Soil zinc content monitoring method and system based on hyperspectral remote sensing

Abstract

The invention relates to the field of soil pollution monitoring in ecologically fragile areas, in particular to a soil zinc content monitoring method and system based on hyperspectral remote sensing. The method comprises the following steps: acquiring spectral reflection data and corresponding zinc element content of a soil sample in an ecologically fragile area by utilizing a hyperspectral remote sensing technology; performing data enhancement on the spectral reflection data, and combining an enhancement result and the zinc element content to obtain an enhanced spectral data set; obtaining an optimal spectrum form through the enhanced spectrum data set; and on the basis of the optimal spectrum form, carrying out characteristic wave band optimization on the enhanced spectrum data set, and completing soil zinc content monitoring according to a priority result. According to the invention, based on the detection data of the soil sample in the ecologically fragile area, the zinc content of the soil is monitored by processing the data and constructing the classification learning model, and the problem of rapid and accurate monitoring of the zinc content of the soil in the ecologically fragile area is solved.

Description

Technical Field

[0001] The present invention relates to the field of soil pollution monitoring in ecologically fragile areas, and in particular to a method and system for monitoring soil zinc content based on hyperspectral remote sensing. Background Art

[0002] Soil is an important carrier of agricultural production and a major component of the ecological environment. Its quality directly affects people's daily lives. Since the content of zinc in the soil is relatively hidden and its performance is lagging, it has not received much attention. However, when the zinc in the soil exceeds a certain amount, it will affect the normal growth of plants and even accumulate in humans and animals through the food chain, posing a great threat to human health. Therefore, how to quickly and effectively detect the zinc content and spatial distribution in the soil is crucial for soil remediation and soil management.

[0003] The traditional method of monitoring the zinc content in soil is to collect a large number of samples and bring them back to the laboratory for measurement using atomic absorption spectroscopy, ultraviolet visible spectrophotometry, atomic fluorescence, inductively coupled plasma mass spectrometry, and inductively coupled plasma emission spectroscopy. This method requires a lot of manpower and material resources and is expensive. In particular, large-scale element content mapping is achieved using spatial interpolation methods, and the results are affected by factors such as the number of sampling points and the spatial distribution of samples. Although some studies have used hyperspectral technology to estimate the content of heavy metals in soil, most of them were conducted in areas with high soil metal content, and no research has been conducted on the zinc content in soil in ecologically fragile areas. It is also unclear what spectral form and characteristic band optimization method is suitable for zinc monitoring in ecologically fragile areas. Summary of the invention

[0004] In view of the shortcomings of existing methods and the needs of practical applications, in order to be applicable to soil zinc content monitoring in ecologically fragile areas, the problem of rapid and accurate monitoring of soil zinc content is solved. On the one hand, the present invention provides a soil zinc content monitoring method based on hyperspectral remote sensing, comprising the following steps: The spectral reflectance data and corresponding zinc content of soil samples in ecologically fragile areas are obtained by using hyperspectral remote sensing technology; the spectral reflectance data are enhanced, and an enhanced spectral data set is obtained by combining the enhancement result and the zinc content; the optimal spectral form is obtained through the enhanced spectral data set; based on the optimal spectral form, the characteristic band of the enhanced spectral data set is optimized, and soil zinc content monitoring is completed according to the priority result. The present invention is based on the detection data of soil samples in ecologically fragile areas, and the soil zinc content monitoring is completed by processing the data and constructing a classification learning model, which solves the problem of rapid and accurate monitoring of soil zinc content in ecologically fragile areas and improves the applicability of soil zinc content monitoring in ecologically fragile areas.

[0005] Optionally, the performing data enhancement on the spectral reflectance data comprises the following steps: The spectral reflection data is first enhanced to obtain a first enhancement result; the spectral reflection data is second enhanced to obtain a second enhancement result; the spectral reflection data is third enhanced to obtain a third enhancement result. The present invention is beneficial to improving the accuracy of the present invention by performing multiple data enhancements on the spectral reflection data.

[0006] Optionally, the first enhancement is performed on the spectral reflectance data to obtain a first enhancement result, which satisfies the following formula: in, represents the first enhancement result, Indicates band The wavelength of represents spectral reflectance data, Indicates the interval between two adjacent wavelengths.

[0007] Optionally, the second enhancement is performed on the spectral reflectance data to obtain a second enhancement result, which satisfies the following formula: in, represents the second enhancement result, Indicates band The wavelength of represents the first enhancement result, Indicates the interval between two adjacent wavelengths.

[0008] Optionally, obtaining an optimal spectral form by enhancing the spectral dataset comprises the following steps: The enhanced spectral data set is split to obtain a spectral data training set and a spectral data verification set; a first machine classification model is constructed, the spectral data training set is used to train the first machine classification model, and the spectral data verification set and the trained first machine classification model are combined to obtain a first verification result; the first verification result is comprehensively analyzed using the determination coefficient, the relative root mean square error and the relative analysis error to obtain the optimal spectral form. The present invention uses a variety of error analysis methods, which is conducive to quickly and accurately obtaining the optimal spectral form.

[0009] Optionally, the method of optimizing characteristic bands of the enhanced spectral data set based on the optimal spectral form and completing soil zinc content monitoring according to the priority results comprises the following steps: Based on the optimal spectral form, the enhanced spectral data set is screened to obtain an optimal spectral data set; the optimal spectral data set is divided to obtain an optimal spectral training data set and an optimal spectral verification data set; the optimal band in the optimal spectral training data set is obtained through a variety of band selection algorithms; a second machine classification model is constructed, the second machine classification model is trained using the optimal band, and the second verification result is obtained by combining the optimal spectral verification data set and the trained second machine classification model; the second verification result is comprehensively analyzed using the determination coefficient, relative root mean square error and relative analysis error to obtain an optimal band selection algorithm; based on the optimal band selection algorithm, soil zinc content monitoring is completed through the trained second machine classification model. The present invention uses data in the optimal spectral form for comprehensive analysis, which further improves the accuracy of the present invention.

[0010] Optionally, the band selection algorithm includes a minimum absolute shrinkage and selection operator algorithm, a genetic algorithm and a gradient boosting decision tree algorithm.

[0011] Optionally, the determination coefficient satisfies the following formula: in, represents the coefficient of determination; The relative root mean square error satisfies the following formula: in, represents the relative root mean square error; The relative analysis error satisfies the following formula: in, represents the relative analytical error, , and The samples are The true value, predicted value and average value of .

[0012] Optionally, the machine classification model is constructed based on the leave-one-out method. The present invention uses the leave-one-out method to construct the machine classification model, which can make full use of data and has high accuracy of results.

[0013] In the second aspect, in order to efficiently execute the soil zinc content monitoring method based on hyperspectral remote sensing provided by the present invention, the present invention also provides a soil zinc content monitoring system based on hyperspectral remote sensing, including a processor, an input device, an output device and a memory, wherein the processor, the input device, the output device and the memory are interconnected, wherein the memory is used to store a computer program, the computer program includes program instructions, and the processor is configured to call the program instructions to execute the soil zinc content monitoring method based on hyperspectral remote sensing as described in the first aspect of the present invention. The soil zinc content monitoring system based on hyperspectral remote sensing of the present invention has a compact structure and stable performance, and can stably execute the soil zinc content monitoring method based on hyperspectral remote sensing provided by the present invention, further improving the overall applicability and practical application ability of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 A flow chart of a soil zinc content monitoring method based on hyperspectral remote sensing provided by an embodiment of the present invention; Figure 2 A framework diagram of a soil zinc content monitoring system based on hyperspectral remote sensing provided by an embodiment of the present invention; Figure 3 A schematic diagram of the structure of a soil zinc content monitoring device based on hyperspectral remote sensing provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0015] The specific embodiments of the present invention will be described in detail below. It should be noted that the embodiments described herein are only for illustration and are not intended to limit the present invention. In the following description, a large number of specific details are set forth in order to provide a thorough understanding of the present invention. However, it is obvious to those of ordinary skill in the art that these specific details do not need to be adopted to implement the present invention. In other examples, in order to avoid confusing the present invention, known circuits, software or methods are not specifically described.

[0016] Throughout the specification, references to "one embodiment," "an embodiment," "an example," or "an example" mean that a particular feature, structure, or characteristic described in conjunction with the embodiment or example is included in at least one embodiment of the present invention. Therefore, the phrases "in one embodiment," "in an embodiment," "an example," or "an example" appearing in various places throughout the specification do not necessarily all refer to the same embodiment or example. In addition, particular features, structures, or characteristics may be combined in one or more embodiments or examples in any suitable combination and / or subcombination. In addition, it should be understood by those of ordinary skill in the art that the figures provided herein are for illustrative purposes and that the figures are not necessarily drawn to scale.

[0017] See also Figure 1In order to be applicable to soil zinc content monitoring in ecologically fragile areas and solve the problem of rapid and accurate soil zinc content monitoring, the present invention provides a soil zinc content monitoring method based on hyperspectral remote sensing, such as Figure 1 As shown, in one embodiment, the method includes the following steps: S1. Use hyperspectral remote sensing technology to obtain spectral reflectance data and corresponding zinc content of soil samples in ecologically fragile areas.

[0018] In the embodiment, the soil samples in the ecologically fragile area are brought back to the room for drying and impurity removal, and then ground into 200-mesh powder samples using a grinder of model RS200. Based on the hyperspectral remote sensing technology, the obtained powder samples are spectrally measured using a FieldSpec4 portable spectrometer in an indoor darkroom environment to obtain spectral reflectance data of the soil samples in the ecologically fragile area; Furthermore, the zinc content of each sample after measuring the spectral reflectance data was measured using a Niton XL3t 950 portable X-ray fluorescence spectrometer. During the measurement, the instrument working mode was set to the mineral (Cu / Zn) mode, and the measurement time of each sample was set to 120 s.

[0019] S2. Perform data enhancement on the spectral reflectance data, and combine the enhancement result with the zinc content to obtain an enhanced spectral data set.

[0020] In the embodiment, the step S2 of performing data enhancement on the spectral reflectance data comprises the following steps: S21. Perform a first enhancement on the spectral reflectance data to obtain a first enhancement result.

[0021] Specifically, the spectral reflectance data is first enhanced to obtain a first enhancement result that satisfies the following formula: in, represents the first enhancement result, Indicates band The wavelength of represents spectral reflectance data, Indicates the interval between two adjacent wavelengths.

[0022] S22: Perform a second enhancement on the spectral reflectance data to obtain a second enhancement result.

[0023] Specifically, the second enhancement is performed on the spectral reflectance data to obtain a second enhancement result, which satisfies the following formula: in, represents the second enhancement result, Indicates band The wavelength of represents the first enhancement result, Indicates the interval between two adjacent wavelengths.

[0024] S23: Perform a third enhancement on the spectral reflectance data to obtain a third enhancement result.

[0025] Specifically, the third enhancement of the spectral reflectance data is performed by connecting the peak points protruding on the spectral curve point by point, and making the outer angle of the broken line at the peak point greater than 180°, thereby forming an envelope, and then dividing the value on the original spectral curve by the corresponding value on the envelope to perform spectral de-envelope processing. In the embodiment, it can be obtained by the Continuum Removed function under the Specrral module in ENVI.

[0026] Furthermore, various enhancement results and the corresponding zinc content are combined to obtain enhanced spectral data corresponding to various spectral forms, and then summarized to form an enhanced spectral data set.

[0027] S3. Obtaining an optimal spectral form through the enhanced spectral dataset.

[0028] In the embodiment, the step S3 of obtaining the optimal spectral form by enhancing the spectral data set comprises the following steps: S31, splitting the enhanced spectral data set to obtain a spectral data training set and a spectral data verification set.

[0029] The enhanced spectral dataset was randomly divided into 2:1 to obtain the spectral data training set and the spectral data validation set.

[0030] S32: construct a first machine classification model, use the spectral data training set to train the first machine classification model, and combine the spectral data verification set and the trained first machine classification model to obtain a first verification result.

[0031] Specifically, a first machine classification model is constructed based on the leave-one-out method, the first machine classification model is trained using the spectral data training set, and then, the spectral data verification set and the trained first machine classification model are combined to obtain a first verification result.

[0032] S33. Comprehensively analyze the first verification result using the determination coefficient, relative root mean square error and relative analytical error to obtain the optimal spectral form.

[0033] Specifically, based on the performance of the corresponding data of each spectral form in the validation set, the spectral form with the best model effect is comprehensively determined in combination with the determination coefficient, relative root mean square error and relative analysis error. The larger the determination coefficient and relative analysis error, and the smaller the relative root mean square error, the better the model effect.

[0034] Furthermore, the coefficient of determination satisfies the following formula: The relative root mean square error satisfies the following formula: The relative analysis error satisfies the following formula: in, , and The samples are The true value, predicted value and average value of .

[0035] S4. Based on the optimal spectral form, the enhanced spectral data set is optimized for characteristic bands, and the soil zinc content monitoring is completed according to the priority results.

[0036] In an embodiment, the step S4 performs characteristic band optimization on the enhanced spectral data set based on the optimal spectral form, and completes soil zinc content monitoring according to the priority result, including the following steps: S41. Based on the optimal spectral form, the enhanced spectral dataset is screened to obtain an optimal spectral dataset.

[0037] Specifically, the enhanced spectral data in a non-optimal spectral form in the above step is removed from the enhanced spectral data set to obtain an optimal spectral data set.

[0038] S42, dividing the optimal spectral data set to obtain an optimal spectral training data set and an optimal spectral verification data set.

[0039] The optimal spectral dataset was randomly divided into 2:1 to obtain the optimal spectral training dataset and the optimal spectral verification dataset.

[0040] S43, obtaining the optimal band in the optimal spectrum training data set through multiple band selection algorithms.

[0041] In an embodiment, the band selection algorithm includes a minimum absolute shrinkage and selection operator algorithm, a genetic algorithm, and a gradient boosting decision tree algorithm.

[0042] Specifically, obtaining the optimal band in the optimal spectrum training data set by the minimum absolute shrinkage and selection operator algorithm includes the following steps: For all variables, i.e., bands, the Lasso class in the linear_model module of the Scikit-learn machine learning library is used for preliminary training to obtain the initial model; The importance of each variable is obtained and ranked through the coef_ parameter, and then the first N features are added to the initial model for verification again. The number of features when the initial model is optimal is determined by the performance of the initial model in the optimal spectral training data set, and the optimal band corresponding to the minimum absolute shrinkage and selection operator algorithm is obtained.

[0043] Furthermore, obtaining the optimal band in the optimal spectrum training data set by a genetic algorithm comprises the following steps: All variables are preliminarily trained using the RandomForestRegressor class in the ensemble module of the Scikit-learn machine learning library to obtain the initial model; Based on the ranking_ parameter, the RFECV class in the feature_selection module of the Scikit-learn machine learning library is used to automatically select features and obtain the optimal band corresponding to the genetic algorithm.

[0044] Furthermore, obtaining the optimal band in the optimal spectrum training data set by using a gradient boosting decision tree algorithm includes the following steps: All variables are preliminarily trained using the GradientBoostingRegress class in the ensemble module of the Scikit-learn machine learning library to obtain the initial model; The importance of each variable is obtained and ranked through the feature_importances_ parameter, and the first N features are added to the initial model. The number of features when the initial model is optimal is determined by the performance of the initial model in the optimal spectral training data set, and the optimal band corresponding to the gradient boosting decision tree algorithm is obtained.

[0045] In other embodiments, other model algorithms that can extract band features, such as neural networks, etc., may also be adopted.

[0046] S44, constructing a second machine classification model, using the optimal band to train the second machine classification model, combining the optimal spectrum verification data set and the trained second machine classification model to obtain a second verification result.

[0047] Specifically, a second machine classification model is constructed based on the leave-one-out method, and the second machine classification model is trained using the optimal band data obtained by various band selection algorithms, and then the second verification result is obtained by combining the optimal spectral verification data set and the trained second machine classification model.

[0048] S45. Comprehensively analyze the second verification result using the determination coefficient, relative root mean square error and relative analysis error to obtain an optimal band selection algorithm.

[0049] Specifically, based on the performance of the corresponding data of each band selection algorithm in the validation set, the band selection algorithm with the best model effect is comprehensively determined in combination with the determination coefficient, relative root mean square error and relative analysis error. The larger the determination coefficient and relative analysis error, and the smaller the relative root mean square error, the better the model effect.

[0050] Furthermore, the coefficient of determination satisfies the following formula: The relative root mean square error satisfies the following formula: The relative analysis error satisfies the following formula: in, , and The samples are The true value, predicted value and average value of .

[0051] S46. Based on the optimal band selection algorithm, soil zinc content monitoring is completed through the trained second machine classification model.

[0052] In the embodiment, the optimal band is extracted from the spectral data of the soil in the monitored area in the ecologically fragile area by using the optimal band selection algorithm, and the soil zinc content monitoring is completed using the trained second machine classification model.

[0053] See also Figure 2 In an embodiment, in order to efficiently execute the soil zinc content monitoring method based on hyperspectral remote sensing provided by the present invention, the present invention also provides a soil zinc content monitoring system based on hyperspectral remote sensing, including: an input device, an output device, a processor, and a memory, wherein the input device, the output device, the processor, and the memory are interconnected, and the memory contains program instructions, and the program instructions are used for the steps of the soil zinc content monitoring method based on hyperspectral remote sensing. The soil zinc content monitoring system based on hyperspectral remote sensing of the present invention has a compact structure and stable performance, and can stably execute the soil zinc content monitoring method based on hyperspectral remote sensing of the present invention, further improving the overall applicability and practical application ability of the present invention.

[0054] In an embodiment, the processor may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. An input device may be used to obtain data information. An output device may be used to output the results obtained by storing program instructions contained in a computer program in a memory provided by the present invention. The memory may include a read-only memory and a random access memory, and provide instructions and data to the processor. A portion of the memory may also include a non-volatile random access memory.

[0055] In yet another alternative embodiment, see Figure 3 In order to efficiently execute the soil zinc content monitoring method based on hyperspectral remote sensing provided by the present invention, this embodiment also provides a soil zinc content monitoring device based on hyperspectral remote sensing, such as Figure 3 As shown, including: The memory 10 is used to store computer programs; the processor 20 is used to execute the computer programs to implement the above-mentioned soil zinc content monitoring method based on hyperspectral remote sensing. The memory 10, the processor 20, the communication interface 31 and the communication bus 32. The memory 10, the processor 20 and the communication interface 31 all communicate with each other through the communication bus 32.

[0056] In an embodiment, the memory 10 is used to store one or more program instructions. The memory 10 may store program instructions for implementing the following functions: Hyperspectral remote sensing technology is used to obtain spectral reflectance data and corresponding zinc content of soil samples in ecologically fragile areas; data enhancement is performed on the spectral reflectance data, and an enhanced spectral data set is obtained by combining the enhancement result and the zinc content; the optimal spectral form is obtained through the enhanced spectral data set; based on the optimal spectral form, characteristic bands of the enhanced spectral data set are optimized, and soil zinc content monitoring is completed according to the priority results.

[0057] In one possible implementation, the memory 10 may include a program storage area and a data storage area, wherein the program storage area may store an operating system, and at least one application required for a function, etc.; the data storage area may store data created during use. In addition, the memory 10 may include a read-only memory and a random access memory, and provide instructions and data to the processor. A portion of the memory may also include NVRAM. The memory stores an operating system and operating instructions, executable modules or data structures, or a subset thereof, or an extended set thereof, wherein the operating instructions may include various operating instructions for implementing various operations. The operating system may include various system programs for implementing various basic tasks and processing hardware-based tasks.

[0058] The processor 20 may be a central processing unit (CPU), an application specific integrated circuit, a digital signal processor, a field programmable gate array or other programmable logic device, and the processor 20 may be a microprocessor or any conventional processor, etc. The processor 20 may call a program stored in the memory 10. The communication interface 31 may be an interface of a communication module, and is used to connect with other devices or systems.

[0059] Of course, it should be noted that Figure 3 The structure shown does not constitute a limitation on the soil zinc content monitoring device based on hyperspectral remote sensing in this embodiment. In actual applications, the soil zinc content monitoring device based on hyperspectral remote sensing may include Figure 3 More or fewer components than shown, or combinations of certain components.

[0060] The embodiment also provides a storage medium having a computer program stored thereon. When the computer program is executed by a processor, the steps of the above-mentioned soil zinc content monitoring method based on hyperspectral remote sensing are implemented.

[0061] The storage medium may include: a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and other media that can store program codes.

[0062] In summary, the present invention is based on the detection data of soil samples in ecologically fragile areas, completes soil zinc content monitoring by processing the data and constructing a classification learning model, solves the problem of rapid and accurate monitoring of soil zinc content in ecologically fragile areas, and improves the applicability of soil zinc content monitoring in ecologically fragile areas.

[0063] Therefore, the present invention effectively overcomes various shortcomings in the prior art and has high industrial utilization value.

[0064] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present invention, and they should all be included in the scope recorded in the present invention.

Claims

1. A method for monitoring soil zinc content based on hyperspectral remote sensing, characterized in that: The method for monitoring soil zinc content based on hyperspectral remote sensing comprises the following steps: Using hyperspectral remote sensing technology, we can obtain spectral reflectance data and corresponding zinc content of soil samples in ecologically fragile areas; Performing data enhancement on the spectral reflectance data, and combining the enhancement result with the zinc content to obtain an enhanced spectral data set; By means of the enhanced spectral data set, an optimal spectral form is obtained; Based on the optimal spectral form, the enhanced spectral data set is optimized for characteristic bands, and soil zinc content monitoring is completed according to the priority results.

2. The method for monitoring soil zinc content based on hyperspectral remote sensing according to claim 1, characterized in that: The step of performing data enhancement on the spectral reflectance data comprises the following steps: Performing a first enhancement on the spectral reflectance data to obtain a first enhancement result; Performing a second enhancement on the spectral reflectance data to obtain a second enhancement result; Perform a third enhancement on the spectral reflectance data to obtain a third enhancement result.

3. The method for monitoring soil zinc content based on hyperspectral remote sensing according to claim 2, characterized in that: The first enhancement is performed on the spectral reflectance data to obtain a first enhancement result, which satisfies the following formula: in, represents the first enhancement result, Indicates band The wavelength of represents spectral reflectance data, Indicates the interval between two adjacent wavelengths.

4. The method for monitoring soil zinc content based on hyperspectral remote sensing according to claim 2, characterized in that: The second enhancement is performed on the spectral reflectance data to obtain a second enhancement result, which satisfies the following formula: in, represents the second enhancement result, Indicates band The wavelength of represents the first enhancement result, Indicates the interval between two adjacent wavelengths.

5. The method for monitoring soil zinc content based on hyperspectral remote sensing according to claim 1, characterized in that: The method of obtaining the optimal spectral form by enhancing the spectral data set comprises the following steps: Splitting the enhanced spectral data set to obtain a spectral data training set and a spectral data verification set; Constructing a first machine classification model, training the first machine classification model using the spectral data training set, and combining the spectral data verification set and the trained first machine classification model to obtain a first verification result; The first verification result is comprehensively analyzed using the determination coefficient, relative root mean square error and relative analytical error to obtain the optimal spectral form.

6. The method for monitoring soil zinc content based on hyperspectral remote sensing according to claim 1, characterized in that: The method of optimizing characteristic bands of the enhanced spectral data set based on the optimal spectral form and completing soil zinc content monitoring according to the priority results includes the following steps: Based on the optimal spectral form, the enhanced spectral dataset is screened to obtain an optimal spectral dataset; Dividing the optimal spectral data set to obtain an optimal spectral training data set and an optimal spectral verification data set; Obtaining the optimal band in the optimal spectrum training data set through a variety of band selection algorithms; Constructing a second machine classification model, using the optimal band to train the second machine classification model, combining the optimal spectrum verification data set and the trained second machine classification model to obtain a second verification result; The second verification result is comprehensively analyzed by using the determination coefficient, relative root mean square error and relative analysis error to obtain an optimal band selection algorithm; Based on the optimal band selection algorithm, soil zinc content monitoring is completed through the trained second machine classification model.

7. The method for monitoring soil zinc content based on hyperspectral remote sensing according to claim 6, characterized in that: The band selection algorithm includes a minimum absolute shrinkage and selection operator algorithm, a genetic algorithm and a gradient boosting decision tree algorithm.

8. The method for monitoring soil zinc content based on hyperspectral remote sensing according to claim 5 or 6, characterized in that: The determination coefficient satisfies the following formula: in, represents the coefficient of determination; The relative root mean square error satisfies the following formula: in, represents the relative root mean square error; The relative analysis error satisfies the following formula: in, represents the relative analytical error, , and The samples are The true value, predicted value and average value of .

9. The method for monitoring soil zinc content based on hyperspectral remote sensing according to claim 5 or 6, characterized in that: The machine classification model is built based on the leave-one-out method.

10. A soil zinc content monitoring system based on hyperspectral remote sensing, characterized in that: The soil zinc content monitoring system based on hyperspectral remote sensing includes: an input device, an output device, a processor, and a memory. The input device, the output device, the processor, and the memory are interconnected. The memory includes program instructions, and the program instructions are used to execute the soil zinc content monitoring method based on hyperspectral remote sensing according to any one of claims 1 to 9.

Images