A Pesticide Residue Detection Method and System Based on Big Data

Abstract

The present invention relates to the technical field of data processing, and particularly to a method and system for detecting pesticide residues based on big data. The method includes: obtaining the surface spectral curve of a target crop and pesticide-related parameters; determining the dynamic adjustment coefficient of the surface spectral curve; for a target band in each band within the surface spectral curve, determining the time decay factor of the target band; determining the spectral sensitivity of the target band; determining the weight of the target band; reducing the dimension of the surface spectral curve, and establishing a pesticide residue detection model to achieve pesticide residue detection based on big data. By extracting the features of each band and calculating the weights, the present invention can ensure that the key information most relevant to pesticide residues is retained during the dimension reduction process, and then the bands are screened according to the weights to remove the noise information irrelevant or less relevant to pesticide residues, which helps to reduce the risks of false alarms and missed detections and improve the accuracy of pesticide residue detection.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a method and system for detecting pesticide residues based on big data. Background Art

[0002] As one of the important challenges in the global food safety field, the research and application of pesticide residue detection technology are of great significance. Among many detection methods, hyperspectral data analysis technology has become the mainstream technical means in modern agricultural detection and analysis due to its advantages of non-destructive, pollution-free and no need for sample pretreatment. The core of this technology is to analyze the spectral characteristics in hyperspectral data to achieve qualitative detection and quantitative evaluation of pesticide residues.

[0003] However, hyperspectral data usually contains hundreds to thousands of bands, which brings significant challenges to data processing. First of all, the hyperspectral data without dimensionality reduction will occupy a large amount of computing resources and storage space, affecting the detection efficiency. Secondly, a large number of features in high-dimensional data may lead to model overfitting, that is, the model overemphasizes the noise or irrelevant information in the data, thus reducing the detection accuracy.

[0004] Therefore, the dimensionality reduction processing of hyperspectral data has become a key link to improve the detection performance. In traditional dimensionality reduction methods, the Principal Component Analysis (PCA) algorithm and the t-Distributed Stochastic Neighbor Embedding (t-SNE) algorithm are more commonly used, but both have certain limitations. The PCA algorithm is based on a linear assumption and may not be able to fully retain the important information in the data when dealing with the hyperspectral data of pesticide residues with nonlinear characteristics; while the t-SNE algorithm can handle nonlinear data well, but the interpretability of its dimensionality reduction results is poor, which is not conducive to subsequent decision-making analysis for pesticide residue detection that requires clear feature information. Summary of the Invention

[0005] In order to solve the problem that when the traditional SMOTE algorithm randomly selects K nearest neighbor points, it will more often select the minority class samples near the majority class samples as neighbors, making the newly generated samples more biased towards the feature space of the majority class samples, thus exacerbating the data imbalance, resulting in a decrease in the recognition ability of the minority class samples when using the dataset to train a classification model subsequently, and further reducing the accuracy of the classification model for pesticide residue detection, the present invention provides a method and system for detecting pesticide residues based on big data.

[0006] In the first aspect, the present invention provides a method for detecting pesticide residues based on big data, adopting the following technical solutions:

[0007] A method for detecting pesticide residues based on big data, comprising: obtaining the surface spectral curve of a target crop among a plurality of crops with known pesticide residue concentrations, as well as the spraying interval, half-life, actual spraying amount, calibrated reference dose, optimal degradation temperature of the pesticide used most recently in each band within the surface spectral curve, and the collection temperature of the target crop; determining the dynamic adjustment coefficient of the surface spectral curve according to the standard deviation and second derivative of each band within the surface spectral curve, as well as the spraying interval and the half-life; for a target band among the bands within the surface spectral curve, determining the time decay factor of the target band according to the spraying interval, half-life, actual spraying amount, and calibrated reference dose of the pesticide used most recently in the target band; determining the spectral sensitivity of the target band according to the standard deviation and second derivative of the target band, the maximum values of the standard deviation and second derivative of each band, the half-life and optimal degradation temperature of the pesticide used most recently in the target band, and the collection temperature of the target crop; determining the weight of the target band according to the dynamic adjustment coefficient, the time decay factor, and the spectral sensitivity of the target band; reducing the dimension of the surface spectral curve according to the magnitude of the weight to obtain the reduced-dimension surface spectral curve, using the target crop with known pesticide residue concentration as the training set of the machine learning algorithm, and establishing a pesticide residue detection model based on the reduced-dimension surface spectral curve to achieve pesticide residue detection based on big data.

[0008] By combining multiple factors such as the standard deviation of the surface spectral curve, the second derivative, the spraying interval, the half-life, and the actual spraying amount, the present invention can accurately determine the dynamic adjustment coefficient, the time decay factor, and the spectral sensitivity, thereby more precisely simulating the time decay process of pesticide residues and better capturing the changes in the surface spectra of different crops. The traditional SMOTE algorithm tends to make the generated samples deviate towards the feature space of the majority class. However, the present invention avoids data bias caused by simple oversampling or undersampling by comprehensively considering environmental factors such as the interval between pesticide sprays and the degradation temperature. Especially in the detection of crops with pesticide residues, it can better capture the features of the minority class (i.e., samples with low or no pesticide residues). By calculating the weight of each band, the highlighting of important information is achieved, enhancing the adaptability of the algorithm. The weight of the band is jointly determined by multiple factors, enabling the key features to be retained even after the spectral curve is dimensionally reduced and avoiding excessive loss of information. During the pesticide residue detection process, the collection temperature, degradation temperature, and spraying history of the crops may all have different effects on the pesticide residue amount. By precisely calculating the spectral sensitivity and comprehensively considering the above factors, it helps to conduct effective residue detection under different environmental conditions. Due to considering dynamic factors and environmental impacts, the present invention can more accurately identify the pesticide residue concentration in the target crops compared with traditional methods, which helps to improve the detection accuracy of the pesticide residue detection system and further enhance the monitoring and management capabilities of the pesticide residue problem. Key features are retained during the dimensional reduction process, and the trained pesticide residue detection model has better generalization ability and can accurately identify the pesticide residue concentration of different types of crops in practical applications without being easily affected by overfitting.

[0009] Further, the dynamic adjustment coefficient satisfies: ; where is the dynamic adjustment coefficient of the surface spectral curve of the target crop, is the number of bands within the surface spectral curve, is the th standard deviation of the bands within the surface spectral curve, is the th second derivative of the bands within the surface spectral curve, and are respectively the spraying interval and half-life of the pesticide used most recently for the th band within the surface spectral curve, is the mean function, is the natural exponential function.

[0010] By combining the standard deviation, the second derivative, the spraying interval, and the half-life, the dynamic adjustment coefficient can sensitively capture the changes related to pesticide residues in the surface spectral curve, thereby improving the detection accuracy of pesticide residues; the exponential function takes into account the non-linear process of pesticide decay, enabling the dynamic adjustment coefficient to automatically adjust the weights of the bands according to time changes, thus better adapting to different time and environmental conditions; the mean function averages the adjustment coefficients of multiple bands, helping to avoid the influence of abnormal data in a single band on the model, reducing overfitting, and enhancing the robustness and stability of the detection system; by comprehensively utilizing the information of different bands, the sensitivity of the model to pesticide residue detection can be enhanced, thereby providing more accurate results in the detection of pesticide residues in different agricultural environments and different types of pesticides.

[0011] Further, the time decay factor satisfies: ; where is the time decay factor of the th band within the surface spectral curve of the target crop, and are respectively the spraying interval and the half-life of the pesticide used most recently for the th band within the surface spectral curve, and are respectively the actual spraying amount and the calibrated reference dose of the pesticide used most recently for the th band within the surface spectral curve, is the natural exponential function.

[0012] The time decay factor of the present invention combines the spraying interval, the half-life, the actual spraying amount, and the calibrated reference dose, and can more accurately reflect the decay process of pesticides in the surface spectrum of the target crop. The exponential decay models the time decay process of pesticide residues, which helps to accurately capture the decay law of pesticide residues over time; by introducing the actual spraying amount and the calibrated reference dose, it can adapt to changes in different pesticide usage amounts, and further more accurately adjust the calculation of the decay factor to ensure that the model can make a more appropriate response to the actual spraying situation; by comprehensively considering the changes in time, spraying amount, and reference dose, the time decay factor can provide a more accurate decay model, thereby improving the accuracy and stability of pesticide residue detection and avoiding possible errors or instability phenomena under different conditions.

[0013] Further, the spectral sensitivity satisfies: ; where is the spectral sensitivity of the th band within the surface spectral curve of the target crop, is the standard deviation of the th band within the surface spectral curve, is the second derivative of the th band within the surface spectral curve, is the maximum value among the standard deviations of each band within the surface spectral curve, is the maximum value among the second derivatives of each band within the surface spectral curve, is the acquisition temperature of the target crop, and are respectively the half-life and the optimal degradation temperature of the pesticide used most recently in the th band within the surface spectral curve, is the absolute value symbol.

[0014] By considering the standard deviation and second derivative of different bands, as well as the effects of temperature and pesticide degradation, the present invention can more accurately describe and predict the spectral sensitivity of different bands, making pesticide detection more precise; the use of pesticides often affects the spectral characteristics of crops, and temperature changes may accelerate or slow down the pesticide degradation process. By introducing the parameters of pesticide half-life and optimal degradation temperature, the effects of these factors on spectral sensitivity can be quantified, thus more accurately reflecting pesticide residues and crop health.

[0015] Further, the weights satisfy: ; where is the weight of the th band within the surface spectral curve of the target crop, is the dynamic adjustment coefficient of the surface spectral curve of the target crop, is the time decay factor of the th band within the surface spectral curve of the target crop, is the spectral sensitivity of the th band within the surface spectral curve of the target crop.

[0016] By combining spectral sensitivity and time decay factor, the present invention can more accurately reflect the variation characteristics of each band during the crop growth cycle, thereby improving the reliability of the analysis results; introducing a dynamic adjustment coefficient enables the system to automatically adjust the weight of each band according to different time points or environmental conditions, adapting to the spectral characteristics of crops at different growth stages and increasing flexibility; the introduction of the time decay factor helps to eliminate the influence brought by external environmental changes (such as climate factors, observation conditions, etc.), making the surface spectral data of crops more stable and accurate.

[0017] Further, performing dimensionality reduction on the surface spectral curve according to the magnitudes of the weights of each band to obtain a dimensionality-reduced surface spectral curve includes: in response to the weight being not higher than a preset dimensionality reduction threshold, removing the target band from the surface spectral curve to obtain the dimensionality-reduced surface spectral curve.

[0018] Further, the machine learning algorithm uses a support vector machine.

[0019] Further, the machine learning algorithm uses a BP neural network.

[0020] Further, to achieve pesticide residue detection based on big data, it includes: for the crop to be detected, obtaining the surface spectral curve after dimensionality reduction of the crop to be detected, inputting the surface spectral curve after dimensionality reduction into the pesticide residue detection model, obtaining the pesticide residue concentration of the crop to be detected, and completing the pesticide residue detection based on big data.

[0021] In a second aspect, the present invention provides a pesticide residue detection system based on big data, adopting the following technical solution:

[0022] A pesticide residue detection system based on big data includes: a processor and a memory, and the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned pesticide residue detection method based on big data is implemented.

[0023] By adopting the above technical solution, the above-mentioned pesticide residue detection method based on big data is generated into a computer program and stored in the memory to be loaded and executed by the processor, so as to manufacture a terminal device according to the memory and the processor, which is convenient to use.

[0024] The present invention has the following technical effects:

[0025] Different from the drawback that the traditional SMOTE algorithm randomly selects K-nearest neighbor points, which easily leads to the new samples being biased towards the feature space of the majority-class samples and exacerbates data imbalance, the present invention processes data by determining dynamic adjustment coefficients, time decay factors, spectral sensitivities, etc., avoiding the problem of exacerbating data imbalance due to unreasonable sample selection, and helping the subsequent classification model to better identify minority-class samples; since the exacerbation of the data imbalance problem is effectively avoided, the training data is more reasonable, thereby improving the identification ability of the subsequent classification model trained using this dataset for minority-class samples, overcoming the defect of the traditional SMOTE algorithm that reduces the identification ability of the classification model for minority-class samples; comprehensively considering multiple relevant factors of the surface spectral curve (such as standard deviation, second derivative, spraying interval, half-life, actual spraying amount, calibration reference dose, optimal degradation temperature, acquisition temperature, etc.) to determine the weights of each band, and establishing a pesticide residue detection model after dimensionality reduction of the surface spectral curve, which considers more factors and processes data more scientifically than traditional methods, can improve the accuracy of the classification model for pesticide residue detection, making the detection results more reliable; analyzing and processing data such as the surface spectral curve of the target crop from multiple dimensions, and establishing a pesticide residue detection model by determining various factors and weights, can make full use of the advantages of big data, making the model more scientific and reliable, and meeting the requirements of modern pesticide residue detection for accuracy and science. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 is the flowchart of the method in an embodiment of the present invention for a pesticide residue detection method based on big data. DETAILED DESCRIPTION OF THE INVENTION

[0027] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0028] An embodiment of the present invention discloses a pesticide residue detection method based on big data, referring to Figure 1 , including steps S1 - S6:

[0029] S1: Obtain the surface spectral curve of the target crop and pesticide-related parameters.

[0030] It should be noted that representative detection areas are selected on target crops (such as corn, wheat, etc.) to ensure that these areas can reflect the overall pesticide residue situation. A suitable hyperspectral data device is selected to ensure that its spectral range and resolution meet the detection requirements. According to the characteristics of the target crop and the detection requirements, the parameters of the hyperspectral data system are set, such as the exposure time (e.g., 0.05 s), scanning speed (e.g., 0.6 cm / s), etc. The hyperspectral data system is used to obtain the surface spectral curve of the detection area of the target crop.

[0031] Obtain the surface spectral curve of the target crop in multiple crops with known pesticide residue concentrations, as well as the spraying interval, half-life, actual spraying amount, calibration reference dose, optimal degradation temperature of the pesticide used on the target crop last time, and the collection temperature of the target crop.

[0032] S2: Determine the dynamic adjustment coefficient of the surface spectral curve.

[0033] It should be noted that the dynamic adjustment coefficient of the surface spectral curve is used to comprehensively consider the influence of the spectral curve's own characteristics and the pesticide spraying time factor on the spectrum; the standard deviation reflects the degree of dispersion of the spectral data in each band, and the second derivative reflects the change rate of the spectral curve in this band. These two parameters describe the characteristics of different bands from the perspective of the spectral curve itself, while the spraying interval and half-life (for example, 7 for organophosphorus) are considered from the perspective of the degradation of pesticides on the crop surface. As time goes by, the residue amount of pesticides will change, thus affecting the spectral curve; therefore, in this step, the dynamic adjustment coefficient is calculated based on the above parameters.

[0034] Determine the dynamic adjustment coefficient of the surface spectral curve according to the standard deviation and second derivative of each band in the surface spectral curve, as well as the spraying interval and the half-life.

[0035] Specifically, the dynamic adjustment coefficient satisfies:

[0036] ;

[0037] In the formula, is the dynamic adjustment coefficient of the surface spectral curve of the target crop, is the number of bands in the surface spectral curve, is the standard deviation of the th band in the surface spectral curve, is the second derivative of the th band in the surface spectral curve, and are respectively the spraying interval and half-life of the pesticide used last time for the th band in the surface spectral curve, is the mean function, is the natural exponential function.

[0038] Among them, if the standard deviation and second derivative of a certain band are large, it indicates that the spectral data of this band has large fluctuations and a high rate of change. The value of will be relatively large, making this band account for a relatively large proportion when calculating the dynamic adjustment coefficient, which means that this band has a greater impact on the overall characteristics of the spectral curve. The larger, the smaller the value of, indicating that the pesticide has degraded for a long time and the residue amount has decreased. At this time, the value of will increase, then increases, which indicates that as the pesticide residue amount decreases, the influence of the self-characteristics of the spectral curve is relatively enhanced.

[0039] S3: For the target band in each band of the surface spectral curve, determine the time decay factor of the target band.

[0040] It should be noted that the pesticide residue amount decays exponentially with time (for example, organophosphorus is 7), and temperature changes will accelerate or delay the degradation process. Quantify the influence of the environment on the residue amount through the exponential decay of the pesticide residue amount with time and temperature changes to avoid the detection value being falsely high due to time lag; at the same time, the difference between the actual spraying dose and the standard dose directly affects the residue concentration. By adjusting the time decay factor with the ratio of the two, the error caused by uneven spraying operations can be reduced.

[0041] Determine the time decay factor of the target band according to the spraying interval, half-life, actual spraying amount of the pesticide used in the target band last time, and the calibrated reference dose.

[0042] Specifically, the time decay factor satisfies:

[0043] ;

[0044] In the formula, is the time decay factor of the th band in the surface spectral curve of the target crop. and are respectively the spraying interval and half-life of the pesticide used in the th band in the surface spectral curve last time. and are respectively the actual spraying amount and calibrated reference dose of the pesticide used in the th band in the surface spectral curve last time. is the natural exponential function.

[0045] Among them, the larger, ​​​The smaller the value, the more it indicates that the pesticide residue is continuously decreasing over time, and the time decay factor will also decrease accordingly; if the actual spraying amount is greater than the calibrated reference dose , then is greater than 1, which will cause the time decay factor to increase. On the contrary, if the actual spraying amount is less than the calibrated reference dose , then the time decay factor will decrease, indicating that the amount of actual pesticide used will affect the decay of pesticide residues.

[0046] S4: Determine the spectral sensitivity of the target band.

[0047] It should be noted that the specific absorption peak of pesticide residues (such as the characteristic response of organophosphorus pesticides in the near-infrared band) may be masked by environmental noise. Through spectral second derivative calculation, weak absorption features can be amplified to improve detection sensitivity; while different types of pesticides (such as carbamates and pyrethroids) have differences in sensitive bands in the spectrum. By screening high-discrimination bands through standard deviation, redundant data interference can be reduced; therefore, the spectral sensitivity of the band is calculated based on the standard deviation and second derivative of the band here.

[0048] Determine the spectral sensitivity of the target band according to the standard deviation and second derivative of the target band, the maximum values of the standard deviation and second derivative in each band, the half-life and optimal degradation temperature of the pesticide used in the target band last time, and the collection temperature of the target crop.

[0049] Specifically, the spectral sensitivity satisfies:

[0050] ;

[0051] In the formula, is the spectral sensitivity of the th band in the surface spectral curve of the target crop, is the standard deviation of the th band in the surface spectral curve, is the second derivative of the th band in the surface spectral curve, is the maximum value of the standard deviation of each band in the surface spectral curve, is the maximum value of the second derivative of each band in the surface spectral curve, is the collection temperature of the target crop, and are the half-life and optimal degradation temperature of the pesticide used in the th band in the surface spectral curve last time, is the absolute value symbol.

[0052] Among them, The larger it is, the higher the discreteness of the data in this band, and it may contain stronger pesticide characteristic signals. That is, the band with a larger standard deviation is more sensitive to the residue amount. The larger it is, the more obvious the characteristic peak of this band, and the higher the sensitivity to the pesticide residue amount. When it is larger, it indicates that the pesticide degradation rate changes and is more strongly affected by the environment. It is necessary to dynamically compensate for the residue error through the proportional term.

[0053] S5: Determine the weight of the target band.

[0054] It should be noted that the weight of the target band is obtained by comprehensively considering the dynamic adjustment coefficient, the time decay factor, and the spectral sensitivity; the dynamic adjustment coefficient reflects the overall dynamic characteristics of the surface spectral curve, the time decay factor reflects the change in the residue amount of the pesticide over time in the target band, and the spectral sensitivity describes the sensitivity of the target band to the pesticide residue and environmental factors. By combining these three parameters through weighting, the obtained weight can more comprehensively evaluate the importance of the target band in pesticide residue detection.

[0055] Determine the weight of the target band according to the dynamic adjustment coefficient, the time decay factor, and the spectral sensitivity of the target band.

[0056] Specifically, the weight satisfies:

[0057] ;

[0058] In the formula, is the weight of the th band in the surface spectral curve of the target crop, is the dynamic adjustment coefficient of the surface spectral curve of the target crop, is the time decay factor of the th band in the surface spectral curve of the target crop, is the spectral sensitivity of the th band in the surface spectral curve of the target crop.

[0059] Among them, when the spectral sensitivity of the band is relatively high, tends to 1, and at this time, the spectral sensitivity is emphasized to obtain the weight; when the time decay factor is relatively large, tends to 0, and at this time, the time decay factor is emphasized to obtain the weight.

[0060] S6: Reduce the dimension of the surface spectral curve and establish a pesticide residue detection model to achieve pesticide residue detection based on big data.

[0061] According to the magnitudes of the weights, the surface spectral curve is dimensionally reduced to obtain the dimensionally reduced surface spectral curve. The target crops with known pesticide residue concentrations are used as the training set of the machine learning algorithm, and a pesticide residue detection model is established based on the dimensionally reduced surface spectral curve to achieve pesticide residue detection based on big data.

[0062] Specifically, the dimensionally reducing the surface spectral curve according to the magnitudes of the weights of each band to obtain the dimensionally reduced surface spectral curve includes:

[0063] In response to the weight not being higher than a preset dimensional reduction threshold, the target band is removed from the surface spectral curve to obtain the dimensionally reduced surface spectral curve.

[0064] Implementers can set the dimensional reduction threshold according to the specific implementation situation. For example, 0.25.

[0065] Specifically, the machine learning algorithm uses a support vector machine.

[0066] In another embodiment, the machine learning algorithm uses a BP neural network.

[0067] Specifically, the achieving of pesticide residue detection based on big data includes:

[0068] For the crop to be detected (the crop to be detected is the same as the target crop), the dimensionally reduced surface spectral curve of the crop to be detected is obtained, and the dimensionally reduced surface spectral curve is input into the pesticide residue detection model to obtain the pesticide residue concentration of the crop to be detected, thus completing the pesticide residue detection based on big data.

[0069] An embodiment of the present invention also discloses a pesticide residue detection system based on big data, including a processor and a memory. The memory stores computer program instructions, and when the computer program instructions are executed by the processor, a pesticide residue detection method based on the present invention is implemented.

[0070] The above system also includes other components well-known to those skilled in the art such as a communication bus and a communication interface. Their settings and functions are known in the art, so they will not be elaborated here.

[0071] The above are all the preferred embodiments of the present invention. The protection scope of the present invention is not limited by this. Therefore, all equivalent changes made according to the structure, shape, and principle of the present invention should be covered within the protection scope of the present invention.

Claims

1. A pesticide residue detection method based on big data, characterized in that: include: Obtaining the surface spectral curve of the target crop among multiple crops with known pesticide residue concentrations, as well as the spraying interval, half-life, actual spraying amount, calibrated reference dose, optimal degradation temperature of the pesticide most recently used in each band within the surface spectral curve, and the collection temperature of the target crop; Determining a dynamic adjustment coefficient of the surface spectrum curve according to the standard deviation and the second-order derivative of each band in the surface spectrum curve, as well as the spraying interval and the half-life; For the target band in each band in the surface spectral curve, the time attenuation factor of the target band is determined according to the spraying interval, half-life, actual spraying amount and calibration reference dose of the pesticide used in the target band most recently; the spectral sensitivity of the target band is determined according to the standard deviation and second-order derivative of the target band, the maximum value of the standard deviation and the maximum value of the second-order derivative of each band, the half-life and optimal degradation temperature of the pesticide used in the target band most recently, and the collection temperature of the target crop; the weight of the target band is determined according to the dynamic adjustment coefficient, the time attenuation factor and the spectral sensitivity of the target band; The weights satisfy: ； In the formula, The first The weight of the band, is the dynamic adjustment coefficient of the surface spectral curve of the target crop, The first The time attenuation factor of each band, The first Spectral sensitivity of each band; According to the size of the weight, the surface spectral curve is reduced in dimension to obtain the reduced-dimensional surface spectral curve. The target crops with known pesticide residue concentrations are used as the training set of the machine learning algorithm. A pesticide residue detection model is established based on the reduced-dimensional surface spectral curve to realize pesticide residue detection based on big data.

2. The pesticide residue detection method based on big data according to claim 1, characterized in that: The dynamic adjustment coefficient satisfies: ； In the formula, is the dynamic adjustment coefficient of the surface spectral curve of the target crop, is the number of bands in the surface spectral curve, The surface spectrum curve The standard deviation of the band, The surface spectrum curve The second-order derivative of the band, and They are the surface spectral curves. The spraying interval and half-life of the most recent pesticide used in each band, is the mean function, is a natural exponential function.

3. The pesticide residue detection method based on big data according to claim 1, characterized in that: The time decay factor satisfies: ； In the formula, The first The time attenuation factor of each band, and They are the surface spectral curves. The spraying interval and half-life of the most recent pesticide used in each band, and They are the surface spectral curves. The actual spraying amount of pesticides used in the last time in each band and the calibrated reference dose, is a natural exponential function.

4. The pesticide residue detection method based on big data according to claim 1, characterized in that: The spectral sensitivity satisfies: ； In the formula, The first The spectral sensitivity of each band, The surface spectrum curve The standard deviation of the band, The surface spectrum curve The second-order derivative of the band, is the maximum value of the standard deviation of each band in the surface spectral curve, is the maximum value of the second-order derivative of each band in the surface spectrum curve, is the collection temperature of the target crop, and They are the surface spectral curves. The half-life and optimal degradation temperature of the pesticide most recently used in each band, is the absolute value symbol.

5. The pesticide residue detection method based on big data according to claim 1, characterized in that: The surface spectral curve is reduced in dimension according to the weight of each band to obtain the surface spectral curve after dimension reduction, including: In response to the weight being no higher than a preset dimensionality reduction threshold, the target band is removed from the surface spectral curve to obtain a surface spectral curve after dimensionality reduction.

6. The pesticide residue detection method based on big data according to claim 1, characterized in that: The machine learning algorithm adopts support vector machine.

7. The pesticide residue detection method based on big data according to claim 1, characterized in that: The machine learning algorithm adopts BP neural network.

8. The pesticide residue detection method based on big data according to claim 1, characterized in that: The method for realizing pesticide residue detection based on big data includes: For the crops to be tested, the surface spectral curves after dimensionality reduction of the crops to be tested are obtained, and the surface spectral curves after dimensionality reduction are input into the pesticide residue detection model to obtain the pesticide residue concentration of the crops to be tested, thereby completing the pesticide residue detection based on big data.

9. A pesticide residue detection system based on big data, characterized in that: include: A processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a pesticide residue detection method based on big data according to any one of claims 1 to 8 is implemented.

Images