A soil health grading system and method

Abstract

The application discloses a kind of soil health grading evaluation system and method, it is related to soil health data processing technical field.The system includes soil data acquisition module, soil health data detection optimization module, remote sensing-sampling data fusion module and health grade remote sensing verification optimization module.The application obtains soil health data detection evaluation value by obtaining soil data and judges whether soil health data detection optimization is carried out, if yes, then carry out data fusion link after optimization, if no, directly carry out data fusion link, obtain soil health grade remote sensing verification deviation degree and judge whether health grade remote sensing verification optimization is carried out, if yes, then carry out soil health grade division after optimization, if no, directly carry out soil health grade division, improve the authenticity of soil health grading evaluation, solve the problem that existing technology exists due to the accumulation of soil data error leads to soil health grading evaluation distortion.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of soil health data processing, in particular to a soil health grading evaluation system and method. BACKGROUND

[0002] In the soil sample collection and pretreatment stage, the standard soil drill is used to obtain the cultivated layer soil sample, which is mixed and reduced by quartile method, and then brought back to the laboratory for natural air drying, grinding, sieving and other pretreatments. In the soil health index determination stage, the soil moisture sensor is used to monitor the soil moisture content in real time, the laser particle size analyzer is used to determine the soil particle composition (such as the proportion of sand, silt and clay), the container density instrument is used to determine the soil bulk density to evaluate the soil compactness; the pH meter is used to quickly detect the soil acidity and alkalinity, the atomic absorption spectrophotometer or the inductively coupled plasma mass spectrometer is used to accurately determine the contents of available nutrients such as nitrogen, phosphorus and potassium in the soil and heavy metals such as cadmium and lead, and the organic matter analyzer is used to determine the soil organic matter content; the soil respiration instrument is used to measure the soil microbial respiration intensity, the high-throughput sequencing technology is used to analyze the soil microbial community structure and diversity, and the enzyme label instrument is used to determine the soil enzyme activity such as urease and phosphatase, so as to comprehensively obtain the basic data of soil health. Then, in the data standardization and weight determination stage, the original data is processed by using the membership function method or the standardization normalization method according to the dimension difference of different indexes to eliminate the influence of the order of magnitude between indexes; then, the weight of each index is determined by combining the principal component analysis method, the analytic hierarchy process or the entropy weight method, and the quantitative and qualitative analysis is comprehensively determined.

[0003] According to the standardized values and corresponding weights of each index, the weighted summation method is used to calculate the soil health comprehensive index, and the geographic information technology is introduced to associate the comprehensive index of each sampling point with the spatial position, generate the soil health spatial distribution map, divide the comprehensive index into six levels of excellent, good, good, medium, poor and poor, and combine the unmanned aerial vehicle remote sensing technology to verify the grades of large area.

[0004] For example, the Chinese invention patent with publication number CN118837524A discloses a farmland healthy soil evaluation method and system, which comprises the following steps: S1: constructing a healthy soil evaluation index system; S2: obtaining soil samples; S3: based on the evaluation index system, using a first evaluation method to obtain the subjective weight value of each evaluation index; S4: simultaneously, analyzing and testing the soil samples and investigating and analyzing the indexes to obtain the objective data of the soil samples; S5: using a second evaluation method to process the objective data to obtain the objective weight value of each evaluation index; S6: using a combination algorithm to process the obtained subjective weight value and objective weight value to obtain a combined weight value; and S7: processing the combined weight value according to a preset scoring rule to obtain the health grade value of the soil sample.

[0005] For example, the disclosure number: CN119985913A Chinese invention patent discloses a kind of farmland soil health evaluation method and system, comprising: by obtaining the remote sensing image data in the target area, the color feature of remote sensing image is decomposed using HIS model, obtains the hue value, saturation value, brightness value of unit grid, constructs color feature vector set, the elements of unit grid color feature vector set are clustered into K kind of unit grid type, with unit grid type corresponding soil type, the soil in the target area is divided into K categories, merge the same type and adjacent unit grid region as independent region, determine the number of sampling points in any independent region, soil sampling is carried out at sampling point, determine the concentration value of each pollutant in the soil sample of independent region, obtain the distribution of each pollutant in the target area, calculate the pollutant single pollution index of independent region according to the obtained dye concentration value, the pollution degree grade of independent region is divided.

[0006] The above-mentioned technology at least has the following technical problems:

[0007] In the process of soil health grading evaluation, the commonly used spatial interpolation method (such as Kriging interpolation, inverse distance weighting method) depends on the spatial distribution density of sampling points and the spatial continuity of soil properties. If the sampling points are unevenly distributed, the Kriging interpolation will deviate from the true value due to the fitting error of the semi-variation function in the sparse data area. If there is a "sudden boundary" in the soil properties (such as sandy soil on one side of a plot and clay soil on the other side, with clear boundary), the inverse distance weighting method will over-average the comprehensive index of the boundary area due to the "smoothing effect", which will cover up the real difference in soil health grade, and there is a problem of mismatch between the interpolation method and the spatial heterogeneity of soil.

[0008] The time of obtaining remote sensing data is different from the time of ground sampling, and the soil health condition may change due to factors such as fertilization and precipitation during this period, which leads to the mismatch between the indirect indicators obtained by remote sensing and the direct indicators measured on the ground, and the grade cannot be effectively verified. Different soil problems may lead to the same spectral response. If only a single spectral index is used for verification, it will not be able to distinguish between "poor vegetation caused by poor soil health" and "poor vegetation caused by climate factors", and the misjudgment rate of the verification result is high. In addition, if the spatial resolution of unmanned aerial remote sensing is lower than the spatial representativeness of the sampling points, one pixel may correspond to multiple sampling points, which will cause averaging deviation when matching data, and there is a problem of distortion of soil health grading evaluation caused by accumulation of soil data errors. SUMMARY

[0009] In order to solve the problem of soil health grading evaluation distortion caused by accumulation of soil data errors in the prior art, the embodiments of the present application provide a soil health grading evaluation system and method. The technical scheme is as follows:

[0010] In one aspect, a soil health grading evaluation system is provided, comprising: a soil data acquisition module, a soil health data detection optimization module, a remote sensing-sampling data fusion module, and a health grade remote sensing verification optimization module; wherein the soil data acquisition module is configured to acquire soil data required for soil health evaluation in a soil sample area and perform soil health detection to obtain a soil health data detection evaluation value for quantifying the accuracy of soil health data in the soil health detection process; the soil health data detection optimization module is configured to determine whether to perform soil health data detection optimization according to the soil health data detection evaluation value, if yes, then perform a remote sensing-sampling data fusion link after soil health data detection optimization, if no, directly perform the remote sensing-sampling data fusion link, and the soil health data detection optimization includes dynamic adjustment of sampling-detection interval time and dynamic adjustment of minimum sampling point spacing; the remote sensing-sampling data fusion module is configured to acquire a remote sensing verification deviation parameter in the remote sensing-sampling data fusion link to obtain a soil health grade remote sensing verification deviation degree for quantifying the deviation degree of the verification conclusion from the true situation when the remote sensing data verifies the soil health grading result; and the health grade remote sensing verification optimization module is configured to determine whether to perform health grade remote sensing verification optimization according to the soil health grade remote sensing verification deviation degree, if yes, then perform soil health grade division after health grade remote sensing verification optimization, if no, directly perform soil health grade division, and the health grade remote sensing verification optimization includes dynamic adjustment of data fusion time window and dynamic adjustment of remote sensing image spatial resolution.

[0011] In another aspect, a soil health grading evaluation method is provided, comprising: acquiring soil data required for soil health evaluation in a soil sample area and performing soil health detection to obtain a soil health data detection evaluation value for quantifying the accuracy of soil health data in the soil health detection process; determining whether to perform soil health data detection optimization according to the soil health data detection evaluation value, if yes, then performing a remote sensing-sampling data fusion link after soil health data detection optimization, if no, directly performing the remote sensing-sampling data fusion link, and the soil health data detection optimization includes dynamic adjustment of sampling-detection interval time and dynamic adjustment of minimum sampling point spacing; acquiring a remote sensing verification deviation parameter in the remote sensing-sampling data fusion link to obtain a soil health grade remote sensing verification deviation degree for quantifying the deviation degree of the verification conclusion from the true situation when the remote sensing data verifies the soil health grading result; and determining whether to perform health grade remote sensing verification optimization according to the soil health grade remote sensing verification deviation degree, if yes, then performing soil health grade division after health grade remote sensing verification optimization, if no, directly performing soil health grade division, and the health grade remote sensing verification optimization includes dynamic adjustment of data fusion time window and dynamic adjustment of remote sensing image spatial resolution.

[0012] The one or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages:

[0013] 1、By obtaining the soil data required for soil health evaluation in the soil sample area and conducting soil health detection, the soil health data detection evaluation value is obtained, which is used to quantify the accuracy of soil health data in the soil health detection process, and whether to optimize the soil health data detection is judged according to the soil health data detection evaluation value to ensure that the precision of the input soil data meets the requirements, reduce the evaluation distortion caused by data error accumulation, and avoid meaningless repeated detection, balance the data quality and detection efficiency; The soil health grade remote sensing verification deviation degree is obtained by obtaining the remote sensing verification deviation parameter in the remote sensing-sampling data fusion link, which is used to quantify the deviation degree of the verification conclusion from the true situation when the remote sensing data verifies the soil health grading result, and whether to optimize the health grade remote sensing verification is judged according to the soil health grade remote sensing verification deviation degree, which can effectively avoid the misjudgment of soil health grade caused by verification deviation, ensure the accuracy and reliability of the soil health grading evaluation result, avoid resource waste when optimization is not needed, improve the evaluation efficiency, and further improve the authenticity of soil health grading evaluation.

[0014] 2、According to the soil health data detection evaluation value, whether to optimize the soil health data detection is judged, which can accurately identify the error problem of soil health data in the collection and determination link, reduce the evaluation distortion caused by data error accumulation, balance the data quality and detection efficiency, provide high-reliability data support for the soil health grading evaluation system, and judge whether to dynamically adjust the sampling-detection interval time according to the soil property change rate, which can shorten the interval to capture real-time changes and avoid interpolation error caused by data lag when the soil property changes fast, and lengthen the interval to reduce redundant sampling and cost waste when the soil property changes slowly, which can ensure that the sampling data accurately match the soil spatial heterogeneity characteristics, balance the sampling efficiency and cost, and provide high-quality data support for the spatial interpolation of soil health grading evaluation, effectively reduce the evaluation distortion risk caused by insufficient data timeliness or excessive sampling; According to the soil sampling point boundary coverage rate, whether to dynamically adjust the minimum distance of the sampling point is judged, which can ensure that the spatial distribution of the sampling point matches the soil spatial heterogeneity characteristics, balance the sampling precision and resource efficiency, provide accurate data support for the spatial interpolation of soil health grading evaluation, effectively reduce the evaluation distortion risk caused by insufficient boundary coverage, and further improve the authenticity of soil health grading evaluation.

[0015] 3. According to the soil health grade remote sensing verification deviation degree, it is judged whether to perform health grade remote sensing verification optimization, the deviation problem in the soil health grading result verification by remote sensing data can be accurately identified, the misjudgment of the soil health grade caused by the verification deviation is reduced, the verification accuracy and the resource cost are balanced, the reliability of the soil health grading evaluation result is provided, according to the remote sensing image acquisition frequency, it is judged whether to perform data fusion time window dynamic adjustment, the window is reduced to reduce the mismatch between remote sensing and ground indexes caused by short-term interference such as fertilization and precipitation when the acquisition frequency is high, the window is reasonably expanded to ensure that there is suitable image available and the data without extreme interference is screened when the acquisition frequency is low, the index deviation and verification failure caused by the improper time window are avoided, and the data synchronization and the remote sensing resource utilization efficiency are balanced, and time dimension support is provided for accurate fusion of remote sensing-soil sampling data; according to the ground sampling depth, it is judged whether to perform remote sensing image spatial resolution dynamic adjustment, the surface interference information is reduced and the expanded variation scale is matched, the remote sensing data and the ground sampling data are accurately matched in the spatial scale, the index mismatch and error accumulation caused by improper resolution are avoided, and then the soil health grading evaluation authenticity is improved. BRIEF DESCRIPTION OF DRAWINGS

[0016] Figure 1 A structure schematic diagram of a soil health grading evaluation system provided by an embodiment of the present application is shown in the figure.

[0017] Figure 2 A sampling-detection interval dynamic adjustment flowchart of a soil health grading evaluation system provided by an embodiment of the present application is shown in the figure.

[0018] Figure 3 A sampling point minimum interval dynamic adjustment flowchart of a soil health grading evaluation system provided by an embodiment of the present application is shown in the figure.

[0019] Figure 4 A flowchart of a soil health grading evaluation method provided by an embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0020] The technical solutions in the present application are described below with reference to the drawings.

[0021] In the embodiments of the present application, the words such as "example", "for example" are used to represent an example, illustration or description. Any embodiment or design scheme described as "example" in the present application should not be interpreted as more preferred or more advantageous than other embodiments or design schemes. In fact, the word "example" is intended to present the concept in a specific way. In addition, in the embodiments of the present application, the meaning expressed by "and / or" can be both, or can be one of the two.

[0022] The embodiment of the present application provides a soil health grading evaluation system and method, solves the problem of soil health grading evaluation distortion caused by accumulation of soil data errors in the prior art, obtains soil health data detection evaluation values by acquiring soil data, judges whether to perform soil health data detection optimization, if yes, performs a data fusion link after optimization, if not, directly performs the data fusion link, obtains soil health grade remote sensing verification deviation degrees by acquiring remote sensing verification deviation parameters, judges whether to perform health grade remote sensing verification optimization, if yes, performs soil health grade division after optimization, if not, directly performs soil health grade division, and the soil health grading evaluation authenticity is improved.

[0023] The technical solution in the embodiment of the present application is to solve the problem of soil health grading evaluation distortion caused by accumulation of soil data errors, and the general idea is as follows:

[0024] The soil health data detection evaluation values are obtained by acquiring soil data required for soil health evaluation in a soil sample area and performing soil health detection, whether to perform soil health data detection optimization is judged according to the soil health data detection evaluation values, if yes, a remote sensing-sampling data fusion link is performed after soil health data detection optimization, if not, the remote sensing-sampling data fusion link is directly performed, soil health grade remote sensing verification deviation degrees are obtained by acquiring remote sensing verification deviation parameters in the remote sensing-sampling data fusion link, whether to perform health grade remote sensing verification optimization is judged according to the soil health grade remote sensing verification deviation degrees, if yes, soil health grade division is performed after health grade remote sensing verification optimization, if not, soil health grade division is directly performed, and the soil health grading evaluation authenticity is improved.

[0025] In order to better understand the above technical solution, the above technical solution will be described in detail in combination with the drawings of the specification and specific embodiments.

[0026] As shown in Figure 1 Fig. 1 is a structural schematic diagram of a soil health grading evaluation system provided by the embodiment of the present application, and the soil health grading evaluation system provided by the embodiment of the present application comprises a soil data acquisition module, a soil health data detection optimization module, a remote sensing-sampling data fusion module and a health grade remote sensing verification optimization module.

[0027] As the first module of a soil health grading evaluation system, the soil data acquisition module is used to acquire soil data required for soil health evaluation in a soil sample area and perform soil health detection, to obtain soil health data detection evaluation values, and to quantify the accuracy of soil health data in the soil health detection process.

[0028] It should be noted that the soil data includes soil sampling point density, soil sampling point average spacing and soil property change rate; wherein the soil sampling point density refers to the ratio of the total number of sampling points in the soil sample area to the total area of the region; the soil sampling point average spacing refers to the arithmetic mean of the straight line distance between adjacent sampling points in the soil sample area; the soil property change rate refers to the change amplitude of the soil property (such as bulk density, organic matter content, pH value, etc.) in unit time.

[0029] The density influence value is obtained by multiplying the soil sampling point density and the relative deviation proportion result of the density threshold value through the density influence factor. Wherein, the relative deviation proportion result refers to the ratio of the absolute value of the difference between the soil sampling point density and the density threshold value to the density threshold value.

[0030] The spacing influence value is obtained by multiplying the soil sampling point average spacing and the relative deviation proportion result of the spacing threshold value through the spacing influence factor. Wherein, the relative deviation proportion result refers to the ratio of the absolute value of the difference between the soil sampling point average spacing and the spacing threshold value to the spacing threshold value.

[0031] The change rate influence value is obtained by multiplying the soil property change rate and the relative deviation proportion result of the change rate threshold value through the change rate influence factor. Wherein, the relative deviation proportion result refers to the ratio of the absolute value of the difference between the soil property change rate and the change rate threshold value to the change rate threshold value.

[0032] The soil health data detection evaluation value is obtained by coupling the density influence value, the spacing influence value and the change rate influence value. Wherein, the coupling processing refers to the addition operation, and the multiplication processing refers to the multiplication operation.

[0033] It is necessary to supplement that the density influence factor, the density threshold value, the interval influence factor, the interval threshold value, the change rate influence factor and the change rate threshold value are obtained from the health grading evaluation database. The soil data has correlation as follows: there is a close correlation among the soil sampling point density, the average interval of soil sampling points and the change rate of soil properties, which jointly determines the accuracy of soil investigation and monitoring. The soil sampling point density refers to the number of sampling points per unit area. The higher the density, the smaller the average interval of sampling points, so that the spatial variation characteristics of soil properties can be captured more finely. On the contrary, low density and large interval may miss the local area with significant changes, resulting in information distortion. The change rate of soil properties reflects the change intensity of soil characteristics (such as organic matter content, pH value, nutrient concentration, etc.) in space or time. The faster the change rate, the greater the difference of soil properties in a short distance, and higher sampling point density and smaller average interval are needed to ensure data representativeness. Therefore, the dynamic coordination among the three is the key to ensuring the reliability of soil investigation data and the accuracy of spatial model. The soil data and the soil health data detection evaluation value have correlation as follows: the soil sampling point density, the average interval of soil sampling points and the change rate of soil properties have positive correlation with the soil health data detection evaluation value. The higher the relative deviation of soil sampling point density and density threshold value, the weaker the ability to capture soil spatial heterogeneity, and the higher the soil health data detection evaluation value. The higher the relative deviation of the average interval of soil sampling points and the interval threshold value, the higher the soil health data detection evaluation value, which means that the soil health status cannot be fully reflected. The higher the relative deviation of the change rate of soil properties and the change rate threshold value, the higher the soil health data detection evaluation value, which indicates that the soil health status fluctuates dramatically in a short distance.

[0034] As a second module of the soil health grading evaluation system, the soil health data detection optimization module is used to determine whether to perform soil health data detection optimization according to the soil health data detection evaluation value. If yes, the remote sensing-sampling data fusion link is performed after soil health data detection optimization. If no, the remote sensing-sampling data fusion link is directly performed. Soil health data detection optimization includes dynamic adjustment of sampling-detection interval and dynamic adjustment of minimum sampling point interval.

[0035] Further, the specific process of determining whether to perform soil health data detection optimization is as follows: if the soil health data detection evaluation value is less than or equal to the data detection setting value, soil health data detection optimization is not performed; if the soil health data detection evaluation value is greater than the data detection setting value, it is determined whether to perform dynamic adjustment of sampling-detection interval according to the change rate of soil properties. If yes, it is determined whether to perform dynamic adjustment of minimum sampling point interval after dynamic adjustment of sampling-detection interval. If no, it is directly determined whether to perform dynamic adjustment of minimum sampling point interval.

[0036] It should be noted that adjusting the sampling-detection interval time based on the change rate of soil properties is essentially to dynamically match the sampling density with the spatial heterogeneity characteristics of the soil, to reduce the mismatch between the interpolation method and the actual spatial distribution of the soil from the source, to provide more accurate data support for subsequent spatial interpolation, to reduce the cost of sampling and detection, and to realize the reasonable allocation of resources. In summary, adjusting the sampling-detection interval time based on the change rate of soil properties can dynamically optimize the spatial distribution density of sampling points, so that the sampling data can more accurately match the spatial heterogeneity characteristics of soil properties, and fundamentally alleviate the error problems caused by insufficient data support or mismatch between data and heterogeneity in Kriging interpolation and inverse distance weighting method, and improve the accuracy and reliability of the spatial interpolation results in soil health classification evaluation.

[0037] As Figure 2As shown, a sampling-detection interval time dynamic adjustment flowchart of a soil health grading evaluation system provided by the embodiment of the application is provided, and the specific logic is as follows: if the soil property change rate is less than the lower limit of the change rate setting, the result of the harmonic average of the change rate correction amount and the health data detection correction amount is input into the detection interval time mapping table for index query to obtain an interval time adjustment factor, it is judged whether the interval time adjustment factor is greater than the adjustment factor setting value, if yes, the adjustment factor correction amount is input into the detection interval time mapping table for index query to obtain an interval time reduction amount, the current sampling-detection interval time is reduced by the interval time reduction amount to obtain a preliminary adjusted interval time, and the result of the interaction of the preliminary adjusted interval time and the adjustment factor correction amount is taken as the next sampling-detection interval time, if no, the adjustment factor reference amount is input into the detection interval time mapping table for index query to obtain an interval time increase amount, the current sampling-detection interval time is added by the interval time increase amount to obtain a preliminary adjusted interval time, and the result of the interaction of the preliminary adjusted interval time and the adjustment factor reference amount is taken as the next sampling-detection interval time, if the soil property change rate is within the change rate setting interval, the sampling-detection interval time dynamic adjustment is not performed, if the soil property change rate is greater than the upper limit of the change rate setting, the result of the harmonic average of the change rate correction amount and the health data detection correction amount is input into the detection interval time mapping table for index query to obtain an interval time regulation factor, it is judged whether the interval time regulation factor is greater than the regulation factor setting value, if yes, the regulation factor correction amount is input into the detection interval time mapping table for index query to obtain an interval time up-regulation amount, the current sampling-detection interval time is coupled with the interval time up-regulation amount to obtain a preliminary adjusted interval time, and the result of the interaction of the preliminary adjusted interval time and the regulation factor correction amount is taken as the next sampling-detection interval time, if no, the regulation factor reference amount is input into the detection interval time mapping table for index query to obtain an interval time down-regulation amount, the current sampling-detection interval time is reduced by the interval time down-regulation amount to obtain a preliminary adjusted interval time, and the result of the interaction of the preliminary adjusted interval time and the regulation factor reference amount is taken as the next sampling-detection interval time.

[0038] As a further specific description, the specific judgment flow of whether to perform sampling-detection interval time dynamic adjustment according to the soil property change rate is as follows:

[0039] If the soil property change rate is less than the change rate lower limit, the change rate correction amount and the health data detection correction amount are harmonically averaged to obtain an interval time adjustment factor by index query in the detection interval time mapping table. When the soil property changes slowly, the sampling-detection interval time can be dynamically adjusted according to the soil health data detection accuracy difference, which can avoid wasting resources due to slow changes and ensure the data timeliness and accuracy balance of the input soil health classification evaluation system by correcting the detection accuracy, thereby reducing the evaluation distortion risk. The change rate correction amount represents the difference between the change rate lower limit and the soil property change rate, and the health data detection correction amount represents the difference between the soil health data detection evaluation value and the data detection set value.

[0040] If the interval time adjustment factor is greater than the adjustment factor set value, an adjustment factor correction amount is input into the detection interval time mapping table for index query to obtain an interval time reduction amount. The current sampling-detection interval time is reduced by the interval time reduction amount to obtain a preliminary adjusted interval time. The result of the interaction between the preliminary adjusted interval time and the adjustment factor correction amount is used as the next sampling-detection interval time, which can avoid the interval being excessively shortened or insufficient due to single reduction, accurately match the soil property change and data detection demand, and ensure that the sampling-detection interval can capture the necessary soil health dynamics without wasting resources. The adjustment factor correction amount represents the difference between the interval time adjustment factor and the adjustment factor set value. The reduction processing refers to subtraction operation, and the interaction processing refers to multiplication operation.

[0041] If not, an adjustment factor reference amount is input into the detection interval time mapping table for index query to obtain an interval time increase amount. The current sampling-detection interval time is added by the interval time increase amount to obtain a preliminary adjusted interval time. The result of the interaction between the preliminary adjusted interval time and the adjustment factor reference amount is used as the next sampling-detection interval time, which can adapt to the slow change of soil properties to reduce redundant sampling and reduce costs, and ensure that the soil health data detection accuracy and timeliness demand can still be met after the interval is extended through interaction processing. This can provide accurate and economic data support for soil health classification evaluation, effectively reduce the risk of data redundancy or information loss caused by improper interval, and the adjustment factor reference amount represents the difference between the adjustment factor set value and the interval time adjustment factor. The addition processing refers to addition operation.

[0042] In the embodiment, the redundant sampling is reduced by extending the interval, the detection cost is reduced, the precision standard and the basic timeliness requirement of the soil health data detection are ensured after the interval is extended through the interactive processing, the key information loss caused by the too long interval is avoided, the accurate and economic data guarantee is provided for the soil health grading evaluation, the data redundancy or information loss risk caused by the improper interval is effectively reduced, and the reliability and practicality of the evaluation result are improved.

[0043] As a further specific description, the judgment of whether to perform the sampling-detection interval time dynamic adjustment further includes:

[0044] If the soil property change rate is within the change rate setting interval, the sampling-detection interval time dynamic adjustment is not performed, and the current sampling-detection interval time is maintained, so as to avoid unnecessary interval adjustment to cause resource consumption and data stability fluctuation, and to ensure that the interval time is adapted to the soil property change rhythm, and the change rate setting interval represents a closed interval formed by the change rate setting lower limit and the change rate setting upper limit.

[0045] If the soil property change rate is greater than the change rate setting upper limit, the result of the harmonic average of the change rate correction amount and the health data detection correction amount is input into the detection interval time mapping table for index query to obtain the interval time control factor, and the change rate correction amount represents the difference between the soil property change rate and the change rate setting upper limit.

[0046] The judgment of whether the interval time control factor is greater than the control factor setting value is performed, if yes, the control factor correction amount is input into the detection interval time mapping table for index query to obtain the interval time up-regulation amount, the current sampling-detection interval time is coupled with the interval time up-regulation amount to obtain the preliminary adjusted interval time, and the result of the interactive processing of the preliminary adjusted interval time and the control factor correction amount is used as the next sampling-detection interval time, so as to avoid the data lag caused by the failure to timely adjust the interval due to the fast change, and to ensure that the interval adjustment does not deviate from the data quality requirement through the detection precision correction, so as to input the data with timeliness and accuracy into the soil health grading evaluation system, reduce the evaluation distortion risk caused by the data lag or insufficient accuracy, and the control factor correction amount represents the difference between the interval time control factor and the control factor setting value. The coupling processing refers to the addition operation.

[0047] If not, the control factor reference amount is input into the detection interval time mapping table for index query to obtain an interval time reduction amount, the current sampling-detection interval time is reduced by the interval time reduction amount to obtain a preliminary adjusted interval time, and the result of the interaction between the preliminary adjusted interval time and the control factor reference amount is taken as the next sampling-detection interval time, which not only adapts to the potential change trend of the soil property through the reduction processing to avoid data lag, but also avoids resource waste caused by excessive interval reduction through the interaction processing, so as to ensure that the sampling-detection interval can meet the timeliness requirement of soil health data, and the control factor reference amount represents the difference between the control factor set value and the interval control factor.

[0048] In the embodiment, the interval is ensured to meet the data timeliness requirement and balance the detection cost, and finally the problems of data lag, insufficient precision or resource waste caused by improper interval are comprehensively reduced, providing precise, efficient and reliable data support for soil health grading evaluation, and reducing the risk of evaluation distortion.

[0049] It should be noted that in the practice of spatial interpolation of soil health grading evaluation, the sampling point boundary coverage (i.e. the coverage of the sampling point to the key spatial boundaries such as the "mutation boundary" of soil property and the boundary of data sparse area) directly determines whether the spatial distribution density of sampling points can match the spatial heterogeneity characteristics of soil, and adjusting the minimum interval of sampling points based on the coverage is essentially to optimize the distribution density of sampling points in the key boundary area, to solve the problem of mismatch between interpolation method and soil spatial heterogeneity from the source, which can maintain the interpolation accuracy and balance the sampling efficiency and cost. In summary, adjusting the minimum interval of sampling points based on the boundary coverage of soil sampling points can dynamically optimize the distribution density of sampling points in the key boundary area, so that the spatial distribution of sampling points can more accurately match the heterogeneity characteristics such as the "mutation boundary" of soil property and the boundary of data sparse area, fundamentally alleviate the fitting error of Kriging interpolation and the "smoothing effect" of inverse distance weighting method, solve the problem of mismatch between interpolation method and soil spatial heterogeneity, and provide more accurate sampling data support for soil health grading evaluation, and finally improve the accuracy and reliability of spatial interpolation results.

[0050] For example, the soil health grading evaluation system can be used to determine the soil health grade of a certain area according to the soil health data of the area. Figure 3As shown, a soil health grading evaluation system sampling point minimum spacing dynamic adjustment flowchart provided by the embodiment of the application, the specific logic is: judging whether the soil sampling point boundary coverage is greater than or equal to the boundary coverage set value, if yes, no sampling point minimum spacing dynamic adjustment is performed, if not, the boundary coverage correction amount and the health data detection correction amount are arithmetically averaged, and the result is input into the minimum spacing mapping table for index query to obtain the minimum spacing adjustment coefficient, judging whether the minimum spacing adjustment coefficient is within the adjustment coefficient set interval, if yes, the result of multiplying the minimum spacing adjustment coefficient and the current sampling point minimum spacing is taken as the next sampling point minimum spacing, if not, judging whether the minimum spacing adjustment coefficient is greater than the adjustment coefficient set upper limit, if yes, the adjustment coefficient correction amount is input into the minimum spacing mapping table for index query to obtain the minimum spacing positive adjustment amount, the result of superimposing the minimum spacing positive adjustment amount and the current sampling point minimum spacing is taken as the preliminary adjusted sampling point minimum spacing, the result of multiplying the preliminary adjusted sampling point minimum spacing and the adjustment coefficient correction amount is taken as the next sampling point minimum spacing, if not, the adjustment coefficient correction amount is input into the minimum spacing mapping table for index query to obtain the minimum spacing negative adjustment amount, the result of subtracting the minimum spacing negative adjustment amount from the current sampling point minimum spacing is taken as the preliminary adjusted sampling point minimum spacing, the result of multiplying the preliminary adjusted sampling point minimum spacing and the adjustment coefficient correction amount is taken as the next sampling point minimum spacing.

[0051] As a further specific description, judging whether to perform sampling point minimum spacing dynamic adjustment, the specific judgment flow is:

[0052] If the soil sampling point boundary coverage is less than the boundary coverage set value, the result of arithmetically averaging the boundary coverage correction amount and the health data detection correction amount is input into the minimum spacing mapping table for index query to obtain the minimum spacing adjustment coefficient, which not only avoids uneven sampling point distribution and large spatial interpolation error caused by insufficient boundary coverage, but also ensures that the data quality after the minimum spacing is corrected meets the standard through detection accuracy correction, providing uniform and accurate sampling data support for spatial interpolation of soil health grading evaluation, and the boundary coverage correction amount represents the difference between the soil sampling point boundary coverage and the boundary coverage set value.

[0053] If yes, the result of multiplying the minimum interval adjustment coefficient and the current sampling point minimum interval is taken as the next sampling point minimum interval, which not only avoids uneven distribution of sampling points caused by excessive or insufficient interval adjustment (such as excessive encryption increasing cost, insufficient adjustment still having coverage gap), but also ensures that the spatial distribution of sampling points adapts to soil boundary heterogeneity, provides uniform and efficient sampling data support for spatial interpolation of soil health classification evaluation, reduces the risk of spatial interpolation error and evaluation distortion caused by improper interval, and the interval adjustment coefficient setting interval represents a closed interval formed by the interval adjustment coefficient setting lower limit and the interval adjustment coefficient setting upper limit. Multiplication processing refers to multiplication operation.

[0054] If no, it is judged whether the minimum interval adjustment coefficient is greater than the interval adjustment coefficient setting upper limit, if yes, the adjustment coefficient correction amount is input into the minimum interval mapping table for index query to obtain the minimum interval positive adjustment amount, the result of superimposing the minimum interval positive adjustment amount and the current sampling point minimum interval is taken as the preliminary adjusted sampling point minimum interval, and the result of multiplying the preliminary adjusted sampling point minimum interval and the adjustment coefficient correction amount is taken as the next sampling point minimum interval, which not only avoids sampling redundancy and cost waste caused by excessive reduction of interval due to too high adjustment coefficient, but also ensures that the adjusted interval still adapts to soil boundary heterogeneity and data detection precision requirement. The adjustment coefficient correction amount represents the difference between the minimum interval adjustment coefficient and the interval adjustment coefficient setting upper limit. Superimposition processing refers to addition operation.

[0055] If no, the adjustment coefficient correction amount is input into the minimum interval mapping table for index query to obtain the minimum interval negative adjustment amount, the result of subtracting the minimum interval negative adjustment amount from the current sampling point minimum interval is taken as the preliminary adjusted sampling point minimum interval, and the result of multiplying the preliminary adjusted sampling point minimum interval and the adjustment coefficient correction amount is taken as the next sampling point minimum interval, which not only avoids insufficient coverage of soil boundary and increase of spatial interpolation error caused by too large interval, but also ensures that the reduced interval adapts to soil boundary heterogeneity and data detection precision requirement, provides high-density and accurate sampling data support for spatial interpolation of soil health classification evaluation, reduces the risk of spatial interpolation distortion or insufficient data representation caused by improper interval, and the adjustment coefficient correction amount represents the difference between the minimum interval adjustment coefficient and the interval adjustment coefficient setting lower limit. Difference processing refers to subtraction operation.

[0056] If the boundary coverage rate of the soil sampling point is greater than or equal to the set boundary coverage rate, the minimum spacing of the soil sampling point is not adjusted, the current minimum spacing of the soil sampling point is maintained, unnecessary spacing adjustment is avoided to cause waste of sampling resources (such as excessive encryption to increase cost) or damage to the rationality of the spatial distribution of the existing sampling points, the spatial distribution of the sampling points can continue to adapt to the soil boundary heterogeneity, uniform and sufficient sampling data support is provided for the spatial interpolation of the soil health classification evaluation, and the risk of spatial interpolation error or resource consumption caused by excessive adjustment is reduced.

[0057] In the embodiment, by dynamically judging the relationship between the boundary coverage rate of the soil sampling point and the set value, combining the boundary coverage correction amount and the health data detection correction amount, the minimum spacing adjustment coefficient is calculated in real time, and different adjustment strategies are adopted according to whether it is in the set interval of the adjustment coefficient, including multiplication processing, superposition processing or difference processing, so as to realize accurate regulation and control of the minimum spacing of the sampling point. This process not only effectively avoids the problems of uneven distribution of sampling points, increased spatial interpolation error, sampling redundancy or resource waste caused by insufficient boundary coverage or excessive adjustment, but also ensures that the spatial distribution of the sampling points always adapts to the soil boundary heterogeneity and the data detection accuracy requirement, provides uniform, efficient and accurate sampling data support for the spatial interpolation of the soil health classification evaluation, significantly reduces the risk of evaluation distortion caused by improper spacing, and maintains the current spacing when the adjustment condition is not met, further optimizing the resource utilization efficiency and system stability.

[0058] As the third module of the soil health classification evaluation system, the remote sensing-sampling data fusion module is used to obtain the remote sensing verification deviation parameter in the remote sensing-sampling data fusion link to obtain the soil health grade remote sensing verification deviation degree, which is used to quantify the deviation degree of the verification conclusion from the true situation when the remote sensing data verifies the soil health classification result.

[0059] It should be noted that the remote sensing verification deviation parameter includes the to-be-compared soil health data detection evaluation value, the remote sensing-sampling time interval and the soil sampling time frequency; wherein, the to-be-compared soil health data detection evaluation value means that if the soil health data detection optimization is performed, the re-acquired soil health data detection evaluation value is recorded as the to-be-compared soil health data detection evaluation value, and if not, the current soil health data detection evaluation value is recorded as the to-be-compared soil health data detection evaluation value; the remote sensing-sampling time interval refers to the time difference between the time of remote sensing image acquisition and the time of soil sampling on the ground; and the soil sampling time frequency refers to the number of soil sampling times per unit time in the soil sample area.

[0060] The to-be-compared soil health data detection evaluation value and the data detection threshold value are multiplied by the data detection influence factor to obtain the data detection influence value. Wherein, the ratio refers to the division operation.

[0061] The time interval influence value is obtained by multiplying the time interval influence factor by the relative deviation proportion result of the remote sensing-sampling time interval and the time interval threshold value. The relative deviation proportion result refers to the ratio of the absolute value of the difference between the remote sensing-sampling time interval and the time interval threshold value to the time interval threshold value.

[0062] The time frequency influence value is obtained by multiplying the time frequency influence factor by the relative deviation proportion result of the soil sampling time frequency and the time frequency threshold value. The relative deviation proportion result refers to the ratio of the absolute value of the difference between the soil sampling time frequency and the time frequency threshold value to the time frequency threshold value.

[0063] The soil health grade remote sensing verification deviation degree is obtained by coupling the data detection influence value, the time interval influence value, and the time frequency influence value. The coupling refers to addition operation, and the multiplication refers to multiplication operation.

[0064] It should be noted that the data detection influencing factors, data detection threshold, time interval influencing factors, time interval threshold, time frequency influencing factors and time frequency threshold are obtained from the health grading evaluation database. There is a correlation between the remote sensing verification deviation parameters, as follows: the three of the to-be-compared soil health data detection evaluation value, the remote sensing-sampling time interval and the soil sampling time frequency have a close correlation, which jointly determines the timeliness, accuracy and comparability of soil health monitoring. The to-be-compared soil health data detection evaluation value is a quantitative index of soil health state obtained by field sampling analysis, and its accuracy depends on the reasonable setting of sampling frequency and the timely supplement of remote sensing data; the remote sensing-sampling time interval refers to the time difference between remote sensing image acquisition and field sampling, if the interval is too long, it will lead to the time mismatch between remote sensing data and field sampling data, weaken the effectiveness of the comparison between the two, and thus increase the to-be-compared soil health data detection evaluation value; the soil sampling time frequency refers to the number of field sampling per unit time, the higher the frequency, the better the dynamic changes of soil health state can be captured, but if the frequency is too high and the remote sensing data update is lagging, it will also cause data redundancy or time synchronization problem, and increase the to-be-compared soil health data detection evaluation value. There is a correlation between the remote sensing verification deviation parameters and the soil health grade remote sensing verification deviation degree, as follows: the three of the to-be-compared soil health data detection evaluation value, the remote sensing-sampling time interval and the soil sampling time frequency have a positive correlation with the soil health grade remote sensing verification deviation degree, which jointly affects the consistency between the remote sensing inversion result and the field soil health state. The to-be-compared soil health data detection evaluation value is a quantitative index of soil health obtained by field sampling analysis, and its accuracy and representativeness directly affect the size of the remote sensing verification deviation degree, if the to-be-compared soil health data detection evaluation value is larger, it will lead to the verification benchmark of the remote sensing inversion result being unreliable, thus magnifying the deviation degree; the remote sensing-sampling time interval refers to the time difference between remote sensing image acquisition and field sampling, if the interval is too long, the soil health state may change during this period, making the remote sensing data and field data not match in time, thus increasing the verification deviation degree; the soil sampling time frequency reflects the density and rhythm of field sampling, if the frequency is too low, the dynamic changes of soil health state will be ignored, the fluctuations of soil health grade cannot be captured in time, and thus the training and verification accuracy of the remote sensing inversion model will be affected, finally leading to the increase of the deviation degree.

[0065] As the fourth module of the soil health grading evaluation system: the health grade remote sensing verification optimization module is used to determine whether to perform health grade remote sensing verification optimization according to the soil health grade remote sensing verification deviation degree, if yes, then perform soil health grade division after health grade remote sensing verification optimization, if no, directly perform soil health grade division, the health grade remote sensing verification optimization includes dynamic adjustment of data fusion time window and dynamic adjustment of remote sensing image spatial resolution.

[0066] Further, the judgment of whether to perform the health level remote sensing verification optimization is specifically as follows: if the soil health level remote sensing verification deviation degree is not greater than the verification deviation degree reference value, the health level remote sensing verification optimization is not performed; if the soil health level remote sensing verification deviation degree is greater than the verification deviation degree reference value, whether to perform the data fusion time window dynamic adjustment is judged according to the remote sensing image acquisition frequency; if yes, after the data fusion time window dynamic adjustment, whether to perform the remote sensing image spatial resolution dynamic adjustment is judged; if no, whether to perform the remote sensing image spatial resolution dynamic adjustment is directly judged.

[0067] It needs to be explained that the adjustment of the remote sensing-soil sampling data fusion time window based on the remote sensing image acquisition frequency is essentially to dynamically match the time acquisition rhythm of the remote sensing data and the change rule of the soil health index, to solve the problems of the mismatch of the remote sensing inversion index and the ground measured index, the spectral response confusion and the data matching deviation from the time dimension, to reduce the accumulation of soil data errors, to avoid the distortion of the soil health grading evaluation, and to further reduce the influence of the accumulation of soil data errors on the grading evaluation. In summary, the adjustment of the remote sensing-soil sampling data fusion time window based on the remote sensing image acquisition frequency can solve the problems of index mismatch, spectral misjudgment and data deviation from the time dimension through the dynamic matching of "frequency-window", reduce the accumulation of soil data errors at each link, ensure the accuracy of the remote sensing and sampling data fusion, provide reliable data support for the soil health grading evaluation, and avoid the distortion of the evaluation results.

[0068] As a further specific description, the specific steps of judging whether to perform the data fusion time window dynamic adjustment are as follows:

[0069] If the remote sensing image acquisition frequency is greater than the acquisition frequency reference upper limit, the result of the weighted average of the verification deviation degree correction amount and the acquisition frequency correction amount is input into the fusion time window mapping table for index query to obtain a window width correction factor. The verification deviation degree correction amount represents the difference between the soil health level remote sensing verification deviation degree and the verification deviation degree reference value, which can generate a window adjustment basis by synchronously combining the remote sensing verification deviation degree and the frequency over-limit degree when the image acquisition frequency is over-limit, to provide a precise reference for the subsequent optimization of the fusion time window width, to avoid the mixing of short-term interference (such as temporary rainfall and fertilization) into the data and the influence on the remote sensing-sampling data fusion accuracy due to the over-wide window at high-frequency acquisition, and to provide an adaptive high-frequency image and accurate and reliable time dimension support for the remote sensing data processing of the soil health grading evaluation, to reduce the risk of fusion data distortion caused by improper window, and the acquisition frequency correction amount represents the difference between the remote sensing image acquisition frequency and the acquisition frequency reference upper limit.

[0070] determining whether the window width correction factor is greater than the width correction factor reference value, if yes, inputting the width correction factor correction amount into the fusion time window mapping table for index query to obtain a fusion time window width reduction coefficient, multiplying the fusion time window width reduction coefficient with the current remote sensing-soil sampling data fusion time window to obtain an upper integer result as the remote sensing-soil sampling data fusion time window of next time, which not only avoids short-term interference (such as temporary precipitation, fertilization) data in high-frequency images due to too wide window, ensures the fusion accuracy of remote sensing-soil sampling data, and also ensures the window width to be a reasonable integer unit through the upper integer to adapt to the actual data processing needs, provides high-quality fusion data support for soil health grading evaluation, reduces the evaluation distortion risk caused by improper window, and the width correction factor correction amount represents the difference between the window width correction factor and the width correction factor reference value. The multiplication processing refers to multiplication operation.

[0071] if no, inputting the width correction factor compensation amount into the fusion time window mapping table for index query to obtain a fusion time window width expansion coefficient, multiplying the fusion time window width expansion coefficient with the current remote sensing-soil sampling data fusion time window to obtain a lower integer result as the remote sensing-soil sampling data fusion time window of next time, which not only avoids insufficient effective data in high-frequency images due to too narrow window, affecting the integrity of remote sensing-soil sampling data fusion, but also ensures the window width to be an integer unit meeting the actual processing needs through the lower integer, and at the same time, the compensation amount is combined to adapt to the deviation and frequency demand, and the width correction factor compensation amount represents the difference between the window width correction factor and the width correction factor reference value.

[0072] In the embodiment, by dynamically adjusting the remote sensing-soil sampling data fusion time window width, the optimization control of the soil health data fusion accuracy and reliability under the high-frequency remote sensing image acquisition environment is realized. The whole mechanism dynamically responds to the remote sensing acquisition frequency and the deviation state, realizes the adaptive optimization of the fusion time window under the premise of ensuring the accuracy of the remote sensing verification of soil health grade, significantly reduces the data distortion risk caused by improper window width setting, and provides high timeliness and high precision remote sensing-sampling fusion data support for soil health grading evaluation.

[0073] As a further specific description, determining whether to perform data fusion time window dynamic adjustment further includes:

[0074] If the remote sensing image acquisition frequency is within the acquisition frequency reference interval, no data fusion time window dynamic adjustment is performed, the current remote sensing-soil sampling data fusion time window can be maintained when the image acquisition frequency is reasonable and stable, which avoids unnecessary window adjustment to break the stability of existing data fusion and increase processing cost, ensures that the window width is adapted to the image acquisition rhythm, and can stably screen effective data without extreme interference for fusion, provides continuous and precision-standard fusion data support for soil health grading evaluation, reduces the risk of fusion error or resource waste caused by excessive adjustment, and the acquisition frequency reference interval represents a closed interval formed by the acquisition frequency reference lower limit and the acquisition frequency reference upper limit.

[0075] If the remote sensing image acquisition frequency is less than the acquisition frequency reference lower limit, the weighted average result of the verification deviation correction amount and the acquisition frequency compensation amount is input into the fusion time window mapping table for index query to obtain the window width control factor, which can generate window adjustment basis by combining the remote sensing verification deviation degree and the frequency deficiency degree when the image acquisition frequency is insufficient, provide accurate reference for optimizing the fusion time window width, avoid insufficient effective data amount and affect the integrity of remote sensing-soil sampling data fusion due to narrow window when the acquisition frequency is low, and ensure that the window adjustment does not deviate from the soil health grade verification accuracy requirement by correcting the verification deviation degree, provide time dimension support for remote sensing data processing for soil health grading evaluation that adapts to low-frequency images and complete data, reduce the risk of incomplete or insufficient precision of fusion data caused by improper window, and the acquisition frequency compensation amount represents the difference between the acquisition frequency reference lower limit and the remote sensing image acquisition frequency.

[0076] If the window width control factor is greater than the width control factor reference value, the width control factor reference value is input into the fusion time window mapping table for index query to obtain the fusion time window width extension coefficient, and the multiplication result of the fusion time window width extension coefficient and the current remote sensing-soil sampling data fusion time window is taken as the next remote sensing-soil sampling data fusion time window, which avoids insufficient effective data amount and ensures the integrity of remote sensing-soil sampling data fusion due to low remote sensing image acquisition frequency and short window, ensures that the window width is an integer unit that meets the actual data processing requirements by taking the integer part, and provides data complete and precision-standard fusion data support for soil health grading evaluation by adapting to the verification deviation and frequency requirement, reduces the risk of incomplete or distorted evaluation caused by improper window, and the width control factor reference value represents the difference between the window width control factor and the width control factor reference value.

[0077] If not, the width control factor lag amount is input into the fusion time window mapping table for index query to obtain a fusion time window width compression coefficient, and a multiplication result of the fusion time window width compression coefficient and the current remote sensing-soil sampling data fusion time window is taken as an integral result as the next remote sensing-soil sampling data fusion time window, which not only avoids poor data timeliness caused by low remote sensing image acquisition frequency and wide window (such as including out-of-date soil state data), but also ensures the timeliness of remote sensing-soil sampling data fusion and the window width to be an integral unit meeting actual data processing requirements, and combines the lag amount with the deviation and frequency requirement to provide fusion data support with high timeliness and precision for soil health grading evaluation, and the width control factor lag amount represents a difference between the width control factor and a width control factor reference value.

[0078] In the embodiment, by establishing a dynamic mapping mechanism of remote sensing image acquisition frequency and fusion time window width, intelligent regulation and optimization of the remote sensing-soil sampling data fusion process under different frequency conditions are realized. The mechanism not only adaptively adjusts the fusion time window according to the real-time change of the image acquisition frequency, but also ensures that the window adjustment is always consistent with the soil health grade verification precision requirement by introducing a verification deviation correction amount, effectively reducing the risk of data distortion, incomplete fusion or poor timeliness caused by improper window setting, and providing high-quality fusion data support with stability, integrity and timeliness for soil health grading evaluation.

[0079] It should be noted that adjusting the remote sensing image spatial resolution based on the ground sampling depth is to match the spatial variation scale of the soil health index under different sampling depths with the spatial capture ability of remote sensing data, to solve the problems of mismatching of remote sensing indirect indicators and ground direct indicators, spectral response confusion and data matching average deviation from the spatial dimension, to reduce the accumulation of soil data errors, to avoid distortion of soil health grading evaluation, and to reduce the accumulation of data errors by reasonable matching. In summary, adjusting the remote sensing image spatial resolution based on the ground sampling depth can solve the problems of index mismatching, spectral misjudgment and average deviation from the spatial dimension through precise matching of "sampling depth-spatial variation scale-remote sensing resolution", reduce the accumulation of soil data errors, ensure the consistency of remote sensing and sampling data in spatial scale, provide reliable spatial data support for soil health grading evaluation, and avoid distortion of evaluation results.

[0080] As a further specific description, whether to perform dynamic adjustment of the remote sensing image spatial resolution is judged, and the specific judgment steps are as follows:

[0081] If the ground sampling depth is greater than the upper limit of the sampling depth reference, the weighted average result of the depth correction amount and the verification deviation correction amount is input into the fusion time window mapping table for index query to obtain a spatial resolution convergence coefficient. The spatial resolution convergence coefficient is multiplied by the current remote sensing image spatial resolution, and the result is rounded up to be the spatial resolution of the next remote sensing image. When the sampling depth is over-limit, the image spatial resolution can be accurately adjusted in combination with the depth over-limit degree and the remote sensing verification deviation degree. The low spatial matching degree of remote sensing data and ground sampling data caused by too deep sampling depth and inappropriate resolution is avoided. The adjusted resolution is also ensured to be an integer unit that meets the actual processing requirements by rounding up. The resolution adjustment is also ensured not to deviate from the soil health evaluation accuracy requirement with the help of the verification deviation degree correction. The depth correction amount represents the difference between the ground sampling depth and the upper limit of the sampling depth reference. The multiplication processing refers to multiplication operation.

[0082] If the ground sampling depth is within the sampling depth reference interval, the remote sensing image spatial resolution is not dynamically adjusted, and the current remote sensing image spatial resolution is maintained. The unnecessary resolution adjustment is avoided to break the existing spatial matching balance of remote sensing data and ground sampling data and increase the data processing cost. The resolution is also ensured to be adapted to the sampling depth to stably ensure the fusion accuracy of the two. The sampling depth reference interval represents a closed interval formed by the lower limit of the sampling depth reference and the upper limit of the sampling depth reference.

[0083] If the ground sampling depth is less than the lower limit of the sampling depth reference, the weighted average result of the depth reference amount and the verification deviation correction amount is input into the fusion time window mapping table for index query to obtain a spatial resolution expansion coefficient. The spatial resolution expansion coefficient is multiplied by the current remote sensing image spatial resolution, and the result is rounded up to be the spatial resolution of the next remote sensing image. When the sampling depth is insufficient, the image spatial resolution can be accurately adjusted in combination with the depth missing degree and the remote sensing verification deviation degree. The remote sensing data cannot accurately capture the details of the shallow soil properties due to too shallow sampling depth and inappropriate resolution, which affects the spatial matching degree with the ground sampling data. The adjusted resolution is also ensured to be an integer unit that meets the actual data processing requirements by rounding up. The resolution adjustment is also ensured not to deviate from the soil health evaluation accuracy requirement with the help of the verification deviation degree correction. The depth reference amount represents the difference between the ground sampling depth and the lower limit of the sampling depth reference.

[0084] In this embodiment, by establishing a dynamic mapping mechanism between ground sampling depth and remote sensing image spatial resolution, the intelligent regulation and optimization of remote sensing data spatial matching accuracy and fusion reliability under different sampling depth conditions are realized. This mechanism not only can adaptively adjust the spatial resolution of remote sensing image according to the real-time change of ground sampling depth, but also can guarantee the scientificity and rationality of resolution adjustment by introducing verification deviation correction quantity, effectively reducing the risk of data fusion distortion, evaluation deviation or inaccurate spatial representation caused by depth-resolution mismatch, and providing remote sensing data support with strong spatial adaptability and high fusion accuracy for soil health grading evaluation.

[0085] As shown in Figure 4 The flow chart of a soil health grading evaluation method provided by the embodiment of the application is shown in the figure. The soil health grading evaluation method comprises: obtaining soil data required for soil health evaluation in a soil sample area and performing soil health detection to obtain soil health data detection evaluation values for quantifying the accuracy of soil health data in the soil health detection process; judging whether to perform soil health data detection optimization according to the soil health data detection evaluation values, if yes, performing a remote sensing-sampling data fusion link after soil health data detection optimization, and if no, directly performing the remote sensing-sampling data fusion link, wherein the soil health data detection optimization comprises dynamic adjustment of sampling-detection interval time and dynamic adjustment of minimum sampling point spacing; obtaining a remote sensing verification deviation parameter in the remote sensing-sampling data fusion link to obtain a soil health grade remote sensing verification deviation degree for quantifying the deviation degree of the verification conclusion from the true situation when the remote sensing data verify the soil health grading result; judging whether to perform health grade remote sensing verification optimization according to the soil health grade remote sensing verification deviation degree, if yes, performing soil health grade division after health grade remote sensing verification optimization, and if no, directly performing soil health grade division, wherein the health grade remote sensing verification optimization comprises dynamic adjustment of a data fusion time window and dynamic adjustment of remote sensing image spatial resolution.

[0086] In this embodiment, by establishing a double feedback mechanism of soil health data detection evaluation values and soil health grade remote sensing verification deviation degrees, dynamic optimization and precision control of the whole process of soil health evaluation are realized. The overall mechanism not only optimizes the sampling strategy and data quality through soil health data detection evaluation values, but also optimizes the remote sensing verification process and fusion accuracy through soil health grade remote sensing verification deviation degrees, forming a closed-loop feedback control from data collection, detection optimization to remote sensing verification and grade division, effectively reducing the risk of evaluation distortion caused by poor data quality, mismatch of time and space scales or large verification deviation, and providing high-precision, high-reliability and high-timeliness technical support for soil health grading evaluation.

[0087] The above merely illustrates the specific embodiments of the present application, but the protection scope of the present application is not limited thereto, any person skilled in the art can easily think of the changes or replacements within the technical range disclosed by the present application, which should be covered in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A soil health grading system, characterized in that, The system comprises a soil data acquisition module, a soil health data detection optimization module, a remote sensing-sampling data fusion module, and a health grade remote sensing verification optimization module. The soil data acquisition module is configured to acquire soil data required for soil health evaluation in a soil sample area and perform soil health detection to obtain a soil health data detection evaluation value, which is used to quantify the accuracy of soil health data in the soil health detection process. The soil health data detection optimization module is configured to determine whether to perform soil health data detection optimization according to the soil health data detection evaluation value. If yes, the remote sensing-sampling data fusion step is performed after the soil health data detection optimization. If no, the remote sensing-sampling data fusion step is directly performed. The soil health data detection optimization includes dynamic adjustment of sampling-detection interval time and dynamic adjustment of minimum sampling point spacing. The remote sensing-sampling data fusion module is configured to acquire a remote sensing verification deviation parameter in the remote sensing-sampling data fusion step to obtain a soil health grade remote sensing verification deviation degree, which is used to quantify the deviation degree of the verification conclusion from the true situation when the remote sensing data is used to verify the soil health grading result. The health grade remote sensing verification optimization module is configured to determine whether to perform health grade remote sensing verification optimization according to the soil health grade remote sensing verification deviation degree. If yes, the soil health grade division is performed after the health grade remote sensing verification optimization. If no, the soil health grade division is directly performed.

2. The soil health grading system of claim 1, wherein, The health grade remote sensing verification optimization includes dynamic adjustment of data fusion time window and dynamic adjustment of remote sensing image spatial resolution. The soil data includes soil sampling point density, soil sampling point average spacing, and soil property change rate. The density influence value is obtained by multiplying the soil sampling point density and the relative deviation proportion of the density threshold value. The spacing influence value is obtained by multiplying the soil sampling point average spacing and the relative deviation proportion of the spacing threshold value. The change rate influence value is obtained by multiplying the soil property change rate and the relative deviation proportion of the change rate threshold value. The soil health data detection evaluation value is obtained by coupling the density influence value, the spacing influence value, and the change rate influence value. The specific process of determining whether to perform soil health data detection optimization is as follows: If the soil health data detection evaluation value is less than or equal to the data detection set value, soil health data detection optimization is not performed. If the soil health data detection evaluation value is greater than the data detection set value, it is determined whether to perform dynamic adjustment of sampling-detection interval time according to the soil property change rate. If yes, it is determined whether to perform dynamic adjustment of minimum sampling point spacing after the dynamic adjustment of sampling-detection interval time. If no, it is directly determined whether to perform dynamic adjustment of minimum sampling point spacing. The specific determination process of determining whether to perform dynamic adjustment of sampling-detection interval time according to the soil property change rate is as follows: If the soil property change rate is less than the change rate setting lower limit, a change rate correction amount, which is used to represent a deviation degree between the change rate setting lower limit and the soil property change rate, and a health data detection correction amount, which is used to represent a deviation degree between the soil health data detection evaluation value and the data detection setting value, are harmonically averaged to obtain an interval time adjustment factor, which is input into the detection interval time mapping table for index query. If the interval time adjustment factor is greater than the adjustment factor setting value, an adjustment factor correction amount, which is used to represent a positive deviation degree between the interval time adjustment factor and the adjustment factor setting value, is input into the detection interval time mapping table for index query to obtain an interval time reduction amount, and a current sampling-detection interval time is reduced by the interval time reduction amount to obtain a preliminary adjusted interval time. The preliminary adjusted interval time and the adjustment factor correction amount are interactively processed to obtain a next sampling-detection interval time. If not, an adjustment factor reference amount, which is used to represent a negative deviation degree between the adjustment factor setting value and the interval time adjustment factor, is input into the detection interval time mapping table for index query to obtain an interval time increase amount, and the current sampling-detection interval time is added by the interval time increase amount to obtain a preliminary adjusted interval time. The preliminary adjusted interval time and the adjustment factor reference amount are interactively processed to obtain a next sampling-detection interval time.

3. A soil health grading system according to claim 2, wherein, The judgment whether to perform the sampling-detection interval time dynamic adjustment further includes: If the soil property change rate is within a change rate setting interval, which represents a closed interval formed by the change rate setting lower limit and a change rate setting upper limit, the sampling-detection interval time dynamic adjustment is not performed. If the soil property change rate is greater than the change rate setting upper limit, a change rate correction amount, which is used to represent a deviation degree between the soil property change rate and the change rate setting upper limit, and a health data detection correction amount are harmonically averaged to obtain an interval time regulation factor, which is input into the detection interval time mapping table for index query. If the interval time regulation factor is greater than a regulation factor setting value, a regulation factor correction amount, which is used to represent a positive deviation degree between the interval time regulation factor and the regulation factor setting value, is input into the detection interval time mapping table for index query to obtain an interval time increase amount, and a current sampling-detection interval time is coupled with the interval time increase amount to obtain a preliminary adjusted interval time. The preliminary adjusted interval time and the regulation factor correction amount are interactively processed to obtain a next sampling-detection interval time. If not, the regulatory factor reference amount is input into the detection interval time mapping table for index query to obtain the interval time reduction amount, the current sampling-detection interval time is reduced by the interval time reduction amount to obtain the preliminary adjusted interval time, and the result of the interaction between the preliminary adjusted interval time and the regulatory factor reference amount is taken as the next sampling-detection interval time, and the regulatory factor reference amount is used to represent the negative deviation degree of the regulatory factor set value from the interval time regulatory factor.

4. The soil health grading system of claim 2, wherein, The judgment of whether to perform the sampling point minimum distance dynamic adjustment includes the following specific judgment process: If the soil sampling point boundary coverage rate is less than the boundary coverage rate set value, the boundary coverage rate correction amount is used to represent the negative deviation degree of the soil sampling point boundary coverage rate from the boundary coverage rate set value, and the boundary coverage rate correction amount and the health data detection correction amount are arithmetically averaged to obtain the minimum distance adjustment coefficient, which is input into the minimum distance mapping table for index query. If the minimum distance adjustment coefficient is within the adjustment coefficient set interval, the minimum distance adjustment coefficient is multiplied by the current sampling point minimum distance to obtain the next sampling point minimum distance, and the adjustment coefficient set interval represents a closed interval formed by the adjustment coefficient set lower limit and the adjustment coefficient set upper limit. If not, it is judged whether the minimum distance adjustment coefficient is greater than the adjustment coefficient set upper limit. If yes, the adjustment coefficient correction amount is input into the minimum distance mapping table for index query to obtain the minimum distance positive adjustment amount, the minimum distance positive adjustment amount is superimposed on the current sampling point minimum distance to obtain the preliminary adjusted sampling point minimum distance, and the preliminary adjusted sampling point minimum distance is multiplied by the adjustment coefficient correction amount to obtain the next sampling point minimum distance. The adjustment coefficient correction amount is used to represent the positive deviation degree of the minimum distance adjustment coefficient from the adjustment coefficient set upper limit. If not, the adjustment coefficient correction amount is input into the minimum distance mapping table for index query to obtain the minimum distance negative adjustment amount, the current sampling point minimum distance is subtracted by the minimum distance negative adjustment amount to obtain the preliminary adjusted sampling point minimum distance, and the preliminary adjusted sampling point minimum distance is multiplied by the adjustment coefficient correction amount to obtain the next sampling point minimum distance. The adjustment coefficient correction amount is used to represent the negative deviation degree of the minimum distance adjustment coefficient from the adjustment coefficient set lower limit. If the soil sampling point boundary coverage rate is greater than or equal to the boundary coverage rate set value, the sampling point minimum distance dynamic adjustment is not performed.

5. The soil health grading system of claim 1, wherein, The remote sensing verification deviation parameter includes a to-be-compared soil health data detection evaluation value, a remote sensing-sampling time interval, and a soil sampling time frequency. The data detection influence factor is multiplied by the proportion of the to-be-compared soil health data detection evaluation value to the data detection threshold value to obtain a data detection influence value. The time interval influence factor is multiplied by the relative deviation proportion of the remote sensing-sampling time interval to the time interval threshold value to obtain a time interval influence value. The time-frequency influence value is obtained by multiplying the time-frequency influence factor with the relative deviation proportion of the time-frequency of soil sampling and the time-frequency threshold value; The data detection influence value, the time interval influence value and the time-frequency influence value are coupled to obtain the soil health grade remote sensing verification deviation degree.

6. The soil health grading system of claim 1, wherein, The specific steps of determining whether to perform health grade remote sensing verification optimization are as follows: If the soil health grade remote sensing verification deviation degree is not greater than the verification deviation degree reference value, health grade remote sensing verification optimization is not performed; If the soil health grade remote sensing verification deviation degree is greater than the verification deviation degree reference value, it is determined whether to perform data fusion time window dynamic adjustment according to the remote sensing image acquisition frequency. If yes, after the data fusion time window is dynamically adjusted, it is determined whether to perform remote sensing image spatial resolution dynamic adjustment. If no, it is directly determined whether to perform remote sensing image spatial resolution dynamic adjustment. The specific steps of determining whether to perform data fusion time window dynamic adjustment are as follows: If the remote sensing image acquisition frequency is greater than the acquisition frequency reference upper limit, the verification deviation degree correction amount and the acquisition frequency correction amount are weighted and averaged to obtain a window width correction factor, which is input into the fusion time window mapping table for index query. The verification deviation degree correction amount is used to represent the deviation degree of the soil health grade remote sensing verification deviation degree and the verification deviation degree reference value. The acquisition frequency correction amount is used to represent the positive deviation degree of the remote sensing image acquisition frequency and the acquisition frequency reference upper limit. If yes, a width correction factor correction amount is input into the fusion time window mapping table for index query to obtain a fusion time window width reduction coefficient. The fusion time window width reduction coefficient and the current remote sensing-soil sampling data fusion time window are multiplied and rounded up to obtain the next remote sensing-soil sampling data fusion time window. The width correction factor correction amount is used to represent the positive deviation degree of the window width correction factor and the width correction factor reference value. If no, a width correction factor compensation amount is input into the fusion time window mapping table for index query to obtain a fusion time window width expansion coefficient. The fusion time window width expansion coefficient and the current remote sensing-soil sampling data fusion time window are multiplied and rounded down to obtain the next remote sensing-soil sampling data fusion time window. The width correction factor compensation amount is used to represent the negative deviation degree of the window width correction factor and the width correction factor reference value.

7. A soil health grading system according to claim 6, wherein, The determination of whether to perform data fusion time window dynamic adjustment further includes: If the remote sensing image acquisition frequency is within the acquisition frequency reference interval, data fusion time window dynamic adjustment is not performed. The acquisition frequency reference interval represents a closed interval formed by the acquisition frequency reference lower limit and the acquisition frequency reference upper limit. If the remote sensing image acquisition frequency is less than the acquisition frequency reference lower limit, the weighted average result of the verification deviation correction amount and the acquisition frequency compensation amount is input into the fusion time window mapping table for index query to obtain a window width regulation factor, and the acquisition frequency compensation amount is used to represent the negative deviation degree of the acquisition frequency reference lower limit from the remote sensing image acquisition frequency. If the window width regulation factor is greater than the width regulation factor reference value, a width regulation factor reference amount is input into the fusion time window mapping table for index query to obtain a fusion time window width extension coefficient, and the multiplication result of the fusion time window width extension coefficient and the current remote sensing-soil sampling data fusion time window is taken as an integer to obtain the next remote sensing-soil sampling data fusion time window, and the width regulation factor reference amount is used to represent the positive deviation degree of the window width regulation factor from the width regulation factor reference value. If not, a width regulation factor lag amount is input into the fusion time window mapping table for index query to obtain a fusion time window width compression coefficient, and the multiplication result of the fusion time window width compression coefficient and the current remote sensing-soil sampling data fusion time window is taken as an integer to obtain the next remote sensing-soil sampling data fusion time window, and the width regulation factor lag amount is used to represent the negative deviation degree of the width regulation factor from the width regulation factor reference value.

8. The soil health grading system of claim 6, wherein, The specific judgment steps are as follows: If the ground sampling depth is greater than the sampling depth reference upper limit, the weighted average result of the depth correction amount and the verification deviation correction amount is input into the fusion time window mapping table for index query to obtain a spatial resolution convergence coefficient, and the multiplication result of the spatial resolution convergence coefficient and the current remote sensing image spatial resolution is taken as an integer to obtain the next remote sensing image spatial resolution, and the depth correction amount is used to represent the positive deviation degree of the ground sampling depth from the sampling depth reference upper limit. If the ground sampling depth is within the sampling depth reference interval, no dynamic adjustment of the remote sensing image spatial resolution is performed, and the sampling depth reference interval represents a closed interval formed by the sampling depth reference lower limit and the sampling depth reference upper limit. If the ground sampling depth is less than the sampling depth reference lower limit, the weighted average result of the depth reference amount and the verification deviation correction amount is input into the fusion time window mapping table for index query to obtain a spatial resolution expansion coefficient, and the multiplication result of the spatial resolution expansion coefficient and the current remote sensing image spatial resolution is taken as an integer to obtain the next remote sensing image spatial resolution, and the depth reference amount is used to represent the negative deviation degree of the ground sampling depth from the sampling depth reference lower limit.

9. A method of soil health grading using the soil health grading system according to any one of claims 1 to 8, wherein, It includes: Obtaining soil data required for soil health evaluation in the soil sample area and performing soil health detection to obtain soil health data detection evaluation values for quantifying the accuracy of soil health data in the soil health detection process; According to the soil health data detection evaluation value, it is judged whether to perform soil health data detection optimization. If yes, the remote sensing-sampling data fusion link is performed after the soil health data detection optimization. If no, the remote sensing-sampling data fusion link is directly performed. The soil health data detection optimization includes dynamic adjustment of sampling-detection interval time and dynamic adjustment of minimum sampling point interval. The soil health grade remote sensing verification deviation degree is obtained by acquiring a remote sensing verification deviation parameter in the remote sensing-sampling data fusion link, and is used to quantify the deviation degree of the verification conclusion from the true situation when the remote sensing data verifies the soil health grading result. According to the soil health grade remote sensing verification deviation degree, it is judged whether to perform health grade remote sensing verification optimization. If yes, the soil health grade division is performed after the health grade remote sensing verification optimization. If no, the soil health grade division is directly performed. The health grade remote sensing verification optimization includes dynamic adjustment of data fusion time window and dynamic adjustment of remote sensing image spatial resolution.

Images