inland tidal flats, and swamps from the land survey data, and create a mask layer in which the pixel value of the water area is set to 0, excluding uninhabited areas of the water.

[0152] 3. Use the raster calculator in ArcGIS software to spatially overlay the mask layer with the population geographic imprint index raster data, eliminate the values ​​of the water area, and obtain the final population geographic imprint index.

[0153] Step S44: Use the population geographic imprint index to correct the initial distribution result extracted in step S37, and the correction formula is as follows:

[0154]

[0155] in, is the corrected population, is the population initially allocated to the classified residential buildings obtained in step S3, is the population geographic footprint index of grid j, It is the largest population geographic imprint index among all calibration grids. is an adjustment parameter that controls the impact of the population distribution indicator weight value on the correction result. A factor that stretches and limits the range of the correction weights.

[0156] in, is an adjustment parameter (after continuous debugging, initially set to 1.5) to control the impact of the population distribution indicator weight value on the correction result. The coefficient representing the range of stretching and limiting correction weights (after continuous debugging, initially set to 1);

[0157] Step S5: Perform zoning statistics on the population distribution correction result according to the administrative division boundary lines of the same level in the administrative division boundary vector data, calculate the population distribution fitting coefficient, iteratively correct the allocation error in the population distribution correction result, and obtain the population distribution correction result

[0158] In some embodiments, step S5 includes:

[0159] Calculating the population distribution fitting coefficient of the administrative division at this level and correcting the grid population number includes: calculating the population distribution fitting coefficient, summing up the cumulative value of all grid population correction values ​​in the administrative division at this level and comparing it with the original population statistics of the administrative division. If the two are not equal, calculating the population distribution fitting coefficient, that is, the ratio of the administrative division population statistics to the grid population cumulative value, the calculation formula is as follows:

[0160]

[0161] Among them, F represents the population distribution fitting coefficient, represents the original population statistics of administrative divisions, It represents the sum of the population values ​​of all grids in the administrative area;

[0162] Correct the grid population, divide the population of each grid by the population distribution fitting coefficient, and get the corrected population of each grid. The formula is as follows:

[0163]

[0164] in, represents the corrected grid population, It represents the sum of the population values ​​of all grids in the administrative area;

[0165] Repeat the correction process for all grids and accumulate the corrected population numbers to check whether they are equal to the original population statistics of the administrative division. If the two are not equal, continue to iterate until the two are equal to obtain the population distribution correction result. The specific calculation process is:

[0166]

[0167] in, Represents the sum of the population numbers after correction for all grids.

[0168] In some embodiments, step S6 further includes:

[0169] Step S6: Calculate the population distribution fitting coefficient, iteratively correct the allocation error in the population distribution correction result, and complete the final spatial matching of residential buildings and population distribution through the zoning adjustment strategy.

[0170] Step S62: Calculate the population distribution fitting coefficient of the secondary administrative divisions and combine it with the zoning adjustment strategy to continue to correct the grid population:

[0171] Based on the sub-administrative division boundaries, the grid population values ​​in each sub-administrative division are counted, and the relative error between the grid population cumulative value and the original population statistical value of each sub-administrative division is calculated. The sub-administrative divisions with relative error absolute values ​​greater than the preset value are marked as areas to be adjusted. The calculation formula is as follows:

[0172]

[0173] Where E represents the relative error, It represents the cumulative sum of the population values ​​of the grids within the secondary administrative division. Indicates the original population statistics of administrative divisions;

[0174] For the area to be adjusted, the sub-administrative division boundary is spatially matched with the grid to determine the region to which each grid belongs;

[0175] Calculate the population adjustment coefficient within the area to be adjusted, that is, the ratio of the population correction value of each sub-administrative area to the population statistical value, and count the population adjustment coefficient of each sub-administrative area within the preset value as 1; then record the ratio of the population correction value of each grid to the population adjustment coefficient as the population number after adjustment of each grid, and take the integer value of the population number after adjustment. The calculation formula is as follows:

[0176]

[0177] in, is the population after grid adjustment, is the population modifier of the sub-district, is the corrected grid population, For administrative divisions

[0178] Initial population statistics;

[0179] Summarize and calculate the difference in population numbers, calculate the difference between the population value of all grids of the secondary administrative divisions and the demographic value, and further correct the population value of each grid. The calculation formula is as follows:

[0180]

[0181] in, represents the sum of the population values ​​of all grids in the administrative area. is the original population statistics of the administrative division. If the difference is greater than 0, all grid population values ​​in the administrative division are sorted from large to small, and 1 is subtracted in sequence until the difference adjustment is 0; if the difference is less than 0, all grid population values ​​in the administrative division are sorted from large to small, and 1 is added in sequence until the difference adjustment is 0;

[0182] The total population after adjustment within the area is further summarized and compared with the actual total population to calculate the relative error of the sub-area summary. The calculation is shown as follows:

[0183]

[0184] in, is the relative error after partition aggregation, is the total population after adjustment, If the relative population error is still greater than the preset value, the iterative operation is repeated until the accuracy requirement is met.

[0185] In some embodiments, for the area to be adjusted, spatially matching the secondary administrative division boundary with the grid to determine the region to which each grid belongs includes:

[0186] Check whether the spatial location of each grid is completely located within a certain secondary administrative division. If the grid crosses the boundaries of multiple administrative divisions, it is necessary to calculate the area occupied by the residential building blocks in the grid within this administrative division. For grids that cross multiple administrative divisions, use the following formula to calculate the proportion of the area of ​​residential building blocks in each administrative division.

[0187]

[0188] in, represents the proportion of the residential building area of ​​the i-th administrative division in the grid, represents the residential building area of ​​the i-th administrative division within the grid, Indicates the area of ​​residential buildings within the grid;

[0189] Determine the administrative division with the largest residential building area within the grid as the administrative division to which the grid belongs.

[0190] In the second aspect, the present application proposes a spatial matching system based on natural resource monitoring and demographic data, including a data extraction module, a patch data matching module, an initial population division module, a verification module, a population correction module and a final matching module;

[0191] The data extraction module is used to execute step S1: extract residential map patch data, house building map patch data, water network, road network, public service facilities, administrative division boundary vector data and DEM data from natural resource survey and monitoring data, and obtain demographic data, OSM road network, POI, night light and vegetation coverage data from the Internet;

[0192] The patch data matching module is used to execute step S2: using spatial position and attribute matching, combined with a random forest algorithm, to spatially match the residential map patch data and the building patch data, and identify and extract the residential map patch data having building attribute characteristics;

[0193] The initial population division module is used to execute step S3: integrating the residential map patch data with housing building attribute characteristics and associating the residential buildings in the data with administrative division information, creating an attribute feature mapping of the residential map patch data with housing building attribute characteristics and the administrative division after association, and using the random forest algorithm and the demographic data to allocate initial population numbers for different types of residential buildings to obtain an initial allocation result;

[0194] The population correction module is used to execute step S4: constructing a population geographic imprint index by integrating multi-source data in the natural resource survey and monitoring data, correcting the initial distribution result, and obtaining a population distribution correction result;

[0195] The verification module is used to execute step S5: perform zoning statistics on the population distribution correction result according to the administrative division boundary lines of the same level in the administrative division boundary vector data, calculate the population distribution fitting coefficient, iteratively correct the allocation error in the population distribution correction result, and obtain the population distribution correction result;

[0196] The final matching module is used to execute step S6: perform zoning statistics on the population distribution correction results according to the secondary administrative division boundaries in the administrative division boundary vector data, calculate the relative error between the initial value of the administrative division population of each partition and the statistical population value, if the absolute value of the relative error value does not exceed the preset value, then complete this spatial matching, if the absolute value of the relative error value exceeds the preset value, record the secondary administrative division as a to-be-corrected area and execute step S5, further iterate and correct until the absolute value of the relative error value does not exceed the preset value, thereby completing the final spatial matching of residential buildings and population distribution.

[0197] In a third aspect, the present application proposes an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above method when executing the computer program.

[0198] In a fourth aspect, the present application proposes a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above method are implemented.

[0199] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0200] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0201] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this disclosure.

[0202] In the embodiments provided in the present disclosure, it should be understood that the disclosed apparatus / computer equipment and methods can be implemented in other ways. For example, the apparatus / computer equipment embodiments described above are merely schematic. For example, the division of modules or units is only a logical function division. There may be other division methods in actual implementation. Multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed may be through some interfaces, indirect coupling or communication connection of the apparatus or unit, which may be electrical, mechanical or other forms.

[0203] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0204] In addition, each functional unit in each embodiment of the present disclosure may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0205] If the integrated module / unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present disclosure implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. The computer program may include computer program code, and the computer program code may be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium. It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, the computer-readable medium does not include electric carrier signals and telecommunication signals.

[0206] The above are only preferred implementations of the present invention. It should be pointed out that a number of modifications and improved technical solutions made by those skilled in the art without departing from the technical solution should also be deemed to fall within the scope of protection required by the claims.

Claims

1. A spatial matching method based on natural resource monitoring and demographic data, characterized by: The following steps are involved: Step S1: extract residential map data, building map data, water network, road network, public service facilities, administrative division boundary vector data and DEM data from natural resource survey and monitoring data, and obtain demographic data, OSM road network, POI, night light and vegetation coverage data from the Internet; Step S2: using spatial position and attribute matching in combination with a random forest algorithm, spatially matching the residential map patch data and the building map patch data, identifying and extracting residential map patch data having building attribute characteristics; Step S3: integrating the residential map patch data with housing building attribute characteristics and associating the residential buildings in the data with the administrative division information, creating an attribute feature mapping of the residential map patch data with housing building attribute characteristics and the administrative division after associating, and using the random forest algorithm and the demographic data to allocate initial population numbers for different types of residential buildings to obtain an initial allocation result; Step S4: constructing a population geographic footprint index by integrating the water network, road network, public service facilities, DEM, nighttime lights and vegetation coverage data, and correcting the initial allocation result to obtain a population distribution correction result; Step S5: performing zoning statistics on the population distribution correction result according to the administrative division boundaries of the same level in the administrative division boundaries, calculating the population distribution fitting coefficient, iteratively correcting the allocation error in the population distribution correction result, and obtaining the population distribution correction result; Step S6: Perform zoning statistics on the population distribution correction results according to the secondary administrative division boundaries in the administrative division boundary vector data, and calculate the relative error between the initial value of the administrative division population and the statistical population value of each division. If the absolute value of the relative error value does not exceed the preset value, the spatial matching is completed. If the absolute value of the relative error value exceeds the preset value, the secondary administrative division is recorded as the area to be corrected and step S5 is executed. Further iterative correction is performed until the absolute value of the relative error value does not exceed the preset value, thereby completing the final spatial matching of residential buildings and population distribution.

2. The method according to claim 1, characterized in that: The step S2 comprises: Step S21: pre-processing the residential map patch data and the house building patch data to ensure that the spatial reference systems of the two are consistent and the integrity of the residential house building patch attribute table is guaranteed, and then the spatial matching rules are specifically implemented: if the house building patch is completely located in the residential map patch, the house building patch is extracted and recorded as a residential house building; if the residential map patch is completely located inside the house building patch, the residential map patch is extracted and given a house building attribute feature; if the residential map patch intersects with the house building patch, the house building patch intersecting with the residential land is extracted, and the integrity of the intersecting patch is retained; if the residential building cannot be matched to the house building type feature, the random forest algorithm is used to infer the house building type attribute feature of the residential map patch; Step S22: Integrate the residential building patch data. According to the spatial matching rules, merge all the identified and extracted residential building patch data into a unified layer. If attribute information is missing in the layer, supplement the corresponding patch attribute information according to the spatial matching rules.

3. The method according to claim 2, characterized in that: The step S3 comprises: Step S31: Associating residential buildings with administrative division information, spatially superimposing the residential buildings with the administrative division boundary layer, and adding the relevant attribute information of the administrative division boundary to the residential building patch attribute table; Step S32: converting the residential building pattern with building attribute characteristics into centroid data, classifying the centroid data according to the building type information, saving each classified centroid data as an independent layer, and retaining the area value of the original residential building pattern in the layer attribute table; Step S33: using the spatial position coordinates of the centroid data and combining the corresponding residential building area as weights, perform spatial kernel density estimation to generate kernel density distribution raster data of various types of residential buildings; Step S34: for the kernel density distribution grid data of various types of residential buildings, the data with a kernel density value of 0 is classified into the first level, and the data with a kernel density value greater than 0 is classified using the natural breakpoint method; Step S35: constructing a feature vector based on the scale of the patch, and constructing an administrative division scale feature vector and a residential building feature vector according to the kernel density distribution level of the classified residential building patch; Step S36: inputting the residential building feature vector, administrative division scale feature vector and administrative division demographic data as training samples into the random forest model to select the optimal parameters; Step S37: Based on the statistical characteristics of the residential building pattern, the initial value of the residential building population is estimated through a random forest model with optimal parameters to obtain an initial allocation result.

4. The method according to claim 3, characterized in that: The step S4 comprises: Step S41: constructing a population geographic footprint index by integrating DEM, public service facilities, road network, water network, nighttime lighting and vegetation coverage data, and correcting the initial population of the area to be corrected. The correction steps are as follows: Step S42: Calculate the single factor weights according to the DEM, public service facilities, road network, water network, night light and vegetation coverage data and generate single factor grid distribution data respectively, wherein the single factor grid distribution data includes DEM single factor grid data, public service facilities single factor grid data, road network single factor grid data, water network single factor grid data, night light single factor grid data and vegetation coverage single factor grid data; Step S43: performing equal-weighted addition on the DEM single-factor raster data, the public service facility single-factor raster data, the road network single-factor raster data, the water network single-factor raster data, the night light single-factor raster data and the vegetation coverage single-factor raster data to generate a population geographic imprint index; Step S44: Use the population geographic imprint index to correct the initial distribution result extracted in step S37, and the correction formula is as follows: in, is the corrected population, is the population initially allocated to the classified residential buildings obtained in step S3, is the population geographic footprint index of grid j, It is the largest population geographic imprint index among all calibration grids. is an adjustment parameter that controls the impact of the population distribution indicator weight value on the correction result. A factor that stretches and limits the range of the correction weights.

5. The method according to claim 4, characterized in that: The step S5 comprises: The population distribution correction result is statistically divided according to the administrative division boundary of the same level in the administrative division boundary vector data, and the cumulative value of all grid population correction values ​​in the administrative division of this level is summarized and compared with the original population statistics of the administrative division. If the two are not equal, the population distribution fitting coefficient is calculated, that is, the ratio of the administrative division population statistics to the grid population cumulative value. The calculation formula is as follows: Among them, F represents the population distribution fitting coefficient, represents the original population statistics of administrative divisions, It represents the sum of the population values ​​of all grids in the administrative area; Correct the grid population, divide the population of each grid by the population distribution fitting coefficient, and get the corrected population of each grid. The formula is as follows: in, represents the corrected grid population, It represents the sum of the population values ​​of all grids in the administrative area; Repeat the correction process for all grids and accumulate the corrected population numbers to check whether they are equal to the original population statistics of the administrative division. If the two are not equal, continue to iterate until the two are equal to obtain the population distribution correction result. The specific calculation process is: in, Represents the sum of the population numbers after correction for all grids.

6. The method according to claim 5, characterized in that: The step S6 comprises: Calculate the population distribution fitting coefficient of the secondary administrative divisions and combine it with the zoning adjustment strategy to continue to correct the grid population: Based on the sub-administrative division boundaries, the grid population values ​​in each sub-administrative division are counted, and the relative error between the grid population cumulative value and the original population statistical value of each sub-administrative division is calculated. The sub-administrative divisions with relative error absolute values ​​greater than the preset value are marked as areas to be adjusted. The calculation formula is as follows: Where E represents the relative error, It represents the cumulative sum of the population values ​​of the grids within the secondary administrative division. Indicates the original population statistics of administrative divisions; For the area to be adjusted, the sub-administrative division boundary is spatially matched with the grid to determine the region to which each grid belongs; Calculate the population adjustment coefficient within the area to be adjusted, that is, the ratio of the population correction value of each sub-administrative area to the population statistical value, and count the population adjustment coefficient of each sub-administrative area within the preset value as 1; then record the ratio of the population correction value of each grid to the population adjustment coefficient as the population number after adjustment of each grid, and take the integer value of the population number after adjustment. The calculation formula is as follows: in, is the population after grid adjustment, is the population modifier of the sub-district, is the corrected grid population, is the original population statistics of administrative divisions; Summarize and calculate the difference in population numbers, calculate the difference between the population value of all grids of the secondary administrative divisions and the demographic value, and further correct the population value of each grid. The calculation formula is as follows: in, represents the sum of the population values ​​of all grids in the administrative area. is the original population statistics of the administrative division; if the difference is greater than 0, all grid population values ​​in the administrative division are sorted from large to small, and 1 is subtracted in sequence until the difference adjustment is 0; if the difference is less than 0, all grid population values ​​in the administrative division are sorted from large to small, and 1 is added in sequence until the difference adjustment is 0; The total population after adjustment within the area is further summarized and compared with the actual total population to calculate the relative error of the sub-area summary. The calculation is shown as follows: in, is the relative error after partition aggregation, is the total population after adjustment, is the actual population number; if the population relative error is still greater than the preset value, the iterative operation is repeated until the accuracy requirement is met.

7. The method according to claim 6, characterized in that: For the area to be adjusted, the secondary administrative division boundary is spatially matched with the grid to determine the region to which each grid belongs, including: Check whether the spatial location of each grid is completely located within a certain secondary administrative division. If the grid crosses multiple administrative division boundaries, it is necessary to calculate the area occupied by the residential building pattern in the grid within this administrative division. For grids that cross multiple administrative divisions, use the following formula to calculate the proportion of the residential building pattern in each administrative division; in, represents the proportion of the residential building area of ​​the i-th administrative division in the grid, represents the residential building area of ​​the i-th administrative division within the grid, Indicates the area of ​​residential buildings within the grid; Determine the administrative division with the largest residential building area within the grid as the administrative division to which the grid belongs.

8. A spatial matching system based on natural resource monitoring and demographic data, characterized by: It includes a data extraction module, a patch data matching module, an initial population division module, a verification module, a population correction module and a final matching module; The data extraction module is used to execute step S1: extract residential map data, building map data, water network, road network, public service facilities, administrative division boundaries and DEM data from natural resource survey and monitoring data, and obtain demographic data, OSM road network, POI, night light and vegetation coverage data from the Internet; The patch data matching module is used to execute step S2: using spatial position and attribute matching, combined with a random forest algorithm, to spatially match the residential map patch data and the building patch data, and identify and extract the residential map patch data having building attribute characteristics; The initial population division module is used to execute step S3: integrating the residential map patch data with housing building attribute characteristics and associating the residential buildings in the data with administrative division information, creating an attribute feature mapping of the residential map patch data with housing building attribute characteristics and the administrative division after association, and using the random forest algorithm and the demographic data to allocate initial population numbers for different types of residential buildings to obtain an initial allocation result; The population correction module is used to execute step S4: constructing a population geographic imprint index by integrating multi-source data in the natural resource survey and monitoring data, correcting the initial distribution result, and obtaining a population distribution correction result; The verification module is used to execute step S5: perform zoning statistics on the population distribution correction result according to the administrative division boundary lines of the same level in the administrative division boundary vector data, calculate the population distribution fitting coefficient, iteratively correct the allocation error in the population distribution correction result, and obtain the population distribution correction result; The final matching module is used to execute step S6: perform zoning statistics on the population distribution correction results according to the secondary administrative division boundaries in the administrative division boundary vector data, calculate the relative error between the initial value of the administrative division population of each partition and the statistical population value, if the absolute value of the relative error value does not exceed the preset value, then complete this spatial matching, if the absolute value of the relative error value exceeds the preset value, record the secondary administrative division as a to-be-corrected area and execute step S5, further iterate and correct until the absolute value of the relative error value does not exceed the preset value, thereby completing the final spatial matching of residential buildings and population distribution.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images