Soil heavy metal content prediction method, device, equipment, medium and product

Abstract

The invention discloses a soil heavy metal content prediction method, device, equipment, medium and product, and relates to the technical field of soil detection, the method comprises the following steps: constructing a multiple regression data set; performing multivariate stepwise regression fitting by adopting the data set to obtain a multivariate stepwise regression model; determining a plurality of soil sampling points from the target soil area, and obtaining a heavy metal content measured value of each soil sampling point; performing heavy metal content inversion on each soil sampling point by adopting a multivariate stepwise regression model to obtain an initial heavy metal content predicted value of each soil sampling point; calculating a difference value between each heavy metal content measured value and the initial heavy metal content predicted value at the corresponding position, and performing spatial interpolation on each difference value by adopting a Kriging interpolation method to obtain spatial distribution of prediction errors; and determining a heavy metal content predicted value of any position on the target soil area according to the spatial distribution of the prediction error and a multivariate stepwise regression model. The soil heavy metal content inversion precision can be improved.

Description

Technical Field

[0001] The present application relates to the field of soil detection technology, and in particular to a method, device, equipment, medium and product for predicting heavy metal content in soil. Background Art

[0002] By utilizing the spectral characteristics of soil containing heavy metals, integrating the measured soil hyperspectral curves, the measured heavy metal content of soil samples, and the advantages of remote sensing images, we can find factors that are sensitive to soil heavy metal stress and build a model to invert the spatial distribution of regional soil heavy metal content. To achieve the goal of accurately inverting the heavy metal content in soil, model construction is extremely critical. Related modeling methods, such as multivariate linear regression, have certain limitations due to the high multicollinearity of independent variables: (1) There is no in-depth analysis of the dependency relationship between variables, and the model has too many variables, resulting in a decrease in the inversion accuracy of soil heavy metal content. (2) Due to the multicollinearity between variables, there are redundant variables in the model, which increases the cost of collecting variable data and model operation. Summary of the invention

[0003] The purpose of this application is to provide a soil heavy metal content prediction method, device, equipment, medium and product, which can improve the accuracy of soil heavy metal content inversion.

[0004] To achieve the above objectives, this application provides the following solutions:

[0005] In a first aspect, the present application provides a method for predicting the content of heavy metals in soil, the method comprising:

[0006] Constructing a data set for multiple regression of soil heavy metal content; the data set includes multiple candidate independent variables and one dependent variable; the dependent variable is the heavy metal content;

[0007] The data set is used to perform multivariate stepwise regression fitting to obtain a multivariate stepwise regression model;

[0008] Determine a plurality of soil sampling points from the target soil area, and obtain the measured value of the heavy metal content at each soil sampling point;

[0009] The multivariate stepwise regression model is used to invert the heavy metal content of each soil sampling point to obtain the initial heavy metal content prediction value of each soil sampling point;

[0010] Calculating the difference between each heavy metal content measurement value and the initial heavy metal content prediction value at the corresponding position, and performing spatial interpolation on each of the differences using the Kriging interpolation method to obtain the spatial distribution of the prediction error;

[0011] According to the spatial distribution of prediction errors and the multivariate stepwise regression model, the predicted value of heavy metal content at any location in the target soil area is determined.

[0012] Optionally, a data set for multiple regression of soil heavy metal content is constructed, specifically including:

[0013] Collect remote sensing images, heavy metal content and spectral reflectance of each soil sample;

[0014] Preprocessing the remote sensing image to obtain multiple soil heavy metal related indexes;

[0015] Performing a tassel-cap transformation on the remote sensing image to obtain a brightness component of the remote sensing image;

[0016] Extracting reflectance of multiple bands from the remote sensing image and performing principal component analysis to obtain multiple principal component bands;

[0017] Preprocessing the spectral reflectance to obtain a preprocessed spectral reflectance;

[0018] Performing outlier elimination processing on the heavy metal content to obtain the heavy metal content after pretreatment;

[0019] For each soil sample, the first-order differential of the spectral reflectance after pretreatment is used as an independent variable, the brightness component, each principal component band and each soil heavy metal related index are used as candidate independent variables, and the heavy metal content after pretreatment is used as a dependent variable;

[0020] The independent variables and the candidate independent variables are all used as variables for collinearity check, and the variables with collinearity greater than the set value are eliminated to obtain the screened independent variables and candidate independent variables; the screened independent variables and candidate independent variables, as well as the corresponding dependent variables constitute data samples; each data sample constitutes a data set.

[0021] Optionally, collect remote sensing images, heavy metal content and spectral reflectance of each soil sample, including:

[0022] Collecting remote sensing images of the soil sample area, wherein the remote sensing images are WorldView-3 remote sensing images;

[0023] Collect multiple soil samples from the soil sample area, the soil collection area of ​​each soil sample is 30 cm×30 cm, the depth is 0-20 cm, and the number and geographic information of each soil sample are recorded, the geographic information includes the type of land feature, longitude and latitude, elevation and slope;

[0024] According to the geographic information of each soil sample, the remote sensing image of the corresponding space is matched for each soil sample;

[0025] The spectral reflectance of each soil sample was measured using an ASD FieldSpec ground spectrometer;

[0026] The heavy metal content of each soil sample was determined.

[0027] Optionally, the remote sensing image is preprocessed to obtain a plurality of soil heavy metal related indexes, specifically including:

[0028] Performing atmospheric correction, radiation correction, orthorectification and image enhancement on the remote sensing image in sequence to obtain an enhanced remote sensing image;

[0029] Extract reflectance of multiple bands from remote sensing images after image enhancement;

[0030] A plurality of soil heavy metal related indexes are extracted from the reflectance of a plurality of bands; the soil heavy metal related indexes include normalized difference vegetation index, enhanced vegetation index and soil water content.

[0031] Optionally, preprocessing the spectral reflectance to obtain the preprocessed spectral reflectance specifically includes:

[0032] Using a Savitzky-Golay filter to filter the spectral reflectance to obtain a filtered spectral reflectance;

[0033] The filtered spectral reflectance is normalized to obtain the preprocessed spectral reflectance.

[0034] Optionally, determining the predicted value of the heavy metal content at any position on the target soil area according to the spatial distribution of the prediction error and the multivariate stepwise regression model specifically includes:

[0035] The predicted value of the heavy metal content at any position is obtained by adding the difference between the initial predicted value of the heavy metal content at any position and the corresponding position in the spatial distribution of the prediction error.

[0036] In a second aspect, the present application provides a soil heavy metal content prediction device, the soil heavy metal content prediction device applies the soil heavy metal content prediction method, and the soil heavy metal content prediction device comprises:

[0037] A data set construction module is used to construct a data set for multiple regression of soil heavy metal content; the data set includes multiple candidate independent variables and one dependent variable; the dependent variable is the heavy metal content;

[0038] A multivariate stepwise regression model determination module is used to perform multivariate stepwise regression fitting using the data set to obtain a multivariate stepwise regression model;

[0039] A heavy metal content determination value determination module is used to determine a plurality of soil sampling points from the target soil area and obtain a heavy metal content determination value of each soil sampling point;

[0040] An initial heavy metal content prediction module is used to use the multivariate stepwise regression model to perform heavy metal content inversion on each of the soil sampling points to obtain a predicted value of the initial heavy metal content of each of the soil sampling points;

[0041] The error distribution determination module is used to calculate the difference between each heavy metal content measurement value and the initial heavy metal content prediction value at the corresponding position, and use the Kriging interpolation method to perform spatial interpolation on each of the differences to obtain the spatial distribution of the prediction error;

[0042] The final heavy metal content prediction value determination module is used to determine the heavy metal content prediction value at any position on the target soil area according to the spatial distribution of the prediction error and the multivariate stepwise regression model.

[0043] In a third aspect, the present application provides a computer device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of any of the above-described methods for predicting heavy metal content in soil.

[0044] In a fourth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of any of the above-described methods for predicting heavy metal content in soil.

[0045] In a fifth aspect, the present application provides a computer program product, including a computer program, which, when executed by a processor, implements the steps of any of the above-mentioned methods for predicting heavy metal content in soil.

[0046] According to the specific embodiments provided in this application, this application discloses the following technical effects:

[0047] The present application provides a method, device, equipment, medium and product for predicting the content of heavy metals in soil. Through multivariate stepwise regression, redundant variables are reduced, data collection costs are lowered, and the Kriging interpolation method is used to perform spatial interpolation on each difference to achieve correction of the prediction results of the multivariate stepwise regression model, thereby improving the accuracy of the inversion of the content of heavy metals in soil. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0049] Figure 1 A schematic diagram of a process for predicting heavy metal content in soil provided in one embodiment of the present application;

[0050] Figure 2 A schematic diagram of the principle of a method for predicting heavy metal content in soil provided in one embodiment of the present application;

[0051] Figure 3 A schematic diagram of the structure of a computer device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0052] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0053] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0054] This application provides a method for predicting soil heavy metal content. Figure 1 and Figure 2 As shown, the soil heavy metal content prediction method includes:

[0055] Step 101: construct a data set for multiple regression of soil heavy metal content; the data set includes multiple candidate independent variables and one dependent variable; the dependent variable is the heavy metal content.

[0056] Step 102: Perform multivariate stepwise regression fitting using the data set to obtain a multivariate stepwise regression model.

[0057] Step 103: Determine a plurality of soil sampling points in the target soil area, and obtain the measured value of the heavy metal content at each soil sampling point.

[0058] Step 104: using the multivariate stepwise regression model to invert the heavy metal content of each soil sampling point to obtain an initial predicted value of the heavy metal content of each soil sampling point.

[0059] Step 105: Calculate the difference between each heavy metal content measurement value and the initial heavy metal content prediction value at the corresponding position, and use the Kriging interpolation method to perform spatial interpolation on each of the differences to obtain the spatial distribution of the prediction error.

[0060] Step 106: Determine the predicted value of the heavy metal content at any location in the target soil area according to the spatial distribution of the prediction error and the multivariate stepwise regression model.

[0061] This application adopts a multivariate stepwise regression modeling method, which integrates multi-source data and avoids the defects of a single data source. The multivariate stepwise regression model is improved by using the Kriging method, which can more accurately invert the soil heavy metal content.

[0062] Multiple regression analysis is a statistical regression analysis method that studies the relationship between one or more independent variables. Multiple stepwise regression analysis is based on multiple regression analysis, but it is superior to general multiple regression methods. The basic idea of ​​this method is to consider the contribution of all variables to the regression equation, perform significance tests on each variable that enters the equation according to its importance, and then eliminate insignificant variables. The addition of new variables may change the significance of previous variables, and the eliminated variables may also become significant due to the addition of new variables. At this time, add them back to the equation until there are no more variables that can be eliminated or added again, and the calculation ends.

[0063] In an exemplary embodiment, step 101 specifically includes:

[0064] Collect remote sensing images, heavy metal content and spectral reflectance of each soil sample.

[0065] The remote sensing image is preprocessed to obtain a plurality of soil heavy metal related indexes.

[0066] The remote sensing image is subjected to a KT transformation to obtain the brightness component of the remote sensing image, specifically including: performing a KT transformation on the WorldView-3 image to obtain the brightness, greenness and wetness components. Since the brightness component is related to soil reflectivity, the brightness component is retained as a factor to participate in the inversion of soil heavy metals.

[0067] Extracting the reflectance of multiple bands from the remote sensing image and performing principal component analysis (PCA transformation) to obtain multiple principal component bands, specifically including: performing PCA transformation on the reflectance of 16 bands extracted from the WorldView-3 remote sensing image, and selecting principal components with cumulative variance contribution rates greater than 85%.

[0068] The spectral reflectance is preprocessed to obtain a preprocessed spectral reflectance.

[0069] The heavy metal content is subjected to an outlier elimination process to obtain a pre-processed heavy metal content.

[0070] For each soil sample, the first-order differential of the spectral reflectance after pretreatment was used as the independent variable, the brightness component, each principal component band and each soil heavy metal related index were used as candidate independent variables, and the heavy metal content after pretreatment was used as the dependent variable.

[0071] The independent variables and the candidate independent variables are all used as variables for collinearity check, and the variables with collinearity greater than the set value are eliminated to obtain the screened independent variables and candidate independent variables; the screened independent variables and candidate independent variables, as well as the corresponding dependent variables constitute data samples; each data sample constitutes a data set.

[0072] The candidate independent variables are screened by multivariate stepwise regression fitting, and the first-order differential of the spectral reflectance and the screened independent variables constitute the independent variables of the multivariate stepwise regression model.

[0073] In an exemplary embodiment, collecting remote sensing images, heavy metal content, and spectral reflectance of each soil sample includes:

[0074] A remote sensing image of a soil sample area is collected, wherein the remote sensing image is a WorldView-3 remote sensing image. The soil sample area is a part of the target soil area.

[0075] Multiple soil samples were collected from the soil sample area according to the principle of random distribution in blocks. The soil sample types covered a variety of land features. In order to correspond to the spatial resolution of the WorldView-3 remote sensing image, the soil collection area of ​​each soil sample was 30cm×30cm and the depth was 0-20cm. The number and geographic information of each soil sample were recorded during the sampling process. The geographic information included land feature type, longitude and latitude, elevation and slope. According to the geographic information of each soil sample, the remote sensing image of the corresponding space was matched for each soil sample.

[0076] The spectral reflectance of each soil sample was measured using an ASD FieldSpec spectrometer. Specifically, the soil samples were air-dried into powder after collection and spectrally measured indoors using an ASD FieldSpec spectrometer. Each soil sample was measured at least three times, and the average of the three spectral data was taken as the spectral reflectance of the soil at that point.

[0077] Determine the heavy metal content of each soil sample, specifically including selecting representative soil samples according to the type of landform and sending them to professional testing institutions to determine the content of heavy metal elements such as As, Cd, Cr, Hg, Pb, and Zn in the soil samples.

[0078] In an exemplary embodiment, the remote sensing image is preprocessed to obtain a plurality of soil heavy metal related indexes, specifically including:

[0079] The remote sensing image is subjected to atmospheric correction, radiation correction, orthorectification and image enhancement in sequence to obtain an image-enhanced remote sensing image.

[0080] The reflectance of multiple bands is extracted from the remote sensing image after image enhancement, specifically including: using the remote sensing image processing platform (ENVI) to extract the reflectance of 16 bands of the WorldView-3 remote sensing image.

[0081] Taking full account of the band characteristics of remote sensing images, multiple soil heavy metal related indices are extracted from the reflectance of multiple bands; the soil heavy metal related indices include normalized difference vegetation index (NDVI), enhanced vegetation index (EVI) and soil moisture content.

[0082] In an exemplary embodiment, preprocessing the spectral reflectance to obtain the preprocessed spectral reflectance specifically includes:

[0083] The spectral reflectance is filtered by using a Savitzky-Golay filter to obtain the filtered spectral reflectance, thereby reducing noise in the spectral data, obtaining smoothed data, and being able to maintain the shape and characteristic peaks of the spectral curve.

[0084] The filtered spectral reflectance is normalized to obtain the preprocessed spectral reflectance, making the spectral data of different bands comparable.

[0085] This application also includes: using ArcGIS software to spatially match field-measured soil spectral data, heavy metal content determination data, and remote sensing image data to ensure that the three types of data can correspond spatially in the construction of a multivariate stepwise regression model.

[0086] In an exemplary embodiment, step 102 specifically includes:

[0087] 1) Model fitting: The data set after preprocessing and variable selection is divided into a training set and a validation set in a ratio of 7:3. The training set data is used to fit the multivariate stepwise regression model. The expression of the multivariate stepwise regression model is:

[0088] Y=b 0 +b 1 x 1 +b 2 x 2 +…+b n x n .

[0089] Among them, Y is the dependent variable, xi (i=1…n) is n independent variables, b i (i=1…n) is the regression coefficient of the equation, b 0 is the constant term of the equation. In fact, the sample data is used to calculate the values ​​of each regression coefficient, and then a linear equation with n independent variables is obtained.

[0090] 2) Stepwise regression: The multivariate stepwise regression method is to find the spectral characteristic bands with good correlation with heavy metal elements based on the correlation analysis between soil heavy metal content and soil reflectance spectrum, and perform multivariate regression analysis on the spectral variables of each heavy metal content and characteristic band. This application is implemented using SPSS software, and the main steps are as follows:

[0091] a. Determine the F test value. The F test can determine whether each variable is significant and use this as the basis for the variable to enter or exclude the equation. The size of the F test value is determined based on the actual situation.

[0092] b. According to the correlation between the independent variable and the dependent variable, enter the regression equation one by one according to their contribution level from large to small. Whenever a new variable enters the equation, its partial regression sum of squares must be calculated, and then the variable with the smallest partial regression sum of squares is selected for F test. If the result is significant, the variable is retained, and other variables in the equation do not need to be eliminated. Otherwise, the variable is eliminated, and the other variables are subjected to F test according to the size of the partial regression sum of squares.

[0093] Evaluate the multivariate stepwise regression model: the evaluation index mainly uses the determination coefficient R 2 , root mean square error RMSE, mean error ME and mean relative error MRE. In addition, the significance test of the regression equation is also an indicator to test the quality of the regression model. The significance test of the regression equation is to test whether the linear relationship between all independent variables and dependent variables of the entire regression equation is close. The F test is often used. The calculation formulas of the evaluation index calculation formula are as follows. According to the principles of the highest regression coefficient and F statistic, the highest R-square, and the smallest root mean square error, the best regression model for soil heavy metal hyperspectral remote sensing monitoring is selected.

[0094]

[0095] in, represents the i-th predicted value, represents the mean value, n represents the sample size, k represents the number of independent variables, Y i represents the ith actual value, Represents the predicted value.

[0096] In an exemplary embodiment, step 104 specifically includes: applying the established and evaluated multivariate stepwise regression model to the entire area, that is, the entire target soil area, calculating the eigenvalue of each pixel in the remote sensing image, where the eigenvalue is the independent variable in the evaluated multivariate stepwise regression model, and inverting the content of heavy metal elements such as As, Cd, Cr, Hg, Pb, and Zn in the soil of the entire area.

[0097] Step 105 specifically includes: taking the difference between the soil heavy metal content measured value and the multivariate stepwise regression model predicted value as a variable, constructing a semivariogram, and using the Kriging interpolation method to perform spatial interpolation on the difference. The interpolated difference map can reflect the spatial distribution law of the model prediction error.

[0098] In an exemplary embodiment, step 106 specifically includes: adding the initial heavy metal content prediction value at any position to the difference value at the corresponding position in the spatial distribution of the prediction error to obtain the heavy metal content prediction value at any position, that is, the improved soil heavy metal content prediction value. The improved model combines the advantages of the multivariate stepwise regression model and the Kriging interpolation method, and can more accurately invert the soil heavy metal content.

[0099] Draw distribution maps: Apply the optimized model to the entire region and use ArcGIS to draw the spatial distribution of heavy metal content in the soil to provide data support for soil quality assessment, ecological security and other work.

[0100] This application fully considers the dependency between variables, adopts methods such as spike-cap transformation and principal component analysis, and uses multivariate stepwise regression to eliminate unnecessary variables, so as to obtain the optimal or most appropriate regression model, thereby improving the accuracy of soil heavy metal content inversion; the traditional multivariate stepwise regression model is improved by using the Kriging method, thereby improving the accuracy of soil heavy metal content inversion.

[0101] This application fully considers the correlation between characteristic bands in the process of establishing the soil heavy metal content model, and constructs the most streamlined and optimized inversion model, which can quickly and efficiently make reasonable spatial predictions of soil heavy metal content in a large area.

[0102] There are many factors that affect the accuracy of soil heavy metal content inversion, and the monitoring difficulty is greater than that of the surface, so a reliable model is very critical. This application makes full use of the brightness component after KT transformation, the principal component after PCA transformation, and NDVI, EVI, soil moisture content, soil organic matter content and measured soil heavy metal content, and uses multivariate stepwise regression and Kriging interpolation methods to construct a high-precision soil heavy metal inversion model. This model can better obtain the complex relationship between soil heavy metal content and multiple factors, which is also a major improvement on the single model.

[0103] Based on the same inventive concept, the embodiment of the present application also provides a soil heavy metal content prediction device for implementing the soil heavy metal content prediction method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more soil heavy metal content prediction device embodiments provided below can refer to the limitations of the soil heavy metal content prediction method above, and will not be repeated here.

[0104] In an exemplary embodiment, the present application provides a soil heavy metal content prediction device, the soil heavy metal content prediction device applies the soil heavy metal content prediction method, and the soil heavy metal content prediction device includes:

[0105] The data set construction module is used to construct a data set for multiple regression of soil heavy metal content; the data set includes multiple candidate independent variables and one dependent variable; the dependent variable is the heavy metal content.

[0106] The multivariate stepwise regression model determination module is used to perform multivariate stepwise regression fitting using the data set to obtain a multivariate stepwise regression model.

[0107] The heavy metal content measurement value determination module is used to determine multiple soil sampling points from the target soil area and obtain the heavy metal content measurement value of each soil sampling point.

[0108] The initial heavy metal content prediction module is used to use the multivariate stepwise regression model to invert the heavy metal content of each soil sampling point to obtain the initial heavy metal content prediction value of each soil sampling point.

[0109] The error distribution determination module is used to calculate the difference between each heavy metal content measurement value and the initial heavy metal content prediction value of the corresponding position, and use the Kriging interpolation method to spatially interpolate each of the differences to obtain the spatial distribution of the prediction error.

[0110] The final heavy metal content prediction value determination module is used to determine the heavy metal content prediction value at any position on the target soil area according to the spatial distribution of the prediction error and the multivariate stepwise regression model.

[0111] In an exemplary embodiment, a computer device is provided. The computer device may be a server or a terminal. The internal structure diagram thereof may be as follows: Figure 3As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. The processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store soil heavy metal content prediction data. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a soil heavy metal content prediction method is implemented.

[0112] Those skilled in the art will understand that Figure 3 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components. In an exemplary embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above-mentioned method embodiments when executing the computer program.

[0113] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0114] In an exemplary embodiment, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0115] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0116] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM may be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0117] The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. The non-relational database may include a distributed database based on blockchain, etc., but is not limited thereto. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a data processing logic of a programmable logic device, etc., but is not limited thereto.

[0118] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0119] This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method and core ideas of this application. At the same time, for those skilled in the art, according to the ideas of this application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.

Claims

1. A method for predicting heavy metal content in soil, characterized in that: The method for predicting soil heavy metal content comprises: Constructing a data set for multiple regression of soil heavy metal content; the data set includes multiple candidate independent variables and one dependent variable; the dependent variable is the heavy metal content; The data set is used to perform multivariate stepwise regression fitting to obtain a multivariate stepwise regression model; Determine a plurality of soil sampling points from the target soil area, and obtain the measured value of the heavy metal content at each soil sampling point; The multivariate stepwise regression model is used to invert the heavy metal content of each soil sampling point to obtain the initial heavy metal content prediction value of each soil sampling point; Calculating the difference between each heavy metal content measurement value and the initial heavy metal content prediction value at the corresponding position, and performing spatial interpolation on each of the differences using the Kriging interpolation method to obtain the spatial distribution of the prediction error; According to the spatial distribution of prediction errors and the multivariate stepwise regression model, the predicted value of heavy metal content at any location in the target soil area is determined.

2. The method for predicting soil heavy metal content according to claim 1, characterized in that: Construct a data set for multiple regression of soil heavy metal content, including: Collect remote sensing images, heavy metal content and spectral reflectance of each soil sample; Preprocessing the remote sensing image to obtain multiple soil heavy metal related indexes; Performing a tassel-cap transformation on the remote sensing image to obtain a brightness component of the remote sensing image; Extracting reflectance of multiple bands from the remote sensing image and performing principal component analysis to obtain multiple principal component bands; Preprocessing the spectral reflectance to obtain a preprocessed spectral reflectance; Performing outlier elimination processing on the heavy metal content to obtain the heavy metal content after pretreatment; For each soil sample, the first-order differential of the spectral reflectance after pretreatment is used as an independent variable, the brightness component, each principal component band and each soil heavy metal related index are used as candidate independent variables, and the heavy metal content after pretreatment is used as a dependent variable; The independent variables and the candidate independent variables are all used as variables for collinearity check, and the variables with collinearity greater than the set value are eliminated to obtain the screened independent variables and candidate independent variables; the screened independent variables and candidate independent variables, as well as the corresponding dependent variables constitute data samples; each data sample constitutes a data set.

3. The method for predicting soil heavy metal content according to claim 2, characterized in that: Collect remote sensing images, heavy metal content and spectral reflectance of each soil sample, including: Collecting remote sensing images of the soil sample area, wherein the remote sensing images are WorldView-3 remote sensing images; Collect multiple soil samples from the soil sample area, the soil collection area of ​​each soil sample is 30 cm×30 cm, the depth is 0-20 cm, and the number and geographic information of each soil sample are recorded, the geographic information includes the type of land feature, longitude and latitude, elevation and slope; According to the geographic information of each soil sample, the remote sensing image of the corresponding space is matched for each soil sample; The spectral reflectance of each soil sample was measured using an ASD FieldSpec ground spectrometer; The heavy metal content of each soil sample was determined.

4. The method for predicting soil heavy metal content according to claim 2, characterized in that: The remote sensing image is preprocessed to obtain multiple soil heavy metal related indexes, including: Performing atmospheric correction, radiation correction, orthorectification and image enhancement on the remote sensing image in sequence to obtain an enhanced remote sensing image; Extract reflectance of multiple bands from remote sensing images after image enhancement; A plurality of soil heavy metal related indexes are extracted from the reflectance of a plurality of bands; the soil heavy metal related indexes include normalized difference vegetation index, enhanced vegetation index and soil water content.

5. The method for predicting soil heavy metal content according to claim 2, characterized in that: Preprocessing the spectral reflectance to obtain the preprocessed spectral reflectance specifically includes: Using a Savitzky-Golay filter to filter the spectral reflectance to obtain a filtered spectral reflectance; The filtered spectral reflectance is normalized to obtain the preprocessed spectral reflectance.

6. The method for predicting soil heavy metal content according to claim 1, characterized in that: Determining the predicted value of heavy metal content at any location on the target soil area according to the spatial distribution of the prediction error and the multivariate stepwise regression model, specifically includes: The predicted value of the heavy metal content at any position is obtained by adding the difference between the initial predicted value of the heavy metal content at any position and the corresponding position in the spatial distribution of the prediction error.

7. A soil heavy metal content prediction device, characterized in that: The soil heavy metal content prediction device applies the soil heavy metal content prediction method according to any one of claims 1 to 6, and the soil heavy metal content prediction device comprises: A data set construction module is used to construct a data set for multiple regression of soil heavy metal content; the data set includes multiple candidate independent variables and one dependent variable; the dependent variable is the heavy metal content; A multivariate stepwise regression model determination module is used to perform multivariate stepwise regression fitting using the data set to obtain a multivariate stepwise regression model; A heavy metal content determination value determination module is used to determine a plurality of soil sampling points from the target soil area and obtain a heavy metal content determination value of each soil sampling point; An initial heavy metal content prediction module is used to use the multivariate stepwise regression model to perform heavy metal content inversion on each of the soil sampling points to obtain a predicted value of the initial heavy metal content of each of the soil sampling points; The error distribution determination module is used to calculate the difference between each heavy metal content measurement value and the initial heavy metal content prediction value at the corresponding position, and use the Kriging interpolation method to spatially interpolate each of the differences to obtain the spatial distribution of the prediction error; The final heavy metal content prediction value determination module is used to determine the heavy metal content prediction value at any position on the target soil area according to the spatial distribution of the prediction error and the multivariate stepwise regression model.

8. A computer device comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the method for predicting the heavy metal content in soil according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for predicting the heavy metal content in soil according to any one of claims 1 to 6 is implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the method for predicting the heavy metal content in soil according to any one of claims 1 to 6 is implemented.

Images