A method, equipment and system for determining the physicochemical properties of terraced soil

Abstract

This invention relates to the field of soil monitoring technology, specifically to a method, equipment, and system for determining the physicochemical properties of terraced soil. The method involves acquiring spectral data from various monitoring points in the terraced fields at different times and in both in-situ and experimental environments. Given the spatial heterogeneity of the physicochemical properties of terraced soils, the method analyzes the intensity differences and similarities of spectral peaks at different monitoring points under each environment, determining similarity indicators and clustering them to obtain clusters. Since soil nutrient changes are reflected in the spectral data, the method analyzes the fluctuation characteristics of spectral peak differences at different times within each cluster under each environment to determine soil nutrient fluctuation indicators. Because the experimental environment is relatively controllable, a target straight line is fitted based on the experimental data as a reference for in-situ spectral correction. Finally, by combining the differences and numerical characteristics of soil nutrient fluctuation indicators at monitoring points within the clusters under in-situ conditions, in-situ environmental parameters, in-situ spectra, and the target straight line, the in-situ spectrum is corrected, thereby improving the accuracy of soil nutrient element determination.

Description

Technical Field

[0001] This invention relates to the field of soil monitoring technology, specifically to a method, equipment, and system for determining the physicochemical properties of terraced soil. Background Technology

[0002] Terraced fields, as the core agricultural ecosystem for soil and water conservation and food production in mountainous and hilly areas, have soil physicochemical properties (such as organic matter content, pH, bulk density, porosity, and nutrient content) that directly determine crop growth suitability, soil fertility maintenance, and ecosystem stability. They are key basic data for guiding precise fertilization, soil and water conservation optimization, and sustainable agricultural development in terraced fields.

[0003] Current technologies for measuring the physicochemical properties of terraced soils typically employ spectral analysis, randomly sampling to obtain soil spectral data at monitoring points and directly performing spectral analysis. However, in real-world scenarios, terraced fields exhibit significant spatial heterogeneity in soil physicochemical properties due to their undulating terrain (significant differences in slope and altitude), diverse planting patterns (crop rotation, intercropping, etc.), and uneven intensity of human intervention (differences in fertilization and cultivation methods). These environmental factors can interfere with spectral data, further increasing its complexity and uncertainty. Therefore, directly analyzing the collected spectral data of terraced soils often fails to accurately reflect the true physicochemical properties of the soil, ultimately leading to significant errors in the measurement results. Summary of the Invention

[0004] To address the significant spatial heterogeneity of soil physicochemical properties in terraced fields due to factors such as undulating terrain, diverse planting patterns, and uneven human intervention, which interfere with spectral data and further increase its complexity and uncertainty, directly analyzing collected terraced soil spectral data often fails to accurately reflect the true physicochemical properties of the soil, ultimately leading to substantial errors in the measurement results. This invention aims to provide a method, equipment, and system for measuring the physicochemical properties of terraced soils. The specific technical solution adopted is as follows:

[0005] A method for determining the physicochemical properties of terraced soil, comprising:

[0006] Spectral data of various monitoring points in the terraced fields at multiple time points under different environments were obtained. The environment included in-situ environment and experimental environment, and the types of spectral data included in-situ spectra and experimental spectra.

[0007] In each environment, the intensity differences and similarities between spectral peaks in the spectral data of different monitoring points are analyzed to determine the change similarity index between monitoring points. In each environment, cluster analysis is performed on all monitoring points based on the change similarity index between monitoring points to obtain clusters. In each cluster in each environment, the fluctuation characteristics of the spectral peak differences of the monitoring points at different time points are analyzed to determine the soil nutrient fluctuation index at each monitoring point.

[0008] The target straight line is fitted based on the soil nutrient fluctuation index, experimental spectrum, and environmental parameters of the monitoring points under the experimental environment. Based on the difference and numerical characteristics of the soil nutrient fluctuation index at the monitoring points in the in-situ cluster, the environmental parameters, in-situ spectrum, and target straight line of the in-situ environment, the in-situ spectrum of the monitoring points is corrected, and the spectral correction data is used to determine soil nutrient elements.

[0009] Furthermore, the method for obtaining the similarity index includes:

[0010] In each environment, the intensity difference of spectral peaks in the spectral data of any two monitoring points at the last time node is analyzed to determine the spectral peak matching index between the two monitoring points.

[0011] In each environment, the similarity between the spectral data of any two monitoring points at the last time node is analyzed to obtain the change similarity factor between the two monitoring points;

[0012] In each environment, the normalized value of the product of the spectral peak matching index and the change similarity factor between any two monitoring points is used as the change similarity index between the spectral data of the two monitoring points in each environment.

[0013] Furthermore, the method for obtaining the spectral peak matching index includes:

[0014] In each environment, for any two monitoring points, compare the spectral peaks in the corresponding spectral data of the two monitoring points at the last time node, match the spectral peaks with the same wavelength as spectral peak pairs, take the difference of the peak values ​​of each spectral peak pair as the peak difference factor of each spectral peak pair, and take the sum of the peak difference factors of all spectral peak pairs as the peak difference parameter of the two monitoring points.

[0015] The number of peak pairs at the two monitoring points is used as the numerator, the sum of the peak difference parameter between the two monitoring points and the preset constant is used as the denominator, and the normalized value of the resulting ratio is used as the peak matching index between the two monitoring points.

[0016] Furthermore, the method for obtaining the change similarity factor includes:

[0017] In each environment, for any two monitoring points, the DTW values ​​of the corresponding spectral data of these two monitoring points at the last time node are negatively correlated and normalized to obtain the similarity factor of the change of spectral data of these two monitoring points under each environmental parameter.

[0018] Furthermore, the method for obtaining the clusters includes:

[0019] Under each environmental parameter, based on the DBSCAN clustering algorithm and the similarity index of changes between monitoring points under the corresponding spectral data, cluster analysis is performed on all monitoring points to obtain clusters of all monitoring points under each environmental parameter. The cluster radius and the minimum number of clusters are preset values.

[0020] Furthermore, the method for obtaining the soil nutrient fluctuation index includes:

[0021] In each cluster under each environment, the spectral peaks of the spectral data of each monitoring point under each environment are compared in the spectral data of each adjacent two time nodes. The absolute value of the difference between the peaks under each same wavelength is calculated as the deviation factor. The mean of all deviation factors under the same wavelength is used as the spectral difference factor of each monitoring point under each adjacent two time nodes.

[0022] The variance of all spectral difference factors corresponding to each monitoring point is used as the difference fluctuation parameter. The value of the sum of the difference fluctuation parameter and the mean of the spectral difference factors is normalized and used as the soil nutrient fluctuation factor for each monitoring point.

[0023] The mean value of the soil nutrient fluctuation factor corresponding to all monitoring points in each cluster is used as the soil nutrient fluctuation index at each monitoring point in each cluster.

[0024] Furthermore, the method for obtaining the target straight line includes:

[0025] In the experimental environment, the experimental environmental parameters at each monitoring point are obtained. The experimental environmental parameters include at least the soil surface moisture content and light intensity.

[0026] In the experimental environment, the mean intensity value, peak value, and half-peak width of the experimental spectrum at the last time node of each monitoring point are obtained, and the experimental fusion vector is formed with the experimental environmental parameters and corresponding soil nutrient fluctuation index at each monitoring point.

[0027] The fused vectors of all monitoring points in the experimental environment are used as the initial training set, and the outputs of the monitoring points are obtained using the WPLASR model. Based on the outputs of the monitoring points and the fused vectors of the monitoring points, the least squares method is used to fit a straight line to obtain the target straight line.

[0028] Furthermore, the method for acquiring the spectral correction data includes:

[0029] In the in-situ environment, within each cluster, the difference between the maximum and minimum soil nutrient fluctuation indices is negatively correlated and normalized, and this normalized value is used as a dynamic constraint parameter.

[0030] The value obtained by negatively mapping the product of the dynamic constraint parameters and the soil nutrient fluctuation index at each monitoring point in the in-situ environment is used as the adjustment coefficient;

[0031] In the in-situ environment, the in-situ environmental parameters at each monitoring point are obtained, and the in-situ environmental parameters include at least the soil surface moisture content and light intensity.

[0032] In the in-situ environment, the mean intensity value, peak value, and half-peak width are obtained from the in-situ spectrum at the last time node of each monitoring point, and are combined with the in-situ environmental parameters and corresponding soil nutrient fluctuation indicators at each monitoring point to form an in-situ fusion vector.

[0033] Substitute the in-situ fusion vector into the target line, multiply the output by the adjustment coefficient, and use the resulting product as the correction coefficient for each monitoring point.

[0034] At each monitoring point, the sum of the correction coefficient and the preset parameter is used as the adjustment degree value. The adjustment degree value is multiplied by the intensity value at each wavelength in the in-situ spectrum to obtain the corrected intensity at each wavelength, thereby obtaining the spectral correction data at each monitoring point.

[0035] A system for measuring the physicochemical properties of terraced soil includes:

[0036] The data acquisition module is used to acquire spectral data of various monitoring points in the terraced fields at multiple time points under different environments. The environment includes in-situ environment and experimental environment, and the types of spectral data include in-situ spectra and experimental spectra.

[0037] The spectral analysis module is used to analyze the intensity differences and similarities between spectral peaks in the spectral data of different monitoring points under each environment, and to determine the change similarity index between monitoring points. Under each environment, cluster analysis is performed on all monitoring points based on the change similarity index between monitoring points to obtain clusters. In each cluster under each environment, the fluctuation characteristics of the spectral peak differences of the monitoring points at different time points are analyzed to determine the soil nutrient fluctuation index at each monitoring point.

[0038] The spectral correction module is used to fit a target straight line based on the soil nutrient fluctuation index, experimental spectrum, and environmental parameters of the experimental environment at the monitoring points under experimental conditions. Based on the difference and numerical characteristics of the soil nutrient fluctuation index at the monitoring points in the cluster under in-situ conditions, the environmental parameters of the in-situ environment, the in-situ spectrum, and the target straight line, the in-situ spectrum of the monitoring points is corrected to obtain spectral correction data for the determination of soil nutrients.

[0039] A device for measuring the physical and chemical properties of terraced soil includes a processor and a memory. The memory stores at least one instruction, at least one program, code set, or instruction set. When the processor loads and executes the at least one instruction, at least one program, code set, or instruction set, it implements the steps of a method for measuring the physical and chemical properties of terraced soil.

[0040] The present invention has the following beneficial effects:

[0041] Spectral data from various monitoring points in the terraced fields were acquired at multiple time points under different environments, encompassing both in-situ and experimental environments, as well as in-situ and experimental spectra. In-situ environmental data accurately reflects the spectral characteristics of the terraced soil under natural conditions, while experimental environmental data helps to eliminate some interfering factors. Due to the spatial heterogeneity of the physicochemical properties of the terraced soil, similarity intensities and similarities between spectral peaks in the spectral data from different monitoring points under each environment were analyzed to determine similarity indices, quantifying the degree of similarity between monitoring points. Based on these similarity indices, cluster analysis was performed on all monitoring points to obtain clusters, grouping monitoring points with similar variation characteristics into one category, thus better reflecting the spatial distribution patterns of the terraced soil. Since changes in soil nutrients are directly reflected in the spectral data, the fluctuations in soil nutrients over time were considered under each environment: the peak difference fluctuation characteristics of the spectral data from monitoring points in the clusters at different time points were analyzed to determine soil nutrient fluctuation indices. Furthermore, since the experimental environment is relatively controllable and can eliminate some interfering factors, a target straight line is fitted based on the soil nutrient fluctuation index, experimental spectrum, and environmental parameters at the monitoring points under the experimental environment. This target straight line can serve as a reference standard for in-situ spectral correction. Finally, based on the differences and numerical characteristics of the soil nutrient fluctuation index at the monitoring points in the cluster under the in-situ environment, the environmental parameters, in-situ spectrum, and target straight line, the in-situ spectrum of the monitoring points is corrected. This can more accurately eliminate the interference of the in-situ environment on the spectral data, obtain more accurate spectral correction data, and thus improve the accuracy of soil nutrient element measurement results. Attached Figure Description

[0042] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0043] Figure 1 This is a flowchart of a method for determining the physicochemical properties of terraced soil according to an embodiment of the present invention;

[0044] Figure 2 A flowchart illustrating a method for obtaining a similarity index according to an embodiment of the present invention;

[0045] Figure 3 This is a system block diagram of a system for measuring the physicochemical properties of terraced soil provided in one embodiment of the present invention;

[0046] Figure 4 This is a schematic diagram of the equipment structure for measuring the physical and chemical properties of terraced soil, provided in one embodiment of the present invention. Detailed Implementation

[0047] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a method, equipment, and system for determining the physicochemical properties of terraced soil according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0048] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0049] The following description, in conjunction with the accompanying drawings, details the specific scheme of the method, equipment, and system for determining the physical and chemical properties of terraced soil provided by this invention.

[0050] Please see Figure 1 The diagram illustrates a flowchart of a method for determining the physicochemical properties of terraced soil according to an embodiment of the present invention. The method includes the following steps:

[0051] Step S1: Obtain spectral data at multiple time points under different environments at various monitoring points in the terraced fields. The environment includes in-situ environment and experimental environment, and the types of spectral data include in-situ spectra and experimental spectra.

[0052] First, topographic parameters such as terrace slope (α), elevation (H), and field boundaries are extracted based on a digital elevation model (DEM). Primary sampling units are then defined according to slope gradient (gentle slope α≤5°, medium slope 5°<α≤15°, steep slope α>15°) and planting pattern (food crops, cash crops, fallow land). Within each primary sampling unit, secondary sub-units are defined based on micro-topographic differences between the field ridge, the center of the field, and the edge of the irrigation and drainage ditch. A "quincunx pattern + key site reinforcement" strategy is adopted, with conventional sampling points (monitoring points) spaced 20-30m apart in the center of the field, and additional reinforced sampling points (monitoring points) added on both sides of the field ridge and along the edge of the irrigation and drainage ditch to ensure coverage of the core area of ​​spatial heterogeneity in the terraced soil. In this embodiment, all monitoring points in subsequent analyses originate from the same primary sampling unit.

[0053] A portable ground-based spectrometer (band range 350-2500nm, spectral resolution ≤3nm) was selected. Before data collection, spectral calibration was performed using a standard white board (repeated once for every 10 sampling points). During data collection, the "three-point method" was used: remove dead leaves, fallen branches, and gravel from the soil surface at the monitoring point, level the measurement surface (area ≥10cm×10cm) with a soil shovel, and place the spectrometer probe perpendicularly 15-20cm above the measurement surface. Spectral data were collected at the center of the measurement surface and 5cm on both sides. Three reflectance spectra were collected at each site, and the average value was calculated as the in-situ spectrum at a single time point for that monitoring point. Environmental parameters of the in-situ environment during data collection were recorded simultaneously, including at least light intensity (L, unit lux) and soil surface water content (θ, unit %). A portable sunshade was used to reduce spectral fluctuation interference caused by direct sunlight and cloud cover.

[0054] Then, at each monitoring point, soil samples were collected in layers using a soil auger, with the soil layer divided into 0-20cm (cultivated layer) and 20-40cm (subcultivated layer). Three replicate samples were collected from each layer and mixed thoroughly. The samples were placed in sealed bags and labeled with the sampling point number, depth, topographic parameters, planting pattern, and other information. After the samples were brought back to the laboratory, impurities were removed, and the samples were ground and sieved through a 2mm sieve. 50g of the processed samples were placed in a spectrometer dish (1cm thick, 5cm in diameter), and laboratory spectral acquisition was performed using an integrating sphere attachment. The spectrometer was set to scan 30 times and integrate for 100ms. Spectra were collected twice from both sides of each sample. After removing outliers, the average value was calculated as the experimental spectrum at a single time point for that monitoring point. At the same time, the environmental parameters of the experimental environment for each sample were obtained, including at least light intensity (L, unit lux) and soil surface moisture content (θ, unit %).

[0055] In-situ and laboratory spectra collected under in-situ and experimental conditions were preprocessed to remove noise bands (350-400nm, 2400-2500nm), and random noise was eliminated using the Savitzky-Golay smoothing method. Spectral interference caused by soil particle size and sample density was corrected by standard normal variable transformation (SNV), providing high-quality data support for the subsequent spectral analysis of the physicochemical properties of terraced soils.

[0056] Step S2: Under each environment, analyze the intensity differences and similarities between spectral peaks in the spectral data of different monitoring points to determine the change similarity index between monitoring points; under each environment, perform cluster analysis on all monitoring points based on the change similarity index between monitoring points to obtain clusters; in each cluster under each environment, analyze the fluctuation characteristics of the spectral peak differences of the monitoring points at different time points to determine the soil nutrient fluctuation index at each monitoring point.

[0057] Measuring the physicochemical properties of terraced soil can accurately reveal the spatial distribution characteristics of core indicators such as soil organic matter content, pH, nutrient concentration, bulk density, and porosity, providing a scientific basis for optimizing the planting layout of terraced crops, formulating precise fertilization plans, and adjusting farming systems. Therefore, it is necessary to measure the nutrient content of terraced soil to determine the differences in nutrients in different areas, thereby planning the types of crops to be planted.

[0058] However, terraced fields have complex terrain, and different areas retain varying amounts of nutrients, as do different levels of sunlight exposure and leaf fall. Therefore, even monitoring points within the same primary sampling unit exhibit significant spatial heterogeneity. Cluster analysis can classify monitoring points into different categories based on their similarity, grouping those with similar soil types together, thus simplifying the complex spatial distribution of soil. Therefore, cluster analysis can be performed on monitoring points under each environment. Before doing so, it is necessary to first determine the similarity between the monitoring points.

[0059] Spectral data contains rich information about soil; different soil components and properties exhibit unique intensity and positional characteristics on spectral peaks. By analyzing the intensity differences and similarities between spectral peaks, the similarities and differences between soils at different monitoring points can be directly reflected from a spectral perspective. Therefore, under each environment, analyzing the intensity differences and similarities between spectral peaks in the spectral data of different monitoring points allows for the determination of similarity indicators between monitoring points, which reflect the similarity between them.

[0060] Preferably, in one embodiment of the present invention, the method for obtaining the similarity index includes:

[0061] Please see Figure 2The diagram illustrates a method flowchart for obtaining a similarity index in one embodiment of the present invention, which includes the following steps:

[0062] Step S201: Under each environment, analyze the intensity difference of spectral peaks in the spectral data of any two monitoring points at the last time node, and determine the spectral peak matching index between the two monitoring points.

[0063] In each environment (referring to experimental and in-situ environments), for any two monitoring points, the spectral peaks in the corresponding spectral data at the last time node are compared. Peaks with the same wavelength are matched and considered as peak pairs. Peaks with the same wavelength usually correspond to the same components or properties in the soil. In this embodiment of the invention, due to the influence of instrument resolution, sample matrix effect, and ambient temperature, the characteristic peak wavelength of the same substance may drift. Therefore, strictly requiring the same wavelength is difficult to achieve in practical implementation scenarios. Thus, in this embodiment of the invention, wavelength errors within 3 units of wavelength are considered as the same wavelength.

[0064] Then, the difference in peak values ​​of each peak pair is used as the peak difference factor for each peak pair. The sum of the peak difference factors of all peak pairs is used as the peak difference parameter for the two monitoring points. The peak difference factor reflects the peak difference characteristics of the two monitoring points on each peak pair. The larger the value, the greater the difference between the two monitoring points in terms of specific soil composition or properties. The peak difference parameter integrates the peak difference factors of all peak pairs. The larger the value, the more significant the difference in the overall spectral properties between the two monitoring points.

[0065] Based on the foregoing analysis, it is known that the peak difference parameter is negatively correlated with the degree of similarity between monitoring points, while the number of peak pairs indicates the number of similar components or properties. Therefore, the number of peak pairs should be positively correlated with the degree of similarity between monitoring points. Thus, the number of peak pairs between these two monitoring points is used as the numerator, and the sum of the peak difference parameter between these two monitoring points and a preset constant is used as the denominator to achieve the aforementioned logic. The normalized value of the resulting ratio is then used as the peak matching index between the two monitoring points. A larger peak matching index indicates a higher degree of similarity in the spectral data between the two monitoring points. Normalization is a technique well-known to those skilled in the art. The normalization function can be linear normalization or standard normalization, etc., and the specific normalization method is not limited here.

[0066] It should be noted that the preset constant is used to prevent the denominator from being 0, and its value can be 0.001. The specific value can be adjusted according to the implementation scenario, and is not limited here. Spectral peaks can be obtained through spectral peak identification, which is a well-known technology and will not be elaborated here.

[0067] Step S202: Under each environment, analyze the similarity between the spectral data of any two monitoring points at the last time node to obtain the change similarity factor between the two monitoring points.

[0068] Dynamic Time Warping (DTW) algorithms can effectively compare the similarity between two sequences. Therefore, under each environment, for any two monitoring points, the DTW value of the corresponding spectral data at the last time node is calculated. The smaller the DTW value, the higher the similarity in the overall shape and trend of the spectral data of the two monitoring points. Therefore, the DTW value is negatively correlated and normalized to correct the logical relationship, obtaining the similarity factor of the spectral data of the two monitoring points under each environmental parameter. The larger the similarity factor, the higher the similarity of the spectral data of the two monitoring points. The negative correlation mapping and normalization here can be performed using the formula... ,in, Let x represent an exponential function with the natural constant e as the base, and let x represent the independent variable.

[0069] It should be noted that the method for obtaining DTW values ​​is a well-known technique, and the specific process will not be elaborated here.

[0070] Step S203: In each environment, for any two monitoring points, combine the spectral peak matching index between the two monitoring points with the change similarity factor to obtain the change similarity index between the two monitoring points in each environment.

[0071] Based on the analysis in step S201, it is known that under each environment, the peak matching index between two monitoring points is positively correlated with the spectral similarity between the two monitoring points, and the variation similarity factor is also positively correlated with the spectral similarity between the two monitoring points. Therefore, under each environment, the normalized value of the product of the peak matching index and the variation similarity factor between any two monitoring points is used as the variation similarity index between the spectral data of the two monitoring points under each environment. The larger the variation similarity index, the more consistent the spectral performance of the two monitoring points is considered to be under each environment, and thus the higher the probability that they will be classified into the same cluster in the subsequent process. Normalization is a technique well known to those skilled in the art. The choice of normalization function can be linear normalization or standard normalization, etc., and the specific normalization method is not limited here.

[0072] After obtaining the similarity index between every two monitoring points under each environment, cluster analysis can be performed on all monitoring points under each environment based on this index to obtain clusters.

[0073] Preferably, in one embodiment of the present invention, the method for obtaining clusters includes:

[0074] Under each environmental parameter, based on the DBSCAN clustering algorithm and the similarity index of changes between monitoring points under the corresponding spectral data, cluster analysis is performed on all monitoring points to obtain clusters of all monitoring points under each environmental parameter. The cluster radius and the minimum number of clusters are preset values.

[0075] It should be noted that in this embodiment of the present invention, the clustering radius is set to 0.3 and the minimum number of clusters is set to 3. The specific values ​​can be adjusted according to the implementation scenario and are not limited here. The DBSCAN clustering algorithm is a well-known technology, and the specific process will not be described in detail here.

[0076] This allows for clustering of all monitoring points under each environment, grouping those with similar spectral data into one category. Different clusters represent sub-regions within each primary sampling unit of the terraced fields where soil physicochemical properties differ. However, given the limited land area, precise fertilization of different regions during planting is impossible; instead, fertilization amounts are determined based on the overall properties of the sub-regions. Therefore, when measuring soil physicochemical properties, the temporal changes in nutrient composition across different sub-regions can be compared to identify the patterns of physicochemical property changes within the corresponding sub-region of each cluster. Consequently, within each cluster under each environment, the fluctuation characteristics of spectral peak differences at different time points are analyzed to determine the soil nutrient fluctuation index at each monitoring point.

[0077] Preferably, in one embodiment of the present invention, the method for obtaining the soil nutrient fluctuation index includes:

[0078] Soil nutrient status may change over time. Analyzing spectral changes at adjacent time points can capture this continuous dynamic process. Therefore, in each cluster under each environment, the spectral peaks of each monitoring point at each adjacent time point are compared in the spectral data corresponding to each environment. The absolute value of the difference between the peaks at each same wavelength is calculated as a deviation factor. The deviation factor reflects the change characteristics of a certain nutrient or property at the monitoring point under this environment at two adjacent time points. The larger the value, the more significant the change. The mean of the deviation factors at all the same wavelengths is used as the spectral difference factor of the monitoring point at each adjacent time point. The spectral difference factor characterizes the overall difference change characteristics of the same monitoring point in two adjacent spectral data. The larger the value, the more obvious the difference change characteristics.

[0079] Variance reflects the dispersion and fluctuation of a set of data. Therefore, the variance of the spectral difference factor at all two adjacent time points at a given monitoring point is used as the difference fluctuation parameter. The larger the difference fluctuation parameter, the more significant the fluctuation of the same nutrient components and characteristics in the spectral data at these two adjacent time points. Then, the value obtained by normalizing the difference fluctuation parameter with the sum of the mean of all spectral difference factors is used as the soil nutrient fluctuation factor for each monitoring point. Based on the above analysis, it can be seen that the larger the soil nutrient fluctuation factor, the more obvious the fluctuation of nutrient components in the soil at that monitoring point. Normalization is a technique well known to those skilled in the art. The choice of normalization function can be linear normalization or standard normalization, etc., and the specific normalization method is not limited here.

[0080] Finally, in each cluster of each environment, the mean value of the soil nutrient fluctuation factor corresponding to all monitoring points is used as the soil nutrient fluctuation index for each monitoring point in that cluster. The larger the soil nutrient fluctuation index, the more significant the change in soil nutrient composition at that monitoring point.

[0081] Thus, we can obtain the soil nutrient fluctuation index at each monitoring point under each environment (in situ environment and experimental environment).

[0082] Step S3: Fit a target straight line based on the soil nutrient fluctuation index, experimental spectrum, and environmental parameters of the monitoring points under the experimental environment; Based on the difference and numerical characteristics of the soil nutrient fluctuation index at the monitoring points in the cluster under the in-situ environment, the environmental parameters, in-situ spectrum, and target straight line of the in-situ environment, correct the in-situ spectrum of the monitoring points to obtain spectral correction data for determining soil nutrient elements.

[0083] Experimental environments typically allow for precise control and adjustment of various environmental parameters, resulting in data with high reliability and consistency. In such relatively stable environments, the relationship between soil nutrient fluctuation indices, spectral characteristics, and environmental parameters is more easily captured and quantified, thus the fitted target line more accurately reflects their inherent patterns. Therefore, a target line can be fitted based on soil nutrient fluctuation indices, experimental spectra, and environmental parameters at monitoring points within the experimental environment, providing an important reference benchmark for in-situ spectral correction.

[0084] Preferably, in one embodiment of the present invention, the method for obtaining the target straight line includes:

[0085] Soil surface moisture content and light intensity are important environmental factors affecting soil spectral characteristics and soil nutrient status. Soil moisture content affects the physical properties of the soil, while light intensity affects the lighting conditions during spectral measurements. Therefore, in the experimental environment, experimental environmental parameters are obtained at each monitoring point, including at least soil surface moisture content and light intensity.

[0086] Then, under experimental conditions, the mean intensity value, peak value, and half-maximum width were obtained from the experimental spectrum at the last time point of each monitoring point. These three indicators are commonly used characteristic parameters in spectral analysis, which can describe the morphology and characteristics of the spectrum from different perspectives and are closely related to the composition and properties of the soil.

[0087] The mean, peak, and half-peak width of the intensity values ​​obtained above are combined with the experimental environmental parameters and corresponding soil nutrient fluctuation indices at each monitoring point to construct an experimental fusion vector. Thus, the experimental fusion vector for each monitoring point can be obtained.

[0088] Finally, the fused vectors of all monitoring points in the experimental environment were used as the initial training set, and the WPLASR model was used to obtain the output of the fused vector for each monitoring point. Based on the output of the monitoring points and the fused vectors of the monitoring points, the least squares method was used to fit a straight line, thereby obtaining the target straight line. The target straight line can provide a concise and clear reference model for subsequent analysis and application, and can also serve as a benchmark for spectral correction in in-situ environments.

[0089] It should be noted that the WPLASR model and the least squares method are well-known techniques, and the specific process will not be elaborated here.

[0090] The target line was obtained by fitting soil nutrient fluctuation indices, experimental spectra, and experimental environmental parameters under experimental conditions. It reflects the intrinsic relationship between soil nutrient-related characteristics under relatively controllable experimental conditions. Although there are differences between the in-situ environment and the experimental environment, some basic physical and chemical processes related to soil nutrients have certain similarities in the two environments. Therefore, the in-situ spectra of the monitoring points can be corrected based on the differences and numerical characteristics of soil nutrient fluctuation indices at the monitoring points in the cluster under the in-situ environment, the environmental parameters of the in-situ environment, the in-situ spectra, and the target line to obtain spectral correction data.

[0091] Preferably, in one embodiment of the present invention, the method for obtaining spectral correction data includes:

[0092] In the in-situ environment, within each cluster, the difference between the maximum and minimum soil nutrient fluctuation indices is negatively correlated and normalized. This normalized value serves as a dynamic constraint parameter. The purpose of this dynamic constraint parameter is to balance the magnitude of subsequent spectral corrections. The larger the difference between the maximum and minimum soil nutrient fluctuation indices, the more significant the differences in soil nutrient status within that cluster. Therefore, the dynamic constraint will be smaller, and the subsequent adjustment magnitude will be larger. This negative correlation mapping can be achieved using the formula... ,in, Let x represent an exponential function with the natural constant e as the base, and let x represent the independent variable.

[0093] When the soil nutrient fluctuation index at the monitoring point is larger in the in-situ environment, a larger adjustment range is also required. Therefore, the value obtained by negatively correlated mapping of the product of the dynamic constraint parameter and the soil nutrient fluctuation index at each monitoring point in the in-situ environment is used as the adjustment coefficient. Based on the previous analysis, it can be seen that the larger the adjustment coefficient, the greater the degree of adjustment of the spectral data at that monitoring point. Given that the calculation process in step S2 shows that the value range of the soil nutrient fluctuation index is 0-1, and the value range of the dynamic constraint parameter is also 0-1, the negative correlation mapping here can be performed using the formula... ,in, It represents the independent variable.

[0094] Then, in situ environmental parameters were obtained at each monitoring point under in situ conditions. These in situ environmental parameters included at least the surface soil moisture content and light intensity.

[0095] Next, in the in-situ environment, the mean intensity, peak intensity, and full width at half maximum (FWHM) of the in-situ spectrum at the last time node of each monitoring point are obtained. These values ​​are then combined with the in-situ environmental parameters and corresponding soil nutrient fluctuation indices at each monitoring point to form an in-situ fusion vector. This fusion vector is substituted into the target line, and the output is multiplied by an adjustment coefficient. The normalized product is then used as the correction coefficient for each monitoring point. The output of the target line reflects the intrinsic relationship between soil nutrient-related characteristics and the environment. The adjustment coefficient considers both the clustering and the soil nutrient characteristics of the monitoring points. Therefore, the correction coefficient obtained by combining these two factors can more accurately and finely adjust the in-situ spectrum of the monitoring points, and the larger the value, the greater the degree of adjustment. Normalization is a technique well-known to those skilled in the art. The normalization function can be linear normalization or standard normalization, etc., and the specific normalization method is not limited here.

[0096] Finally, at each monitoring point, the sum of the correction coefficient and the preset parameter is used as the adjustment degree value. The adjustment degree value is multiplied by the intensity value at each wavelength in the in-situ spectrum (the last time node) to obtain the corrected intensity at each wavelength, thus obtaining the spectral correction data at each monitoring point.

[0097] It should be noted that, in order to prevent over-adjustment, the preset parameter in this embodiment of the present invention is set to 1.

[0098] After obtaining the spectral correction data at each monitoring point, the continuous projection algorithm (SPA) or competitive adaptive reweighted sampling (CARS) can be used to screen the subset of bands with the highest correlation to the target nutrient (such as OM and TN) from the entire band (simplifying calculation and improving model robustness). For example, organic matter can be screened at 760nm, 1240nm, 1720nm, and 2100nm; nitrogen can be screened at 550nm, 950nm, and 1550nm; and phosphorus can be screened at 1000nm, 1450nm, and 2300nm.

[0099] It should be noted that the Continuous Projection Algorithm (SPA) or Competitive Adaptive Reweighted Sampling (CARS) are well-known techniques, and their specific processes will not be elaborated here.

[0100] In this embodiment of the invention, all numerical values ​​involved in the calculation have undergone data preprocessing to eliminate the influence of dimensions. The specific means of eliminating the influence of dimensions are well known to those skilled in the art and will not be limited or described in detail here.

[0101] In summary, spectral data from various monitoring points in the terraced fields were obtained at multiple time points under different environments, encompassing both in-situ and experimental environments, as well as in-situ and experimental spectra. In-situ environmental data accurately reflects the spectral characteristics of the terraced soil under natural conditions, while experimental environmental data helps to eliminate some interfering factors. Due to the spatial heterogeneity of the physicochemical properties of the terraced soil, similarity indices were determined by analyzing the intensity differences and similarities of spectral peaks in the spectral data from different monitoring points under each environment. This quantifies the similarity between monitoring points, and cluster analysis is then performed on all monitoring points based on these similarity indices to obtain clusters. This allows monitoring points with similar change characteristics to be grouped together, better reflecting the spatial distribution patterns of the terraced soil. Since changes in soil nutrients are directly reflected in the spectral data, the fluctuations in soil nutrients over time were considered under each environment. The peak difference fluctuation characteristics of the spectral data from monitoring points in the clusters at different time points were analyzed to determine soil nutrient fluctuation indices. Furthermore, since the experimental environment is relatively controllable and can eliminate some interfering factors, a target straight line is fitted based on the soil nutrient fluctuation index, experimental spectrum, and environmental parameters at the monitoring points under the experimental environment. This target straight line can serve as a reference standard for in-situ spectral correction. Finally, based on the differences and numerical characteristics of the soil nutrient fluctuation index at the monitoring points in the cluster under the in-situ environment, the environmental parameters, in-situ spectrum, and target straight line, the in-situ spectrum of the monitoring points is corrected. This can more accurately eliminate the interference of the in-situ environment on the spectral data, obtain more accurate spectral correction data, and thus improve the accuracy of soil nutrient element measurement results.

[0102] This invention also provides a system for measuring the physicochemical properties of terraced soils. Please refer to [link / reference]. Figure 3 The diagram shows a system block diagram, including a data acquisition module 301 for implementing step S1 in the above method embodiment; a spectral analysis module 302 for implementing step S2 in the above method embodiment; and a spectral correction module 303 for implementing step S3 in the above method embodiment.

[0103] It should be noted that the system provided in the above embodiments is only an example of the division of the above functional modules. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the computer device can be divided into different functional modules to complete all or part of the functions described above. In addition, the terraced soil physicochemical property determination system and the terraced soil physicochemical property determination method embodiment provided in the above embodiments belong to the same concept, and their specific implementation process can be found in the method embodiment, which will not be repeated here.

[0104] This invention also provides an apparatus for measuring the physicochemical properties of terraced soils. Please refer to [link / reference]. Figure 4The diagram shows a schematic of the device structure, including a processor 400, a memory 401, a bus 402, and a communication interface 403. The processor 400, the communication interface 403, and the memory 401 are connected via the bus 402. The memory 401 may contain a high-speed random access memory, and the bus 402 may be an ISA bus, a PCI bus, or an EISA bus, etc. The processor 400 may be an integrated circuit chip with signal processing capabilities. The memory 401 stores at least one instruction, at least one program, a code set, or an instruction set. When the processor loads and executes the at least one instruction, at least one program, a code set, or an instruction set, it implements the steps in a method for determining the physicochemical properties of terraced soil.

[0105] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0106] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A method for determining the physicochemical properties of terraced field soil, characterized in that, The method includes: Spectral data of various monitoring points in the terraced fields at multiple time points under different environments were obtained. The environment included in-situ environment and experimental environment, and the types of spectral data included in-situ spectra and experimental spectra. In each environment, the intensity differences and similarities between spectral peaks in the spectral data of different monitoring points are analyzed to determine the change similarity index between monitoring points. In each environment, cluster analysis is performed on all monitoring points based on the change similarity index between monitoring points to obtain clusters. In each cluster in each environment, the fluctuation characteristics of the spectral peak differences of the monitoring points at different time points are analyzed to determine the soil nutrient fluctuation index at each monitoring point. The target line is fitted based on the soil nutrient fluctuation index, experimental spectrum and environmental parameters of the monitoring points under the experimental environment; the in-situ spectrum of the monitoring points is corrected based on the difference and numerical characteristics of the soil nutrient fluctuation index of the monitoring points in the cluster under the in-situ environment, the environmental parameters of the in-situ environment, the in-situ spectrum and the target line, and the spectral correction data is used to determine the soil nutrient elements. The methods for obtaining the similarity index include: In each environment, the intensity difference of spectral peaks in the spectral data of any two monitoring points at the last time node is analyzed to determine the spectral peak matching index between the two monitoring points. In each environment, the similarity between the spectral data of any two monitoring points at the last time node is analyzed to obtain the change similarity factor between the two monitoring points; In each environment, the normalized value of the product of the spectral peak matching index and the change similarity factor between any two monitoring points is used as the change similarity index between the spectral data of the two monitoring points in each environment. The method for obtaining the soil nutrient fluctuation index includes: In each cluster under each environment, the spectral peaks of the spectral data of each monitoring point under each environment are compared in the spectral data of each adjacent two time nodes. The absolute value of the difference between the peaks under each same wavelength is calculated as the deviation factor. The mean of all deviation factors under the same wavelength is used as the spectral difference factor of each monitoring point under each adjacent two time nodes. The variance of all spectral difference factors corresponding to each monitoring point is used as the difference fluctuation parameter. The value of the sum of the difference fluctuation parameter and the mean of the spectral difference factors is normalized and used as the soil nutrient fluctuation factor for each monitoring point. The mean value of the soil nutrient fluctuation factor corresponding to all monitoring points in each cluster is used as the soil nutrient fluctuation index at each monitoring point in each cluster.

2. The method for determining the physicochemical properties of terraced soil according to claim 1, characterized in that, The method for obtaining the spectral peak matching index includes: In each environment, for any two monitoring points, compare the spectral peaks in the corresponding spectral data of the two monitoring points at the last time node, match the spectral peaks with the same wavelength as spectral peak pairs, take the difference of the peak values ​​of each spectral peak pair as the peak difference factor of each spectral peak pair, and take the sum of the peak difference factors of all spectral peak pairs as the peak difference parameter of the two monitoring points. The number of peak pairs at the two monitoring points is used as the numerator, the sum of the peak difference parameter between the two monitoring points and the preset constant is used as the denominator, and the normalized value of the resulting ratio is used as the peak matching index between the two monitoring points.

3. The method for determining the physicochemical properties of terraced soil according to claim 1, characterized in that, The method for obtaining the change similarity factor includes: In each environment, for any two monitoring points, the DTW values ​​of the corresponding spectral data of these two monitoring points at the last time node are negatively correlated and normalized to obtain the similarity factor of the change of spectral data of the two monitoring points under each environmental parameter.

4. The method for determining the physicochemical properties of terraced soil according to claim 1, characterized in that, The method for obtaining the clusters includes: Under each environmental parameter, based on the DBSCAN clustering algorithm and the similarity index of changes between monitoring points under the corresponding spectral data, cluster analysis is performed on all monitoring points to obtain clusters of all monitoring points under each environmental parameter. The cluster radius and the minimum number of clusters are preset values.

5. The method for determining the physicochemical properties of terraced soil according to claim 1, characterized in that, The method for obtaining the target straight line includes: In the experimental environment, the experimental environmental parameters at each monitoring point are obtained. The experimental environmental parameters include at least the soil surface moisture content and light intensity. In the experimental environment, the mean intensity value, peak value, and half-peak width of the experimental spectrum at the last time node of each monitoring point are obtained, and the experimental fusion vector is formed with the experimental environmental parameters and corresponding soil nutrient fluctuation index at each monitoring point. The fused vectors of all monitoring points in the experimental environment are used as the initial training set, and the outputs of the monitoring points are obtained using the WPLASR model. Based on the outputs of the monitoring points and the fused vectors of the monitoring points, the least squares method is used to fit a straight line to obtain the target straight line.

6. The method for determining the physicochemical properties of terraced soil according to claim 1, characterized in that, The method for obtaining the spectral correction data includes: In the in-situ environment, within each cluster, the difference between the maximum and minimum soil nutrient fluctuation indices is negatively correlated and normalized, and this normalized value is used as a dynamic constraint parameter. The value obtained by negatively mapping the product of the dynamic constraint parameters and the soil nutrient fluctuation index at each monitoring point in the in-situ environment is used as the adjustment coefficient; In the in-situ environment, the in-situ environmental parameters at each monitoring point are obtained, and the in-situ environmental parameters include at least the soil surface moisture content and light intensity. In the in-situ environment, the mean intensity value, peak value, and half-peak width are obtained from the in-situ spectrum at the last time node of each monitoring point, and are combined with the in-situ environmental parameters and corresponding soil nutrient fluctuation indicators at each monitoring point to form an in-situ fusion vector. Substitute the in-situ fusion vector into the target line, multiply the output by the adjustment coefficient, and use the resulting product as the correction coefficient for each monitoring point. At each monitoring point, the sum of the correction coefficient and the preset parameter is used as the adjustment degree value. The adjustment degree value is multiplied by the intensity value at each wavelength in the in-situ spectrum to obtain the corrected intensity at each wavelength, thereby obtaining the spectral correction data at each monitoring point.

7. A system for determining the physicochemical properties of terraced soil, characterized in that, The system includes: The data acquisition module is used to acquire spectral data of various monitoring points in the terraced fields at multiple time points under different environments. The environment includes in-situ environment and experimental environment, and the types of spectral data include in-situ spectra and experimental spectra. The spectral analysis module is used to analyze the intensity differences and similarities between spectral peaks in the spectral data of different monitoring points under each environment, and to determine the change similarity index between monitoring points. Under each environment, cluster analysis is performed on all monitoring points based on the change similarity index between monitoring points to obtain clusters. In each cluster under each environment, the fluctuation characteristics of the spectral peak differences of the monitoring points at different time points are analyzed to determine the soil nutrient fluctuation index at each monitoring point. The spectral correction module is used to fit a target straight line based on the soil nutrient fluctuation index, experimental spectrum and environmental parameters of the monitoring points under experimental conditions; based on the difference and numerical characteristics of the soil nutrient fluctuation index of the monitoring points in the cluster under in-situ conditions, the environmental parameters of the in-situ environment, the in-situ spectrum and the target straight line, the in-situ spectrum of the monitoring points is corrected to obtain spectral correction data for the determination of soil nutrient elements. The methods for obtaining the similarity index include: In each environment, the intensity difference of spectral peaks in the spectral data of any two monitoring points at the last time node is analyzed to determine the spectral peak matching index between the two monitoring points. In each environment, the similarity between the spectral data of any two monitoring points at the last time node is analyzed to obtain the change similarity factor between the two monitoring points; In each environment, the normalized value of the product of the spectral peak matching index and the change similarity factor between any two monitoring points is used as the change similarity index between the spectral data of the two monitoring points in each environment. The method for obtaining the soil nutrient fluctuation index includes: In each cluster under each environment, the spectral peaks of the spectral data of each monitoring point under each environment are compared in the spectral data of each adjacent two time nodes. The absolute value of the difference between the peaks under each same wavelength is calculated as the deviation factor. The mean of all deviation factors under the same wavelength is used as the spectral difference factor of each monitoring point under each adjacent two time nodes. The variance of all spectral difference factors corresponding to each monitoring point is used as the difference fluctuation parameter. The value of the sum of the difference fluctuation parameter and the mean of the spectral difference factors is normalized and used as the soil nutrient fluctuation factor for each monitoring point. The mean value of the soil nutrient fluctuation factor corresponding to all monitoring points in each cluster is used as the soil nutrient fluctuation index at each monitoring point in each cluster.

8. A device for measuring the physicochemical properties of terraced field soil, characterized in that, It includes a processor and a memory, the memory storing at least one instruction, at least one program, code set or instruction set, the steps of the method for determining the physical and chemical properties of terraced soil as described in any one of claims 1-6 being implemented when at least one instruction, at least one program, code set or instruction set is loaded and executed by the processor.

Images