GIS-based land classification depth measurement method and system

Abstract

The invention discloses a GIS-based land classification depth measurement method and system, relates to the technical field of land classification depth measurement, and aims to solve the problems that in traditional land depth measurement work, due to the fact that the number of measurement points is limited and distribution is lack of systematicness, the real change condition of land depth is difficult to comprehensively and accurately reflect, local depth abnormal areas are easy to omit, and the measurement accuracy is poor. According to the method, multi-source geographic data are integrated through the GIS technology, land depth information can be comprehensively and accurately reflected, and the problems that traditional measurement is low in precision, measurement point distribution is lack of systematicness, and data processing and analysis capacity is weak are solved. In the aspect of region division, the consistency of land features in the region is ensured by integrating a multi-factor consideration method, a dynamic adjustment mechanism of measurement point layout can be flexibly arranged according to the actual condition of the land, a slope calculation and depth estimation method is combined, the complex relation between geographic elements is fully considered, and estimation errors are effectively reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of land classification depth measurement, and specifically to a land classification depth measurement method and system based on GIS. Background Art

[0002] With the rapid development of social economy, the rational development, utilization and management of land resources have become a vital issue. As one of the key attributes of land resources, the accurate measurement of land depth is indispensable to many fields, such as agricultural production, urban planning, geological exploration and environmental protection.

[0003] In this regard, the patent document with publication number CN113360587B discloses a land surveying and mapping device and a surveying and mapping method based on GIS technology, including: a data acquisition module, used to collect terrain data of the land to be detected based on GIS technology; a data processing module, used to process the terrain data to obtain surveying and mapping data; a drawing module, used to draw a topographic map of the land to be detected based on the surveying and mapping data. By collecting the terrain data of the land to be detected based on GIS technology, the accuracy and efficiency of data acquisition are improved. By processing the terrain data to obtain surveying and mapping data, the reliability of the surveying and mapping data is guaranteed, so that the final topographic map can better reflect the actual situation of the land.

[0004] In traditional land depth measurement work, due to the limited number of measurement points and the lack of systematic distribution, it is difficult to fully and accurately reflect the real changes in land depth. It is easy to miss local depth abnormal areas, resulting in large deviations in measurement results. The data obtained are often discrete and isolated, lacking effective integration and in-depth analysis methods, making it difficult to conduct a comprehensive and in-depth exploration of the spatial distribution characteristics of land depth.

[0005] In view of the above problems, a GIS-based land classification depth measurement method and system are proposed. Summary of the invention

[0006] The purpose of the present invention is to provide a GIS-based land classification depth measurement method and system, which solves the problem of omissions in measurement in the background technology.

[0007] To achieve the above object, the present invention provides the following technical solution: a GIS-based land classification depth measurement system, comprising:

[0008] Area division module, measurement point layout module, measurement data acquisition module, calculation verification module, depth estimation adjustment module and data integration module;

[0009] The regional division module includes:

[0010] The data collection submodule collects topographic data, geological structure data and soil type data;

[0011] The analysis and processing submodule conducts comprehensive analysis and processing of the collected multi-source data, uses geographic information system technology to perform data superposition and cluster analysis, and divides the region into multiple areas based on the analysis and processing results, and lists them as area one, area two, and area three;

[0012] The measurement point layout module is used to layout measurement points in each area, including:

[0013] The basic layout submodule uses uniform grid layout measurement points in relatively flat areas with uniform land properties, and uses dense measurement points in areas close to regional boundaries or with large terrain changes, based on the area and the requirements of the multi-source data collected;

[0014] Dynamic adjustment submodule, encrypts measurement points and arranges measurement points in a circular outward diffusion pattern with this point as the center, and calculates the circle radius and measurement point density according to terrain changes;

[0015] The measurement data acquisition module is used to collect measurement data, communicate with and control the measurement instrument, send measurement instructions, receive and store measurement data, record land depth data, measurement point coordinate data and measurement time information, and store them in the database management system;

[0016] The calculation and verification module is used to calculate and verify the slope of land depth change. The spatial regression model is introduced according to the data sent by the measurement data acquisition module for calculation. The measurement data is divided into a training set and a verification set. The cross-validation and bootstrap techniques are used to verify and evaluate the slopes calculated by different models.

[0017] The depth estimation and adjustment module is used to estimate the land depth at the boundary of area 2 and adjust the measurement plan. According to the slope determined in area 1 and the boundary characteristics of area 2, the estimation model is used to calculate the estimated land depth at the boundary of area 2. During the measurement process of area 2 and area 3, the measured depth value is compared with the estimated depth value in real time. If the error exceeds the threshold range, the measurement is suspended for a comprehensive inspection, and the measurement plan is adjusted according to the inspection results.

[0018] The data integration module is used to integrate data and improve the model. After completing the measurement of each area, the measurement data is integrated and cleaned, outliers and duplicate data are removed, and standardized. The integrated data is used to build a three-dimensional spatial model of land depth. Based on the data of the entire measurement process, the slope estimation model and depth measurement model are reviewed and improved, and the model parameters and variable weights are adjusted.

[0019] Preferably, the data acquisition submodule acquires topographic data through satellite remote sensing images, and when acquiring geological structure data, a combination of geological drilling and geophysical exploration methods is used.

[0020] Preferably, when the analysis and processing submodule performs cluster analysis, the clustering algorithm adopted is the K-means clustering algorithm, and the formula for calculating the cluster center is: Among them, C j is the jth cluster center, x i is the i-th data point, δ ij is an exponential function, when the data point x i When it belongs to cluster j, δ ij =1, otherwise δ ij =0, n is the total number of data points.

[0021] Preferably, when the basic layout submodule arranges the measurement points in a uniform grid, the calculation formula for the grid spacing is: Where s is the area of ​​region 1, m is the number of preset basic measurement points, and V is the terrain relief of region 1.

[0022] Preferably, when the dynamic adjustment submodule calculates the circular radius r, the calculation formula is: Where Δh is the standard deviation of the land depth change within a certain range centered on the outlier point, which is calculated using the elevation data of adjacent measurement points. θ is the terrain slope angle, which is calculated based on the terrain data. k3 is the adjustment factor. If there is no terrain undulation in the area, the k3 value range is [1,1.5]. If there is terrain undulation in the area, the k3 value range is [1.5,2.5].

[0023] Preferably, when the calculation verification module uses the spatial regression model to calculate the slope, the spatial autoregression model formula used is: y=ρWy+Xβ+∈, where y is the land depth vector, ρ is the spatial autoregression coefficient, W is the spatial weight matrix, representing the spatial adjacency relationship between each measurement point, X is the explanatory variable matrix, including the measurement point coordinates, terrain feature variables, β is the explanatory variable coefficient vector, ∈ is the random error vector, and when calculating the cross-validation error, the root mean square error formula is used: where y i is the actual value of the i-th validation set data point, is the corresponding predicted value, and Z is the number of data points in the validation set.

[0024] Preferably, when dividing the measured data into a training set and a validation set, a stratified sampling method is used to stratify the data according to factors such as land type and terrain characteristics, and each layer of data is divided into a training set and a validation set according to a preset ratio of 7:3, and when using the bootstrap technique, the number of bootstrap sample sets N is b Determined according to the original data volume N, the formula is Where f is the sample set adjustment coefficient, and its value range is [1.5-2.5].

[0025] Preferably, when the depth estimation adjustment module uses the estimation model to calculate the estimated value of the land depth at the boundary of the second region, the estimation model is a multivariate linear regression model, and its formula y p =β0+β1x1+β2x2+…β n x n , where y p To estimate the land depth, β0, β1, β2, β n are regression coefficients, determined by training with area 1 measurement data, x1, x2, … x n is the explanatory variable, including the slope value of area 1. During the measurement process of areas 2 and 3, if the average error between the measured depth values ​​and the estimated depth values ​​of m consecutive measurement points is greater than the set threshold ∈1, it is determined that the measurement error exceeds the threshold range, and the measurement is suspended for a comprehensive inspection. The value range of m is [5, 10], and the value range of ∈1 is [0.1, 0.3].

[0026] Preferably, the data integration module uses the Kriging interpolation method when constructing the three-dimensional spatial model of land depth, and its semivariogram model is a Gaussian model, and the formula is: Where γ(A) is the semivariogram function value, A is the distance between sample points, Q0 is the nugget value, which indicates the variation caused by random error, Q is the arch height, which indicates the variation caused by spatial autocorrelation, and T is the range, which indicates the range of spatial autocorrelation.

[0027] An application method of GIS-based land classification depth measurement system:

[0028] S1, start the measurement system and complete the initialization and data preparation work, use the regional division module to collect multi-source data for regional division, and determine region 1, region 2 and region 3;

[0029] S2, in area 1, basic measurement points are laid out according to the area and data requirements. In flat and uniform areas, grids are laid out at specific intervals. The distribution of measurement points is encrypted or adjusted at boundaries and places with large terrain changes. If an abnormal point of depth change is encountered, a circular diffusion layout is carried out with it as the center and the appropriate radius and density are determined. The instrument is controlled to collect and store data through the measurement data acquisition module. The calculation and verification module introduces a spatial autoregressive model to calculate the slope, uses stratified sampling to divide the training set and the verification set, and uses cross-validation and bootstrap technology to verify and evaluate, and determine the optimal slope calculation model and parameters.

[0030] S3, based on the results of area one, area two is measured. The depth estimation and adjustment module uses a multivariate linear regression model to estimate the boundary depth of area two. The measurement points of area two are arranged according to the experience of area one and adjusted as needed. During the measurement, the measured and estimated depths are compared. If the error exceeds the threshold, the inspection is suspended and continued after adjustment. The above steps are repeated to measure area three.

[0031] Compared with the prior art, the present invention has the following beneficial effects:

[0032] 1. The present invention provides a GIS-based land classification depth measurement method and system, which improves the measurement efficiency and avoids the drawbacks of manual measurement on large areas of land that consumes a lot of manpower, material resources and time. By integrating multi-source geographic data through GIS technology, the land depth information can be fully and accurately reflected, overcoming the problems of low traditional measurement accuracy, lack of systematic distribution of measurement points, and weak data processing and analysis capabilities. In terms of regional division, the method of comprehensive multi-factor consideration ensures the consistency of land characteristics within the region and provides an accurate basis for subsequent measurements. The dynamic adjustment mechanism of the measurement point layout can be flexibly arranged according to the actual land conditions to ensure measurement accuracy under complex terrain. At the same time, the slope calculation and depth estimation method fully considers the complex relationship between geographical elements, and has an error monitoring and adjustment mechanism to effectively reduce the estimation error. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 is a flow chart of the present invention;

[0034] Figure 2 This is a flow chart of slope calculation and verification and depth estimation and adjustment of the present invention. DETAILED DESCRIPTION

[0035] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0036] In order to further understand the content of the present invention, the present invention is described in detail in conjunction with the accompanying drawings.

[0037] Combination Figure 1-Figure 2 The present invention provides a GIS-based land classification depth measurement system, comprising:

[0038] Area division module, measurement point layout module, measurement data acquisition module, calculation verification module, depth estimation adjustment module and data integration module;

[0039] The regional division module includes:

[0040] The data acquisition submodule acquires topographic data through satellite remote sensing images, and when acquiring geological structure data, it uses geological drilling combined with geophysical exploration methods to acquire topographic data, geological structure data, and soil type data;

[0041] The analysis and processing submodule performs comprehensive analysis and processing on the collected multi-source data, and uses geographic information system technology to perform data superposition and cluster analysis. Through multi-source data fusion, more comprehensive and accurate terrain data can be obtained, reducing the limitations of a single data source and improving the accuracy of subsequent analysis and measurement. When performing cluster analysis, the clustering algorithm used by the analysis and processing submodule is the K-means clustering algorithm, and the formula for calculating the cluster center is: Among them, C j is the jth cluster center, x i is the i-th data point, δ ij is an exponential function, when the data point x i When it belongs to cluster j, δ ij =1, otherwise δ ij = 0, n is the total number of data points, and the region is divided into multiple regions according to the analysis and processing results, and listed as region 1, region 2 and region 3;

[0042] In addition, in the process of analyzing the superposition of geological structure data and topographic data, in order to determine the intersection of geological structure lines and topographic contour lines, first of all, it is necessary to accurately digitize and extract geological structure lines and topographic contour lines through GIS software;

[0043] The intersection angle is calculated by using the vector calculation method to calculate the angle between the tangent direction vectors of the two lines at the intersection. The terrain slope is calculated based on the digital elevation model using the neighborhood difference algorithm;

[0044] When the elevation difference is calculated for a 3×3 or 5×5 pixel window around the central pixel and the slope is calculated in combination with the distance, when the intersection angle is less than α1, the value range is [30°, 45°], and the terrain slope at the intersection point is greater than the set slope value β1, the value range is [15°, 25°], the area is divided separately for subsequent key measurement and analysis.

[0045] The measurement point layout module is used to layout measurement points in each area, including: when using a uniform grid layout for measurement points in relatively flat areas with uniform land properties, further optimize the grid layout according to the land use type. For example, in grasslands, consider the layout of crop planting directions so that measurement points can better reflect the characteristics of land depth related to agricultural production. For uphill slopes with undulations, reasonably adjust the grid spacing to ensure that measurement points can capture changes in land depth.

[0046] By optimizing the grid layout in combination with land use types, the measurement points can more effectively reflect the land depth characteristics under different land use types, improve the relevance of measurement data to practical applications, and provide more targeted data support for land development and management.

[0047] The basic layout submodule adopts uniform grid layout measurement points in relatively flat areas with uniform land properties, and adopts dense measurement points in areas close to regional boundaries or with large terrain changes, according to the area and the requirements of the multi-source data collected. When the basic layout submodule arranges measurement points in a uniform grid, the calculation formula for the grid spacing is: Where s is the area of ​​region 1, m is the number of preset basic measurement points, and V is the terrain relief of region 1;

[0048] When the measurement points are placed close to the region boundary, the distance D from the measurement point to the boundary satisfies the formula The measurement of the length B of the regional boundary can be carried out by calculating the length of the digitized path of the regional boundary and calculating the length of the line element through GIS software; the number of boundary measurement points n is preset. b It is pre-set according to the complexity of the boundary and the measurement accuracy requirements. Generally, complex boundaries, such as those with many twists and turns and large curvature, have more presets, while simple boundaries have fewer presets. The boundary adjustment coefficient ranges from [0.5, 1.5], which is determined according to the complexity of the terrain in the boundary area. For areas with relatively flat terrain and slow boundary changes, the value is [0.5, 0.8], while for areas with large terrain undulations and large boundary curvature, the value is [1.2, 1.5], to ensure that the distribution of measurement points in the boundary area is reasonable and can effectively reflect the changes in land depth at the boundary.

[0049] The dynamic adjustment submodule encrypts the measurement points and arranges the measurement points in a circular outward diffusion pattern with the point as the center. The circular radius and the measurement point density are calculated according to the terrain changes. When the dynamic adjustment submodule calculates the circular radius r, the calculation formula is: Where Δh is the standard deviation of the land depth change within a certain range centered on the outlier point, which is calculated using the elevation data of adjacent measurement points; θ is the terrain slope angle, which is calculated based on terrain data; k3 is the adjustment factor. If there is no terrain undulation in the area, the k3 value range is [1,1.5]; if there is terrain undulation in the area, the k3 value range is [1.5,2.5];

[0050] The measurement data acquisition module is used to collect measurement data, communicate with and control the measurement instrument, send measurement instructions, receive and store measurement data, record land depth data, measurement point coordinate data and measurement time information, and store them in the database management system;

[0051] The calculation and verification module is used to calculate and verify the slope of land depth change. The spatial regression model is introduced for calculation according to the data sent by the measurement data acquisition module. The measurement data is divided into a training set and a verification set. The slopes calculated by different models are verified and evaluated using cross-validation and bootstrap techniques. When the calculation and verification module uses the spatial regression model to calculate the slope, the spatial autoregression model formula used is: y=ρWy+Xβ+∈, where y is the land depth vector, ρ is the spatial autoregression coefficient, W is the spatial weight matrix, which represents the spatial adjacency relationship between each measurement point, X is the explanatory variable matrix, including the measurement point coordinates and terrain feature variables, β is the explanatory variable coefficient vector, ∈ is the random error vector, and when calculating the cross-validation error, the root mean square error formula is used: where y i is the actual value of the i-th validation set data point, is the corresponding predicted value, Z is the number of validation set data points, and when the measured data is divided into training set and validation set, the stratified sampling method is adopted to stratify the data according to factors such as land type and terrain characteristics. Each layer of data is divided into the training set and the validation set according to the preset ratio of 7:3. When the bootstrap method is used, the number of bootstrap sample sets N b Determined according to the original data volume N, the formula is Where f is the sample set adjustment coefficient, and its value range is [1.5-2.5];

[0052] The depth estimation and adjustment module is used to estimate the land depth at the boundary of area 2 and adjust the measurement plan. According to the slope determined in area 1 and the boundary characteristics of area 2, the estimation model is used to calculate the estimated land depth at the boundary of area 2. During the measurement process of area 2 and area 3, the measured depth value is compared with the estimated depth value in real time. If the error exceeds the threshold range, the measurement is suspended for a comprehensive inspection, and the measurement plan is adjusted according to the inspection results.

[0053] When the depth estimation adjustment module uses the estimation model to calculate the estimated land depth at the boundary of the second area, the estimation model is a multivariate linear regression model, and its formula y p =β0+β1x1+β2x2+…β n x n , where y p To estimate the land depth, β0, β1, β2, β n are regression coefficients, determined by training with area 1 measurement data, x1, x2, … x n is the explanatory variable, including the slope value of area 1. During the measurement of areas 2 and 3, if the average error between the measured depth value and the estimated depth value of m consecutive measurement points is greater than the set threshold ∈1, it is determined that the measurement error exceeds the threshold range, and the measurement is suspended for a comprehensive inspection. The value range of m is [5, 10], and the value range of ∈1 is [0.1, 0.3];

[0054] When using the estimation model to calculate the estimated land depth at the boundary of Region 2, in addition to considering factors such as the slope value of Region 1, the terrain characteristic values ​​of the boundary points, and the soil type-related values, time series analysis is also introduced to analyze the changing trends of land depth in different seasons and years. In particular, in areas that are greatly affected by precipitation and groundwater fluctuations, the time factor is incorporated into the estimation model to improve the accuracy of the prediction;

[0055] The introduction of time series analysis can better capture the changing patterns of land depth in the time dimension, especially for areas that are greatly affected by natural factors. It can make the depth estimation model more in line with the actual situation, reduce prediction errors, and provide more reliable prediction data for the long-term planning and management of land resources.

[0056] The data integration module is used to integrate data and improve the model. After completing the measurement of each area, the measurement data is integrated and cleaned, outliers and duplicate data are removed, and standardized. The integrated data is used to build a three-dimensional spatial model of land depth. Based on the data of the entire measurement process, the slope estimation model and the depth measurement model are reviewed and improved, and the model parameters and variable weights are adjusted. When constructing the three-dimensional spatial model of land depth, the data integration module uses the Kriging interpolation method, and its semivariogram model is a Gaussian model. The formula is: Where γ(A) is the semivariogram function value, A is the distance between sample points, Q0 is the nugget value, which indicates the variation caused by random error, Q is the arch height, which indicates the variation caused by spatial autocorrelation, and T is the range, which indicates the range of spatial autocorrelation.

[0057] An application method of GIS-based land classification depth measurement system:

[0058] S1, start the measurement system and complete the initialization and data preparation work, use the regional division module to collect multi-source data for regional division, and determine region 1, region 2 and region 3;

[0059] S2, in area 1, basic measurement points are laid out according to the area and data requirements. In flat and uniform areas, grids are laid out at specific intervals. The distribution of measurement points is encrypted or adjusted at boundaries and places with large terrain changes. If an abnormal point of depth change is encountered, a circular diffusion layout is carried out with it as the center and the appropriate radius and density are determined. The instrument is controlled to collect and store data through the measurement data acquisition module. The calculation and verification module introduces a spatial autoregressive model to calculate the slope, uses stratified sampling to divide the training set and the verification set, and uses cross-validation and bootstrap technology to verify and evaluate, and determine the optimal slope calculation model and parameters.

[0060] S3, based on the results of area one, area two is measured. The depth estimation and adjustment module uses a multivariate linear regression model to estimate the boundary depth of area two. The measurement points of area two are arranged according to the experience of area one and adjusted as needed. During the measurement, the measured and estimated depths are compared. If the error exceeds the threshold, the inspection is suspended and continued after adjustment. The above steps are repeated to measure area three.

[0061] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0062] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A GIS-based land classification depth measurement system, characterized in that: include: Area division module, measurement point layout module, measurement data acquisition module, calculation verification module, depth estimation adjustment module and data integration module; The regional division module includes: The data collection submodule collects topographic data, geological structure data and soil type data; The analysis and processing submodule conducts comprehensive analysis and processing of the collected multi-source data, uses geographic information system technology to perform data superposition and cluster analysis, and divides the region into multiple areas based on the analysis and processing results, and lists them as area one, area two, and area three; The measurement point layout module is used to layout measurement points in each area, including: The basic layout submodule uses uniform grid layout measurement points in relatively flat areas with uniform land properties, and uses dense measurement points in areas close to regional boundaries or with large terrain changes, based on the area and the requirements of the multi-source data collected; Dynamic adjustment submodule, encrypts measurement points and arranges measurement points in a circular outward diffusion pattern with this point as the center, and calculates the circle radius and measurement point density according to terrain changes; The measurement data acquisition module is used to collect measurement data, communicate with and control the measurement instrument, send measurement instructions, receive and store measurement data, record land depth data, measurement point coordinate data and measurement time information, and store them in the database management system; The calculation and verification module is used to calculate and verify the slope of land depth change. The spatial regression model is introduced according to the data sent by the measurement data acquisition module for calculation. The measurement data is divided into a training set and a verification set. The cross-validation and bootstrap techniques are used to verify and evaluate the slopes calculated by different models. The depth estimation and adjustment module is used to estimate the land depth at the boundary of area 2 and adjust the measurement plan. According to the slope determined in area 1 and the boundary characteristics of area 2, the estimation model is used to calculate the estimated land depth at the boundary of area 2. During the measurement process of area 2 and area 3, the measured depth value is compared with the estimated depth value in real time. If the error exceeds the threshold range, the measurement is suspended for a comprehensive inspection, and the measurement plan is adjusted according to the inspection results. The data integration module is used to integrate data and improve the model. After completing the measurement of each area, the measurement data is integrated and cleaned, outliers and duplicate data are removed, and standardized. The integrated data is used to build a three-dimensional spatial model of land depth. Based on the data of the entire measurement process, the slope estimation model and depth measurement model are reviewed and improved, and the model parameters and variable weights are adjusted.

2. The GIS-based land classification depth measurement system according to claim 1, characterized in that: The data acquisition submodule obtains topographic data through satellite remote sensing images, and when collecting geological structure data, it adopts geological drilling combined with geophysical exploration methods.

3. The GIS-based land classification depth measurement system according to claim 1 is characterized by: When the analysis and processing submodule performs cluster analysis, the clustering algorithm adopted is the K-means clustering algorithm, and the formula for calculating the cluster center is: Among them, C j is the jth cluster center, x i is the i-th data point, δ ij is an exponential function, when the data point x i When it belongs to cluster j, δ ij =1, otherwise δ ij =0, n is the total number of data points.

4. The GIS-based land classification depth measurement system according to claim 1 is characterized by: When the basic layout submodule arranges the measurement points in a uniform grid, the calculation formula for the grid spacing is: Where s is the area of ​​region 1, m is the number of preset basic measurement points, and V is the terrain relief of region 1.

5. The GIS-based land classification depth measurement system according to claim 1 is characterized by: When the dynamic adjustment submodule calculates the circular radius r, the calculation formula is: Where Δh is the standard deviation of the land depth change within a certain range centered on the outlier point, which is calculated using the elevation data of adjacent measurement points. θ is the terrain slope angle, which is calculated based on terrain data. k3 is the adjustment factor. If there is no terrain undulation in the area, the k3 value range is [1,1.5]. If there is terrain undulation in the area, the k3 value range is [1.5,2.5].

6. The GIS-based land classification depth measurement system according to claim 1 is characterized by: When the calculation verification module uses the spatial regression model to calculate the slope, the spatial autoregression model formula used is: y=ρWy+Xβ+∈, where y is the land depth vector, ρ is the spatial autoregression coefficient, W is the spatial weight matrix, which represents the spatial adjacency relationship between each measurement point, X is the explanatory variable matrix, including the measurement point coordinates and terrain feature variables, β is the explanatory variable coefficient vector, ∈ is the random error vector, and when calculating the cross-validation error, the root mean square error formula is used: where y i is the actual value of the i-th validation set data point, is the corresponding predicted value, and Z is the number of data points in the validation set.

7. The GIS-based land classification depth measurement system according to claim 6 is characterized by: When dividing the measured data into training set and validation set, the stratified sampling method is adopted to stratify the data according to factors such as land type and terrain characteristics. Each layer of data is divided into training set and validation set according to the preset ratio of 7:

3. When using the bootstrap method, the number of bootstrap sample sets N is b Determined according to the original data volume N, the formula is Where f is the sample set adjustment coefficient, and its value range is [1.5-2.5].

8. The GIS-based land classification depth measurement system according to claim 1 is characterized by: When the depth estimation adjustment module uses the estimation model to calculate the estimated land depth at the boundary of the second area, the estimation model is a multivariate linear regression model, and its formula y p =β0+β1x1+β2x2+…β n x n , where y p To estimate the land depth, β0, β1, β2, β n are regression coefficients, determined by training with area 1 measurement data, x1, x2, … x n is the explanatory variable, including the slope value of area 1. During the measurement process of areas 2 and 3, if the average error between the measured depth values ​​and the estimated depth values ​​of m consecutive measurement points is greater than the set threshold ∈1, it is determined that the measurement error exceeds the threshold range, and the measurement is suspended for a comprehensive inspection. The value range of m is [5, 10], and the value range of ∈1 is [0.1, 0.3].

9. The GIS-based land classification depth measurement system according to claim 1, characterized in that: The data integration module uses the Kriging interpolation method when constructing the three-dimensional spatial model of land depth, and its semivariogram model is a Gaussian model, and the formula is: Where γ(A) is the semivariogram function value, A is the distance between sample points, Q0 is the nugget value, which indicates the variation caused by random error, Q is the arch height, which indicates the variation caused by spatial autocorrelation, and T is the range, which indicates the range of spatial autocorrelation.

10. The application method of the GIS-based land classification depth measurement system according to any one of claims 1 to 9, characterized in that: S1, start the measurement system and complete the initialization and data preparation work, use the regional division module to collect multi-source data for regional division, and determine region 1, region 2 and region 3; S2, in area 1, basic measurement points are laid out according to the area and data requirements. In flat and uniform areas, grids are laid out at specific intervals. The distribution of measurement points is encrypted or adjusted at boundaries and places with large terrain changes. If an abnormal point of depth change is encountered, a circular diffusion layout is carried out with it as the center and the appropriate radius and density are determined. The instrument is controlled to collect and store data through the measurement data acquisition module. The calculation and verification module introduces a spatial autoregressive model to calculate the slope, uses stratified sampling to divide the training set and the verification set, and uses cross-validation and bootstrap technology to verify and evaluate, and determine the optimal slope calculation model and parameters. S3, based on the results of area one, area two is measured. The depth estimation and adjustment module uses a multivariate linear regression model to estimate the boundary depth of area two. The measurement points of area two are arranged according to the experience of area one and adjusted as needed. During the measurement, the measured and estimated depths are compared. If the error exceeds the threshold, the inspection is suspended and continued after adjustment. The above steps are repeated to measure area three.

Images