Land surveying and mapping data analysis system based on multi-source data fusion

Abstract

The invention relates to the technical field of data fusion, in particular to a land surveying and mapping data analysis system based on multi-source data fusion, which comprises a grid unification module, a fluctuation extraction module, a data rejection module, a region compression module and a fusion analysis module. According to the method, a space unit consistency basis is constructed through unified serial number comparison, multi-source measurement value serial number coordination is completed in combination with spatial position residual difference, stable serial numbers meeting the vertical precision requirement are recognized in a difference fluctuation extraction mode, and data availability is quantitatively judged through numerical value deviation frequency in the serial numbers and an abnormal proportion. Eliminating a low-roughness region based on region standard deviation change to effectively compress redundant space paragraphs, realizing terrain elevation information fusion and land change analysis through weighted calculation of a space measurement value, and realizing fusion expression of a land surveying and mapping result; and the measurement precision consistency, the spatial feature expression capability and the change trend identification effect under multi-source topographic data fusion are effectively improved.

Description

Technical Field

[0001] This invention relates to the field of data fusion technology, and in particular to a land surveying data analysis system based on multi-source data fusion. Background Technology

[0002] Data fusion technology involves coordinating data from multiple sources in terms of time, space, or attributes to form a consistent, accurate, and complete comprehensive information output. This includes the acquisition, preprocessing, registration, integrated processing of registered information, and maintenance of data consistency. It is widely used in various scenarios such as remote sensing monitoring, intelligent transportation, environmental assessment, and geographic information systems. Traditional land surveying refers to the process of extracting land feature boundaries and analyzing terrain using a single or limited data source, such as total station measurements, GPS positioning, or aerial remote sensing images. It involves unified analysis and modeling of surface spatial information to assist in land use status surveys and spatial change identification. Typically, it uses methods such as manual interpretation of remote sensing images combined with on-site sampling point surveys to obtain the information required for surveying, and then uses image processing tools to complete tasks such as target extraction and attribute assignment.

[0003] In the current land surveying data analysis process, a single or limited number of data sources are used as the source of surveying information. The surveying model is formed by relying on manual image interpretation and on-site sampling. Due to the limited data coverage and human identification errors, it is difficult to achieve high-precision modeling in complex terrain or variable areas. Especially in areas where the distribution of land features is uneven or where spatial changes are drastic, there are problems such as missing measurement information, ambiguous attribute assignments, and inconsistent boundary extraction. This can easily lead to a decrease in the accuracy of surveying results and misjudgment of land change identification, thus limiting the applicability of data consistency maintenance and comprehensive information integration and analysis. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of existing technologies and propose a land surveying data analysis system based on multi-source data fusion.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: a land surveying data analysis system based on multi-source data fusion includes: The raster unification module acquires multi-source terrain data, establishes a unified UTM grid code, confirms the consistency of spatial units corresponding to the numbering by combining spatial location coordinates, and generates a multi-source numbering unified control group. The fluctuation extraction module extracts spatial measurement values, calculates the range, and obtains the maximum difference sequence based on the multi-source numbered unified control group. It then compares the difference with the vertical accuracy level of the DEM and filters the stable unit number group to obtain the spatial fluctuation stable number set. The data elimination module counts the frequency of numerical deviations of each data source based on the spatial fluctuation stability number set, calculates the abnormal proportion value of each data source, and eliminates data sources whose proportion value exceeds the horizontal positioning accuracy to obtain the distribution record of available sources participating in the fusion. The regional compression module extracts the numerical sequence of available data sources in a continuous spatial region based on the distribution records of available sources participating in the fusion, calculates the standard deviation within the continuous region, removes regions with a standard deviation lower than the terrain roughness threshold, and obtains regional information cropped data.

[0006] As a further aspect of the present invention, the multi-source numbering unified control group includes numbering mapping relationship, spatial location correspondence relationship, and coordinate system unified coding; the spatial fluctuation stable numbering set includes precision conformity numbering set, spatial stability identifier, and difference threshold judgment basis; and the available source distribution record participating in fusion includes available data source identifier, distribution information within the numbering group, and source removal statistics.

[0007] As a further aspect of the present invention, the grid unification module includes: The number extraction submodule obtains the numbering information of multi-source terrain data, extracts and collects the numbered data, lists the number sets according to the data sources, matches each number set with spatial location coordinates, and generates number coordinate combination results. The intersection matching submodule compares the spatial location coordinates of the same number in different data sources based on the number coordinate combination results. It calculates the spatial residual of each coordinate group through unified projection transformation, determines whether the residual is lower than the spatial reference system offset threshold, and counts the number combinations that meet the conditions to obtain the consistent number matching comparison results. The grid coding construction submodule, based on the consistent number matching comparison results, divides the corresponding number combinations into UTM grids, generates unified grid unit numbers according to the spatial boundary and coordinate range of the area to which the number belongs, and generates a unified control group of multi-source numbers.

[0008] As a further aspect of the present invention, the spatial residuals of each coordinate group are implemented using the formula: ; Calculations are performed, in which, The representative number index is Spatial location residual value, The representative number index is The x-coordinate value extracted from data source A. The representative number index is The x-coordinate value extracted from data source B. The representative number index is The ordinate value extracted from data source A. The representative number index is The ordinate value extracted from data source B. This represents the sum of the differences in the x-coordinates of all numbers from item 1 to item N between data source A and data source B. This represents the sum of the differences in the ordinates of all numbers from the 1st to the Nth item between data source A and data source B. This indicates the total number of numbers involved in the calculation of spatial location residuals.

[0009] As a further aspect of the present invention, the fluctuation extraction module includes: The measurement extraction submodule obtains the spatial location corresponding to each group number in the multi-source numbered unified control group, collects the spatial measurement elevation values ​​of all data sources at each location, records the corresponding number and source identifier, collects them according to the number and establishes a measurement sequence to generate a spatial measurement set; The difference operation submodule calculates the numerical difference between the maximum and minimum values ​​in each group based on the numbered measurement sequence in the spatial measurement set, extracts the range values ​​under all numbers and constructs the maximum difference sequence to obtain the numbered range sequence. The stability marking submodule, based on the range value corresponding to each number in the numbered range sequence, and according to the DEM vertical accuracy level benchmark value, discriminates and marks the range value, classifies the numbers with values ​​less than or equal to the benchmark value into stable unit number groups, and obtains the spatial fluctuation stable number set.

[0010] As a further aspect of the present invention, the data removal module includes: The frequency statistics submodule extracts the spatial measurement elevation values ​​of each data source in the consecutive numbered group based on the spatial fluctuation stability number set, accumulates the number of deviation records, counts the deviation frequency of each data source in each numbered group, and generates source deviation frequency data. The ratio judgment submodule calculates the ratio of the number of deviations of each data source in the source deviation frequency data to the total number of corresponding numbers, obtains the abnormal ratio value of each data source in each number group, compares the ratio value with the set horizontal positioning accuracy threshold, obtains the list of data sources marked as unavailable, and obtains the unavailable source number record. The source removal submodule removes the corresponding data records of the data sources in each number group according to the unusable source number records, retains the remaining usable source records, and collects and maps the numbers, locations and spatial measurement elevation values ​​of the usable data sources with number identifiers to obtain the distribution records of usable sources participating in the fusion.

[0011] As a further aspect of the present invention, the region compression module includes: The sequence extraction submodule extracts available data source measurement values ​​from each group of continuous spatial regions based on the distribution records of available sources participating in the fusion, collects spatial measurement elevation values ​​by region number, organizes the number sequence and corresponding values ​​to form a numerical sequence set, and generates a continuous region numerical sequence. The standard deviation calculation submodule calculates the standard deviation of the measured values ​​under each group number based on the continuous region numerical sequence, extracts the standard deviation value of each number region, establishes a mapping list between the number and the corresponding standard deviation, and obtains the regional standard deviation set. The region clipping submodule compares the standard deviation corresponding to each number in the region standard deviation set with the terrain roughness threshold, filters out regions with standard deviations below the threshold and records the corresponding numbers, summarizes the remaining region numbers and numerical information, and obtains region information clipping data.

[0012] As a further aspect of the present invention, the system further includes: The fusion analysis module clips data based on the regional information, aggregates regional spatial measurement values, calculates weighted average elevation values, and performs land area comparative analysis to generate fusion analysis results of multi-source land surveying data. The regional information cropping data includes the cropped region number, terrain roughness classification identifier, and spatial continuous region range.

[0013] As a further aspect of the present invention, the fusion analysis module includes: The measurement summary submodule extracts all available spatial measurement elevation values ​​from each available data source under each region number based on the region information cropping data, integrates the measurement values ​​and corresponding source weight information of each data source according to the number, forms a combination sequence of number and weight measurement values, and generates a measurement weight sequence set; The weighted calculation submodule calculates the weighted average elevation value within each group of numbers based on the measured values ​​and weights of all data sources under each number in the measured value weight sequence set, extracts the weighted average value of each number and organizes it into a number index list to obtain a set of weighted elevation values. The regional comparison submodule assigns the weighted average elevation value of each number in the weighted elevation value set to the corresponding region according to the land parcel division number, compares the weighted elevation difference range corresponding to the number in each region, summarizes the elevation difference results between regions, and obtains the multi-source data fusion analysis results of land surveying.

[0014] As a further aspect of the present invention, the weighted average elevation value is calculated using the following formula: ; Calculations are performed, in which, Representative number The weighted average elevation value below, Representative number The corresponding number Elevation measurements from one data source Representative number Next The weight values ​​assigned to each data source. Representative number The total number of available data sources included below. The index number of the data source. An identifier index for space numbering.

[0015] Compared with the prior art, the advantages and positive effects of the present invention are as follows: In this invention, a spatial unit consistency foundation is established through unified numbering and comparison. The numbering coordination of multi-source measurement values ​​is completed by combining spatial location residual differences. Stable numbers that meet vertical accuracy requirements are identified by the difference fluctuation extraction method. Data availability is judged by the frequency of numerical deviation and the proportion of anomalies within the number. Low-roughness areas are eliminated based on regional standard deviation changes to effectively compress redundant spatial segments. Topographic elevation information fusion and land change analysis are achieved through weighted calculation of spatial measurement values. This enables the accurate establishment of spatial unit correspondences across source data, the quantitative elimination of abnormal measurement data, the dynamic trimming of regional data quality, and the fusion expression of land surveying results. This effectively improves the measurement accuracy consistency, spatial feature expression ability, and trend identification effect under multi-source topographic data fusion. Attached Figure Description

[0016] Figure 1 This is a system flowchart of the present invention; Figure 2 This is a system module diagram of the present invention. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0018] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0019] Please see Figure 1 and Figure 2 A land surveying data analysis system based on multi-source data fusion includes: The raster unification module acquires the numbering information from multi-source terrain data, performs the intersection of number sets and establishes a unified UTM grid code (using EPSG: 32600 series coordinate system). Combined with the spatial location coordinates of each data source under each number, the consistency of the spatial units corresponding to the number is confirmed by comparing the projection transformation residuals (conforming to the spatial reference system standard) and a multi-source numbering unification control group is generated. The fluctuation extraction module extracts spatial measurement values ​​within each control group based on the multi-source numbered unified control group, calculates the range and obtains the maximum difference sequence, compares the difference with the DEM vertical accuracy level (refer to the set standard ≤10cm), and marks the numbers that meet the DEM vertical accuracy level as stable unit number groups, thus obtaining the spatial fluctuation stable number set; The data removal module counts the frequency of numerical deviation of each data source in the consecutive numbered group based on the spatial fluctuation stability number set, calculates the abnormal proportion value of each data source in the numbered group, and marks the data source with the proportion value exceeding the horizontal positioning accuracy as an unusable data source and removes it to obtain the distribution record of available sources participating in the fusion. The regional compression module extracts the numerical sequence of available data sources in a continuous spatial region based on the distribution records of available sources participating in the fusion, calculates the standard deviation of each continuous region, removes regions with a standard deviation lower than the terrain roughness threshold (using the elevation change rate index in the data quality elements), records the corresponding region number distribution, and obtains regional information to prune the data. The fusion analysis module trims data based on regional information, aggregates spatial measurement values ​​from all available data sources in the trimmed region, calculates the weighted average elevation value for each group number (in accordance with the "Geospatial Data Fusion Specification"), and uses the weighted average elevation value to conduct comparative analysis of land areas, generating fusion analysis results of multi-source land surveying data.

[0020] The multi-source numbering unified control group includes numbering mapping relationships, spatial location correspondence relationships, and coordinate system unified coding. The spatial fluctuation stable numbering set includes the accuracy conforming numbering set, spatial stability identifier, and difference threshold judgment criteria. The distribution records of available sources participating in the fusion include available data source identifiers, distribution information within the numbering group, and statistical information of removed sources. The regional information cropped data includes cropped regional numbers, terrain roughness classification identifiers, and spatial continuous area range. The multi-source data fusion analysis results of land surveying include weighted elevation results, regional fusion layers, and land comparison indicators.

[0021] Please see Figure 2 The grid unification module includes: The number extraction submodule obtains the numbering information of multi-source terrain data, extracts and collects the numbered data, lists the number sets according to the data sources, matches each number set with spatial location coordinates, and generates number coordinate combination results. The number extraction submodule acquires multi-source terrain data numbering information. First, it needs to clarify the source type and format of the multi-source terrain data, such as data from UAV aerial photography, elevation DEM, and LiDAR point cloud data. It then extracts unique numbering fields (map sheet number, layer ID, data block number, etc.) from each source data by reading its metadata, serving as the basis for number extraction. Next, it identifies the data type in the extracted numbering fields. For example, strings need to be converted to a unified encoding format (UTF-8) to ensure compatibility, while numeric numbers are uniformly converted to integers and then aggregated according to data source category. By traversing each data source numbering set, its number is spatially indexed and bound to the corresponding map sheet boundary, which can be done using coordinate files (such as the accompanying GeoTransform information in GeoTIFF). Alternatively, extract the coordinates (X0, Y0) of the top left corner of the map sheet and the resolution (Δx, Δy) from the projected coordinate system file, and calculate the spatial boundary range corresponding to each number. For example, for a number 1234, the coordinates of the top left corner of the map sheet are (500000, 4200000), the resolution is 30 meters, and the map sheet size is 1000x1000 pixels. Then the coordinates of the bottom right corner are (500000+30×1000, 4200000−30×1000) = (530000, 4170000). Finally, the number 1234 and its map sheet boundary coordinates are combined into a one-to-one correspondence (1234, (500000, 4200000, 530000, 4170000)). After all the number sets are constructed, the output is the combination result of the number and the coordinate range.

[0022] The intersection matching submodule compares the spatial location coordinates of the same number in different data sources based on the number coordinate combination results. It calculates the spatial residual of each coordinate group through unified projection transformation, determines whether the residual is lower than the spatial reference system offset threshold, and counts the number combinations that meet the conditions to obtain the consistent number matching comparison results. Spatial residuals are applied to each coordinate group using the following formula: ; Calculations are performed, in which, The representative number index is Spatial location residual value, The representative number index is The x-coordinate value extracted from data source A. The representative number index is The x-coordinate value extracted from data source B. The representative number index is The ordinate value extracted from data source A. The representative number index is The ordinate value extracted from data source B. This represents the sum of the differences in the x-coordinates of all numbers from item 1 to item N between data source A and data source B. This represents the sum of the differences in the ordinates of all numbers from the 1st to the Nth item between data source A and data source B. This indicates the total number of numbers involved in the calculation of spatial location residuals.

[0023] The coordinates of the top left corner of image number 1234 in the image are: , The image size is 1000×1000 pixels, with a resolution of 30 meters. The boundary calculation method is derived based on the GeoTransform matrix in the image metadata. The coordinates of the lower right corner of the image are: ; ; The actual matched image number comes from three data sources: LiDAR, DEM, and UAV remote sensing imagery. The image sheet number is 1234, and its coordinates in the three data types are as follows: Data source A (LiDAR): Top left corner coordinates , ; Data source B (DEM): Top left corner coordinates , ; Data source C (drone imagery): Top left corner coordinates , , , ; Data source D: , , , ; Set the number of the number set to Calculate the differences of each item step by step and sum them: ; ; Average difference: ; ; Substitute the spatial residual values ​​of numbers 1234 (corresponding to k=1) into the calculation: ; The results show that the spatial deviation of the upper left corner coordinates of the corresponding map in the lidar and DEM source for number 1234 is 5.58 meters, which exceeds the 5-meter threshold of the conventional spatial reference system offset tolerance. Therefore, number 1234 does not meet the matching condition and needs to be excluded in the subsequent grid coding step.

[0024] Spatial location residuals represent the actual geometric offset between the spatial coordinates of the same map sheet numbered in different data sources within multi-source topographic data. It is a quantitative expression of the positional differences of two or more spatial entities corresponding to the same map sheet number under the same index. This value reflects the positioning offset caused by factors such as the accuracy of the acquisition equipment, projection transformation errors, inconsistencies in coordinate systems, or differences in map sheet boundary definitions across different data sources, and is measured in meters. A smaller residual value indicates a more accurate match between the map sheet numbers across multiple data sources, with a higher degree of coordinate overlap. A larger residual value indicates significant spatial inconsistencies between data sources, affecting subsequent map sheet registration, unified grid coding, or spatial overlay analysis. Therefore, this indicator is a key reference for judging spatial consistency of map sheet numbers, establishing a multi-source unified grid comparison system, and improving the accuracy of data fusion.

[0025] The formula's calculation logic is based on the quantification of spatial geometric differences in residual offsets. Each coordinate difference term in the formula... and These represent the differences in single-point coordinates of the two data sources in the horizontal and vertical directions, reflecting the original point-to-point displacement; an overall mean difference term is introduced. and The aim is to enhance the ability to identify local deviations by weighting or differentiating the single-point offset of each numbered position with the overall system offset trend of the entire numbered set. Addition and subtraction operations are used to emphasize the cumulative trend or relative elimination effect of local deviations under system offset, respectively; squaring is used to unify the offsets in the positive and negative directions into positive values ​​and amplify the influence of larger deviation points; finally, the Euclidean distance in two-dimensional space is calculated by taking the square root of the sum of squares, ensuring that the horizontal and vertical coordinate offsets are considered together and the output result is in true spatial distance units, forming a comprehensive spatial location residual value as a metric for consistency judgment.

[0026] The grid coding construction submodule, based on the consistent numbering matching comparison results, combines the corresponding numbers in the EPSG:32600 series coordinate system to perform UTM grid division, generates unified grid unit numbers according to the spatial boundary and coordinate range of the area to which the number belongs, and generates a unified control group of multi-source numbers; The grid coding construction submodule, based on the consistent numbering matching results, divides each matching number combination into a UTM grid according to the EPSG:32600 series coordinate system. First, it determines the UTM zone number based on the coordinate boundary of the corresponding area (e.g., if 117°E falls in zone 50, then EPSG:32650 is used). Next, it divides the spatial boundary coordinates of each number group according to a fixed grid size (e.g., 1km × 1km), using the upper left corner of the boundary as the starting point and numbered by row and column. Let the upper left corner coordinates be (Xmin, Ymax) = (600000, 4300000), then the first grid cell is numbered Grid_600_4300. Moving to the right, each 1km increments the column number by 1, and moving downwards, each 1km decrements the row number by 1. For example, the lower right corner coordinates are (604000, 4...). If the grid number is 296000, then the grid numbering within the grid ranges from Grid_600_4300 to Grid_604_4296. Subsequently, for each matching number, add the set of all corresponding grid numbers to construct a unified grid cell numbering lookup table to achieve full coverage pairing between number combinations and UTM grids. For example, if the coverage area of ​​number 8765 is (600000, 4300000) - (602000, 4298000), then it contains grid cells Grid_600_4300, Grid_601_4300, Grid_600_4299, Grid_601_4299, Grid_600_4298, and Grid_601_4298. Finally, number 8765 corresponds to 6 UTM grid numbers, forming a unified control group for multi-source numbering.

[0027] Please see Figure 2 The fluctuation extraction module includes: The measurement extraction submodule obtains the spatial location corresponding to each group number in the multi-source numbered unified control group, collects the spatial measurement elevation values ​​of all data sources at each location, records the corresponding number and source identifier, collects them according to the number and establishes a measurement sequence to generate a spatial measurement set; The measurement extraction submodule obtains the spatial location corresponding to each group number in the multi-source unified control group. First, it retrieves and parses the coordinate range corresponding to each number in the unified control group, using its center point or representative coordinates as the reference point for measurement extraction. Then, for this spatial location, it iterates through all record units with measurement information in all data sources, extracting the elevation value of that location from each data source using the number as an index. Image or raster values ​​are then extracted via data file reading. For example, the center point coordinates of number 9012 are (501000, 4210000). In source A, its corresponding DEM raster value is 243.2 meters, while in source B... The measured value is 242.7 meters in source C and 243.4 meters in source C. Therefore, the three values ​​are recorded as (9012, A, 243.2), (9012, B, 242.7), and (9012, C, 243.4) respectively, based on the number and source identifier. This process is repeated to group all measured values ​​for that number according to their source, forming a measured value sequence [243.2, 242.7, 243.4]. This process continues to iterate through all unified number groups, constructing a complete list of corresponding measured values. Using the number as the key and the measured value sequence as the value, a key-value mapping structure for the spatial measured value set is established, enabling the collection and structured management of measured values ​​for all numbers in multi-source data.

[0028] The difference operation submodule calculates the numerical difference between the maximum and minimum values ​​in each group based on the sequence of numbered measurements in the spatial measurement set, extracts the range values ​​under all numbers and constructs the maximum difference sequence to obtain the numbered range sequence; The difference calculation submodule performs maximum and minimum value extraction on each set of measured values ​​based on the numbered measured value sequences in the spatial measured value set. It extracts endpoint data by searching for the maximum and minimum values ​​in each set of sequences, defining the maximum measured value as Maxᵢ and the minimum measured value as Minᵢ in each set, and calculating the difference ΔHᵢ = Maxᵢ − Minᵢ. For example, the measured value sequence corresponding to number 9012 is [243.2, 242.7, 243.4], then Max = 243.4, Min = 242.7, and ΔH = 0.7 meters. This process is performed on all measured value sequences with numbers in sequence to construct the range results corresponding to all numbers. The mapping relationship between the number and the ΔH value is recorded using a list or array structure, such as (9012, 0.7), (9013, 0.15), (9014, 0.42), etc. Then, all ΔH values ​​are arranged in numerical order to form a range sequence, which can be used for subsequent fluctuation range judgment and number classification operations.

[0029] The stability marking submodule is based on the range value corresponding to each number in the number range sequence. According to the DEM vertical accuracy level benchmark value of 10 cm, the threshold is set to distinguish and mark the range value. The numbers with values ​​less than or equal to the benchmark value are classified into stable unit number groups, thus obtaining the spatial fluctuation stable number set. The stability labeling submodule classifies each ΔH value based on the range value corresponding to each number in the range sequence. The discrimination benchmark value is set to 0.10 meters, which comes from the common vertical accuracy requirements of DEM data and is set in combination with the average elevation accuracy index of the 1:5000 mapping standard as the basis for stability screening. All numbers are sequentially traversed for their ΔH values. If ΔHᵢ≤0.10, the number ᵢ is included in the stable unit number group; otherwise, it is discarded. For example, the range corresponding to number 9013 is 0.15>0.10, which does not meet the requirement, so it is not selected. The range of number 9015 is 0.08 meters≤0.10, so this number is identified as a stable number and added to the number set result. Finally, the discrimination process classifies all numbers to generate a spatial fluctuation stable number set.

[0030] Please see Figure 2 The data removal module includes: The frequency statistics submodule extracts the spatial measurement elevation values ​​of each data source in the consecutive numbered group based on the spatial fluctuation stability number set, accumulates the number of deviation records, counts the deviation frequency of each data source in each numbered group, and generates source deviation frequency data. The frequency statistics submodule extracts spatial measurement elevation values ​​from each data source in a consecutively numbered group based on the stable spatial fluctuation number set. First, it extracts the corresponding spatial coordinates and elevation measurements from each data source for each stable number group. The measurements from each data source are then structured and grouped according to their number and source. For example, the number group [1001, 1002, 1003] has measurement sets A: [245.1, 245.3, 245.2], B: [244.9, 245.0, 244.8], and C: [246.2, 247.0, 246.5] under source A, source B, and source C, respectively. Next, the median value Mᵢ is calculated for each measurement sequence according to its number as the baseline value. For example, 10 The measured values ​​of the three sources with ID 01 are 245.1, 244.9, and 246.2. After sorting, the median value is 245.1. Then, the deviation value of each source measured value from the baseline value Mᵢ is calculated as Dᵢⱼ=|Hᵢⱼ−Mᵢ|. The deviation status is determined by setting the deviation judgment limit to 0.3 meters. If Dᵢⱼ>0.3, it is counted as one deviation. For example, in ID 1001, the measured value of source C, 246.2, is 1.1 different from Mᵢ=245.1, which is greater than 0.3, so it is a deviation. The deviation value of source B is 0.2, which is not exceeded, so it is not counted. After accumulating all IDs, the number of deviations is 0 for source A, 1 for source B, and 3 for source C. Finally, the total number of deviations is counted according to the data source and output as source deviation frequency data.

[0031] The ratio judgment submodule calculates the ratio of the number of deviations of each data source to the total number of corresponding numbers in the source deviation frequency data to obtain the abnormal ratio value of each data source in each number group. The ratio value is compared with the set horizontal positioning accuracy threshold of 15% to obtain a list of data sources marked as unavailable and obtain the unavailable source number record. The ratio judgment submodule calculates the ratio of the number of deviations of each data source to the total number of corresponding numbers in the source deviation frequency data. First, it reads the total number of deviations Nᵢ and the total number of numbers T for each source in the number group, and calculates the abnormal ratio Rᵢ=Nᵢ / T. For example, if there are 20 number groups and source C has 5 deviations, then R_C=5 / 20=0.25. Then, it compares this ratio value with a set threshold. The horizontal positioning accuracy threshold is set to 15%, which is converted to a value of 0.15. All data sources with Rᵢ>0.15 need to be marked and their number records are added to the unusable source number list. For example, if R_A=0.05, R_B=0.10, and R_C=0.25, then only source C is marked as unusable. Finally, it obtains all number records involving number groups under this source to form an unusable source number record table, which is used for subsequent data removal and fusion processing.

[0032] The source removal submodule removes the corresponding data records of the data sources in each number group according to the unavailable source number records, retains the remaining available source records, and collects and maps the numbers, locations and spatial measurement elevation values ​​of the available data sources with number identifiers to obtain the distribution records of available sources participating in the fusion. The source removal submodule removes data records from the corresponding data sources in each number group based on the unavailable source number. First, it iterates through all numbers in the number group to identify whether there is a corresponding unavailable source number identifier. If there is, the data item of that source is deleted from its measurement record list. For example, the original records for number 1001 were A: 245.1, B: 244.9, and C: 246.2. After source C is marked as unavailable, the measurement values ​​of this group are updated to A: 245.1 and B: 244.9. The remaining source data is retained in the records. Then, the data structure is reconstructed for each remaining source number, recording its number, remaining available source identifier, and measurement value. For example, the output for number 1001 is (1001, A, 245.1) and (1001, B, 244.9). All number groups are processed in sequence. Finally, all available source records are summarized to form a number-source-elevation value triplet structure for fusion. An index mapping is established according to the number to generate the distribution records of available sources participating in fusion.

[0033] Please see Figure 2 The region compression module includes: The sequence extraction submodule extracts available data source measurement values ​​from each group of continuous spatial regions based on the distribution records of available sources participating in the fusion. It then collects spatial measurement elevation values ​​by region number, organizes the numbered sequence and corresponding values ​​to form a numerical sequence set, and generates a continuous regional numerical sequence. The sequence extraction submodule, based on the available source distribution records participating in the fusion, first groups spatial regions according to their serial numbers. Adjacent or spatially contiguous serial numbers are grouped into a continuous region. The continuity between serial numbers is determined by the difference in coordinate distance. For example, the top-left corner coordinates of serial numbers A_001 and A_002 are (500000, 4200000) and (500000, 4199000) respectively, with a vertical interval of 1000 meters. If the grid resolution is set to 1km, then the two serial numbers constitute a continuous region. Next, spatial elevation measurements are extracted sequentially from the available source records within this continuous region, and the elevation values ​​under each serial number are then sorted according to... Data sources are aggregated into a unified data structure within the region. For example, region R1 consists of A_001 to A_005, with each number having measured values ​​from sources A, B, and C, such as A_001=[245.0, 245.3, 245.1], A_002=[244.9, 245.2, 245.0], etc. Then, the numerical sequence of region R1 can be seen as an integration of the corresponding measured value sets under multiple numbers. The numbers and corresponding values ​​are combined into a list structure and named as a numerical sequence set. It is archived separately by region, and finally, a continuous regional numerical sequence is constructed with the region number as the key and the set of elevation measurements under its corresponding consecutive numbers as the value.

[0034] The standard deviation calculation submodule calculates the standard deviation of the measured values ​​under each group number based on the continuous regional numerical sequence, extracts the standard deviation values ​​of each numbered region, establishes a mapping list between the number and the corresponding standard deviation, and obtains the regional standard deviation set. The standard deviation calculation submodule calculates the standard deviation of the measured values ​​under each group number based on the continuous regional numerical sequence. First, it iterates through the set of numbers for each region, calling its spatial elevation measurement set for each number. For example, if number A_001 has measured values ​​[245.0, 245.3, 245.1] under sources A, B, and C, then the mean of this set is calculated as μ = (245.0 + 245.3 + 245.1) / 3 = 245.13. Next, the squared difference between each value and the mean is calculated: (245.0 − 245.13)² + (245.3 − 245.13)² + (245.1 − 245.13)² ≈ 0.0169 + 0.0289 + 0.0009 = 0.0467. Finally, the average and square root of the above sums are taken to obtain the standard deviation σ = ... ≈0.125. Using this method, the standard deviation of each number in the region is extracted, and a one-to-one correspondence structure list between the number and the standard deviation is constructed. The list format is as follows: [(A_001, 0.125), (A_002, 0.087), (A_003, 0.054)]. Finally, the standard deviations of all numbered regions are summarized to form a set of regional standard deviations.

[0035] The region clipping submodule compares the standard deviation of each number in the region standard deviation set with the terrain roughness threshold of 5 cm, filters out the numbered regions with standard deviations below the threshold and records the corresponding numbers, summarizes the remaining region numbers and numerical information, and obtains the region information clipping data. The regional cropping submodule compares the standard deviation of each number in the regional standard deviation set with the terrain roughness threshold of 5 cm. First, it performs a screening operation on each item in the standard deviation set, extracts the standard deviation value σᵢ corresponding to the number, and compares it with the set threshold σ0 = 0.05 meters. If σᵢ < σ0, it is considered that its spatial elevation change does not meet the characteristics of a rough region and should be removed. For example, the standard deviation value of number A_003 is 0.042, which is less than 0.05, so A_003 is marked as a removal number. The standard deviation value of number A_001 is 0.125, which is greater than 0.05, so it is retained. It iterates through all standard deviation records and performs judgment processing. The number records that meet the conditions are summarized into the removal number list, and the number records that do not meet the conditions are retained and a new number set is created. At the same time, the corresponding elevation value set is also organized and saved. The final output is regional information cropping data composed of the removal number records and the retained regional numbers and corresponding value information.

[0036] Please see Figure 2 The fusion analysis module includes: The measurement summary submodule extracts all available spatial measurement elevation values ​​from each available data source under each region number based on the region information cropping data, integrates the measurement values ​​and corresponding source weight information of each data source according to the number, forms a combination sequence of number and weight measurement values, and generates a measurement weight sequence set; The measurement aggregation submodule extracts all available spatial measurement elevation values ​​from data sources corresponding to each region number based on regional information cropping. First, using each region number as an index, it iterates through the spatial measurement record table corresponding to that number, extracting the elevation measurements and source identifiers of all data sources that were not excluded. An initial weight value is assigned to each data source based on its origin. This weight value is set according to the data source acquisition method and resolution. For example, for three types of data sources: Source A is an airborne lidar (resolution better than 0.3 meters), with a weight set to 0.5; Source B is an aerial image DEM (resolution 0.5 meters), with a weight set to... 0.3, Source C is a public topographic map raster DEM (1 meter resolution), with a weight set to 0.2. Each weight value satisfies the condition 0 < wᵢ < 1 and ∑wᵢ = 1. For number A_005, if the three source measurements are 246.1 (A), 245.9 (B), and 246.3 (C), the measurement weight combination sequence is [(246.1, 0.5), (245.9, 0.3), (246.3, 0.2)]. In the data structure, this sequence is a two-dimensional array consisting of the measurement value and the weight, with the number as the key and the value as the two-dimensional array. The above extraction is continuously performed on all numbers to finally form a measurement weight sequence set.

[0037] The weighted calculation submodule calculates the weighted average elevation value within each group of numbers based on the measured values ​​and weights of all data sources under each number in the measured weight sequence set, extracts the weighted average value of each number and organizes it into a number index list to obtain the weighted elevation value set. The weighted average elevation value is calculated using the following formula: ; Calculations are performed, in which, Representative number The weighted average elevation value below, Representative number The corresponding number Elevation measurements from one data source Representative number Next The weight values ​​assigned to each data source. Representative number The total number of available data sources included below. The index number of the data source. An identifier index for space numbering.

[0038] The elevation measurements of the three data sources under number A_005 are as follows (collected using spatial measurement equipment): Data source S1 (airborne lidar): Elevation measurement Meters, resolution 0.25 meters; Data source S2 (aerial image DEM): elevation measurement Meters, resolution 0.5 meters; Data source S3 (topographic raster DEM): Elevation measurements Meters, resolution 1.0 meter; Based on the data acquisition method and accuracy standards, the weights of the three types of data sources are set as follows: Based on the vertical accuracy of the lidar being better than ±0.15 meters; Based on the standard ±0.3 meter elevation accuracy of aerial imagery DEM; Based on the common error level of approximately ±0.5 meters for topographic raster DEM; Substitute the above elevation measurements and weights into the formula: ; Calculate the numerator: ; ; ; Sum of numerators: ; Sum of denominators: ; Weighted average elevation values: ; The results show that the weighted average elevation value of A_005 obtained by the fusion of three source data is 246.08 meters. This value will be used as the unique elevation feature value of A_005 in the number index list and will be used in subsequent analysis tasks such as the comparison of the maximum and minimum difference ΔH between numbers within the plot and the average value across plots.

[0039] The weighted average elevation value refers to the representative elevation value obtained by fusing elevation measurement results from different data sources under the same spatial number, according to the weights set by factors such as their acquisition accuracy, resolution, and reliability. This value comprehensively considers the differences in measurement quality and distribution density of multi-source data. By assigning higher weights to high-precision, low-error data, it makes them dominate the result, while retaining the spatial information and supplementary effects provided by other data sources. This results in a more balanced geographic elevation parameter in terms of data consistency, reliability, and spatial representation accuracy. This parameter can serve as the core input basis for subsequent regional analysis, parcel difference statistics, and terrain modeling.

[0040] The formula couples the elevation measurements from each data source with their importance in the fusion system. Each elevation measurement is multiplied by its corresponding weight parameter, representing the degree of influence of that measurement on the overall average. All weighted measurements are summed to form the contribution of the overall measurement. The denominator sums all weight values ​​as a normalization factor to constrain the final result to stay within the physical range of the measurements themselves. Finally, the numerator is divided by the denominator to express the fusion mean of all measurements under their respective weights, reflecting the spatial consistency fusion calculation of multi-source elevation data under unified numbering.

[0041] The regional comparison submodule assigns the weighted average elevation value of each number in the weighted elevation value set to the corresponding region according to the land parcel division number, compares the weighted elevation difference range corresponding to the number in each region, summarizes the elevation difference results between regions, and obtains the fusion analysis results of multi-source land surveying data. The regional comparison submodule, based on the land parcel division scheme, constructs a table corresponding to land parcel numbers and regions. Numbers A_001 to A_004 are assigned to parcel T1, and numbers A_005 to A_008 are assigned to parcel T2. The weighted average elevation values ​​for each number within parcel T1 are extracted: A_001: 245.12 meters, A_002: 244.88 meters, A_003: 245.20 meters, A_004: 244.97 meters. The maximum value within T1 is 245.20 meters, and the minimum value is 244.88 meters, with an elevation difference ΔH_T1 = 245.20 − 244.88 = 0.32 meters. The weighted average elevation value for number A_005 is 246.08 meters, and it is included in the regional elevation structure as the first elevation index value in the T2 region numbering. The remaining elevations in area T2 are A_006: 246.30 meters, A_007: 246.15 meters, and A_008: 245.97 meters. The maximum elevation within T2 is 246.30 meters, and the minimum is 245.97 meters. Therefore, the elevation difference is calculated as: ΔH_T2 = 246.30 − 245.97 = 0.33 meters. The average elevations within the two plots are calculated as follows: For area T1, the average is (245.12 + 244.88 + 245.20 + 244.97) / 4 = 245.0425 meters; for area T2, the average is (246.08 + 246.30 + 246.97) / 4 = 245.0425 meters. (0.15 + 245.97) / 4 = 246.125 meters; the average elevation difference between plots ΔH group = 246.125 − 245.0425 = 1.0825 meters; the weighted elevation value of 246.08 meters obtained by the fusion calculation of number A_005, as the first data point of area T2, is in a core position in the calculation logic of internal difference ΔH_T2 and inter-group difference ΔH group, directly affecting the accuracy of the T2 average elevation expression and the comparison of elevation structure between plots; finally, the elevation difference range of all area numbers, the regional center elevation index and the inter-group elevation difference are integrated according to the plot combination method to generate the land surveying multi-source data fusion analysis results.

[0042] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A land surveying data analysis system based on multi-source data fusion, characterized in that: The system includes: The raster unification module acquires multi-source terrain data, establishes a unified UTM grid code, confirms the consistency of spatial units corresponding to the numbering by combining spatial location coordinates, and generates a multi-source numbering unified control group. The fluctuation extraction module extracts spatial measurement values, calculates the range, and obtains the maximum difference sequence based on the multi-source numbered unified control group. It then compares the difference with the vertical accuracy level of the DEM and filters the stable unit number group to obtain the spatial fluctuation stable number set. The data elimination module counts the frequency of numerical deviations of each data source based on the spatial fluctuation stability number set, calculates the abnormal proportion value of each data source, and eliminates data sources whose proportion value exceeds the horizontal positioning accuracy to obtain the distribution record of available sources participating in the fusion. The regional compression module extracts the numerical sequence of available data sources in a continuous spatial region based on the distribution records of available sources participating in the fusion, calculates the standard deviation within the continuous region, removes regions with a standard deviation lower than the terrain roughness threshold, and obtains regional information cropped data.

2. The land surveying data analysis system based on multi-source data fusion according to claim 1, characterized in that: The multi-source numbering unified control group includes numbering mapping relationship, spatial location correspondence relationship, and coordinate system unified coding. The spatial fluctuation stable numbering set includes precision conformity numbering set, spatial stability identifier, and difference threshold judgment basis. The available source distribution record participating in fusion includes available data source identifier, distribution information within the numbering group, and source removal statistics.

3. The land surveying data analysis system based on multi-source data fusion according to claim 1, characterized in that: The grid unification module includes: The number extraction submodule obtains the numbering information of multi-source terrain data, extracts and collects the numbered data, lists the number sets according to the data sources, matches each number set with spatial location coordinates, and generates number coordinate combination results. The intersection matching submodule compares the spatial location coordinates of the same number in different data sources based on the number coordinate combination results. It calculates the spatial residual of each coordinate group through unified projection transformation, determines whether the residual is lower than the spatial reference system offset threshold, and counts the number combinations that meet the conditions to obtain the consistent number matching comparison results. The grid coding construction submodule, based on the consistent number matching comparison results, divides the corresponding number combinations into UTM grids, generates unified grid unit numbers according to the spatial boundary and coordinate range of the area to which the number belongs, and generates a unified control group of multi-source numbers.

4. The land surveying data analysis system based on multi-source data fusion according to claim 3, characterized in that: The spatial residuals for each coordinate group are calculated using the following formula: ； Calculations are performed, in which, The representative number index is Spatial location residual value, The representative number index is The x-coordinate value extracted from data source A. The index represents the x-coordinate value extracted from data source B. The representative number index is The ordinate value extracted from data source A. The representative number index is The ordinate value extracted from data source B. This represents the sum of the differences in the x-coordinates of all numbers from item 1 to item N between data source A and data source B. This represents the sum of the differences in the ordinates of all numbers from the 1st to the Nth item between data source A and data source B. This indicates the total number of numbers involved in the calculation of spatial location residuals.

5. The land surveying data analysis system based on multi-source data fusion according to claim 1, characterized in that: The fluctuation extraction module includes: The measurement extraction submodule obtains the spatial location corresponding to each group number in the multi-source numbered unified control group, collects the spatial measurement elevation values ​​of all data sources at each location, records the corresponding number and source identifier, collects them according to the number and establishes a measurement sequence to generate a spatial measurement set; The difference operation submodule calculates the numerical difference between the maximum and minimum values ​​in each group based on the numbered measurement sequence in the spatial measurement set, extracts the range values ​​under all numbers and constructs the maximum difference sequence to obtain the numbered range sequence. The stability marking submodule, based on the range value corresponding to each number in the numbered range sequence, and according to the DEM vertical accuracy level benchmark value, discriminates and marks the range value, classifies the numbers with values ​​less than or equal to the benchmark value into stable unit number groups, and obtains the spatial fluctuation stable number set.

6. The land surveying data analysis system based on multi-source data fusion according to claim 1, characterized in that: The data removal module includes: The frequency statistics submodule extracts the spatial measurement elevation values ​​of each data source in the consecutive numbered group based on the spatial fluctuation stability number set, accumulates the number of deviation records, counts the deviation frequency of each data source in each numbered group, and generates source deviation frequency data. The ratio judgment submodule calculates the ratio of the number of deviations of each data source in the source deviation frequency data to the total number of corresponding numbers, obtains the abnormal ratio value of each data source in each number group, compares the ratio value with the set horizontal positioning accuracy threshold, obtains the list of data sources marked as unavailable, and obtains the unavailable source number record. The source removal submodule removes the corresponding data records of the data sources in each number group according to the unusable source number records, retains the remaining usable source records, and collects and maps the numbers, locations and spatial measurement elevation values ​​of the usable data sources with number identifiers to obtain the distribution records of usable sources participating in the fusion.

7. The land surveying data analysis system based on multi-source data fusion according to claim 1, characterized in that: The region compression module includes: The sequence extraction submodule extracts available data source measurement values ​​from each group of continuous spatial regions based on the distribution records of available sources participating in the fusion, collects spatial measurement elevation values ​​by region number, organizes the number sequence and corresponding values ​​to form a numerical sequence set, and generates a continuous region numerical sequence. The standard deviation calculation submodule calculates the standard deviation of the measured values ​​under each group number based on the continuous region numerical sequence, extracts the standard deviation value of each number region, establishes a mapping list between the number and the corresponding standard deviation, and obtains the regional standard deviation set. The region clipping submodule compares the standard deviation corresponding to each number in the region standard deviation set with the terrain roughness threshold, filters out regions with standard deviations below the threshold and records the corresponding numbers, summarizes the remaining region numbers and numerical information, and obtains region information clipping data.

8. The land surveying data analysis system based on multi-source data fusion according to claim 1, characterized in that: The system also includes: The fusion analysis module clips data based on the regional information, aggregates regional spatial measurement values, calculates weighted average elevation values, and performs land area comparative analysis to generate fusion analysis results of multi-source land surveying data. The regional information cropping data includes the cropped region number, terrain roughness classification identifier, and spatial continuous region range.

9. The land surveying data analysis system based on multi-source data fusion according to claim 8, characterized in that: The fusion analysis module includes: The measurement summary submodule extracts all available spatial measurement elevation values ​​from each available data source under each region number based on the region information cropping data, integrates the measurement values ​​and corresponding source weight information of each data source according to the number, forms a combination sequence of number and weight measurement values, and generates a measurement weight sequence set; The weighted calculation submodule calculates the weighted average elevation value within each group of numbers based on the measured values ​​and weights of all data sources under each number in the measured value weight sequence set, extracts the weighted average value of each number and organizes it into a number index list to obtain a set of weighted elevation values. The regional comparison submodule assigns the weighted average elevation value of each number in the weighted elevation value set to the corresponding region according to the land parcel division number, compares the weighted elevation difference range corresponding to the number in each region, summarizes the elevation difference results between regions, and obtains the multi-source data fusion analysis results of land surveying.

10. The land surveying data analysis system based on multi-source data fusion according to claim 9, characterized in that: The weighted average elevation value is calculated using the following formula: ； Calculations are performed, in which, Representative number The weighted average elevation value below, Representative number The corresponding number Elevation measurements from one data source Representative number Next The weight values ​​assigned to each data source. Representative number The total number of available data sources included below. The index number of the data source. An identifier index for space numbering.

Images