Liquorice growth environment monitoring method and system

Abstract

The invention provides a licorice growth environment monitoring method and system, and relates to the technical field of environment monitoring. The method comprises the following steps: obtaining a plurality of licorice samples; acquiring a plurality of high-definition sample licorice images and growth height data of the sample licorice; inputting the high-definition sample licorice images into a trained prediction insect damage probability model, and obtaining prediction insect damage probability data of multiple sample licorice at multiple moments of the current monitoring period; determining a growth abnormality score according to the predicted insect damage probability data and the growth height data; acquiring climate parameters; determining a climate anomaly score according to the climate parameters; obtaining nutrient element content information and moisture content information in the soil sample; determining a soil abnormity score according to the nutrient element content information and the moisture content information; and generating a licorice growth environment monitoring report. According to the invention, the probability that licorice is attacked by pests can be predicted, and the comprehensiveness of monitoring the growth environment of licorice is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of environmental monitoring, and in particular to a method and system for monitoring the growth environment of liquorice. Background Art

[0002] In the related art, CN115266610A provides a plant growth monitoring method and system. The method is applied to a plant growth monitoring system, and the system includes: a spectral data monitoring module, a growth environment data monitoring module, a main control module, and an Internet of Things communication component module arranged in the plant growth area. The method includes: obtaining the plant reflection spectrum data of sunlight and the preset band of the plant through the spectral data monitoring module; obtaining the growth environment parameter data of the plant through the growth environment data monitoring module; obtaining the location information of the plant through the main control module, storing the spectral data, the growth environment parameter data and the location information, and calculating the vegetation index based on the spectral data and the growth environment parameter data; sending the growth environment parameter data, the vegetation index and the location information to the Internet of Things cloud through the Internet of Things communication component module.

[0003] CN107121535A provides an effective crop growth environment monitoring system, including an information acquisition module, a controller module, an information transmission module and a remote monitoring module, wherein the information acquisition module is used to collect data information and video information of the crop growth environment, and the information transmission module transmits the collected data information and video information of the crop growth environment to the remote monitoring module by wireless transmission, and the remote monitoring module judges the condition of the crop growth environment according to the received data, and generates corresponding operation instructions according to the judgment result, and transmits the operation instructions to the controller module through the information transmission module, and the controller module performs corresponding operations according to the received instructions. The beneficial effects of this scheme are: effective monitoring of environmental factors affecting the growth of crops is achieved, corresponding measures are taken according to the monitoring situation, and the health of the crop growth environment is ensured.

[0004] Therefore, although the related art can monitor the growth environment, the related art does not take into account the probability of being infested by insects during the growth process, that is, it is difficult to predict the probability of licorice being infested by insects, and thus it is difficult to comprehensively monitor the growth environment of licorice.

[0005] The information disclosed in the background technology section of this application is only intended to deepen the understanding of the general background technology of this application, and should not be regarded as an admission or any form of suggestion that the information constitutes the prior art already known to those skilled in the art. Summary of the invention

[0006] The present invention provides a method and system for monitoring the growth environment of licorice, which can solve the technical problems that the related technology is difficult to predict the probability of licorice being damaged by insect pests and is difficult to comprehensively monitor the growth environment of licorice.

[0007] According to a first aspect of the present invention, a method for monitoring a licorice growth environment is provided, comprising: sampling licorice in a licorice growth area to obtain a plurality of sample licorice; taking high-definition photographs of the sample licorice at multiple times in a current monitoring cycle to obtain a plurality of high-definition sample licorice images and growth height data of the sample licorice; inputting the high-definition sample licorice images into a trained predicted pest probability model to obtain predicted pest probability data of a plurality of sample licorice at multiple times in the current monitoring cycle; determining a growth anomaly score according to the predicted pest probability data and the growth height data; obtaining climate parameters of the licorice growth area at multiple times in the current monitoring cycle, wherein the climate parameters include temperature parameters, humidity parameters and light intensity parameters; determining a climate anomaly score according to the climate parameters; collecting soil samples in the licorice growth area at multiple times in the current monitoring cycle and performing spectral component analysis to obtain nutrient element content information and moisture content information in the soil samples; determining a soil anomaly score according to the nutrient element content information and the moisture content information; and generating a licorice growth environment monitoring report according to the growth anomaly score, the climate anomaly score and the soil anomaly score.

[0008] Furthermore, the training step of the predicted pest probability model includes: obtaining historical high-definition sample licorice images at multiple moments in multiple historical monitoring cycles, wherein the previous historical monitoring cycle of the current monitoring cycle is the first historical monitoring cycle; inputting the historical high-definition sample licorice images into the trained predicted pest probability model to obtain historical predicted pest probability data of multiple sample licorice at multiple moments in multiple historical monitoring cycles; performing feature extraction processing on the licorice rhizome area in the historical high-definition sample licorice images through the rhizome feature coding level of the trained image recognition neural network model to obtain historical rhizome feature vectors of each historical sample licorice; performing feature extraction processing on the licorice leaf area in the historical high-definition sample licorice images through the leaf feature coding level of the trained image recognition neural network model to obtain historical leaf feature vectors of each historical sample licorice; performing feature extraction processing on the licorice fruit area in the historical high-definition sample licorice images through the fruit feature coding level of the trained image recognition neural network model to obtain historical fruit feature vectors of each historical sample licorice; obtaining historical pest probability data of multiple sample licorice at multiple moments in multiple historical monitoring cycles; according to the formula

[0009] Determine the training loss function Loss of the pest prediction probability model, where R i，s，kis the historical root feature vector of the i-th sample of licorice at the k-th moment in the s-th historical monitoring period, The nth historical monitoring period k The historical root feature vector of the i-th sample licorice at the moment, For R i，s，k The transposed vector, X i，s，k is the historical leaf feature vector of the i-th sample licorice at the k-th moment in the s-th historical monitoring period, The nth historical monitoring period k The historical leaf feature vector of the i-th sample licorice at the moment, For X i，s，k The transposed vector, Y i，s，k is the historical fruit feature vector of the i-th sample licorice at the k-th moment in the s-th historical monitoring period, The nth historical monitoring period k The historical fruit feature vector of the i-th sample licorice at the moment, Y i，s，k The transposed vector, G i，s，k is the historical pest probability data of the i-th sample licorice at the k-th moment in the s-th historical monitoring period, G i，s，p is the historical predicted pest probability data of the i-th sample licorice at the k-th moment in the s-th historical monitoring period, n i is the number of sample licorice, n s is the number of historical monitoring cycles, n k is the number of monitoring cycle times, i≤n i , s≤n s , k≤n k , and i, s, k, n i 、n s and n k are all positive integers; according to the training loss function, the predicted pest probability model is trained to obtain the trained predicted pest probability model.

[0010] Furthermore, a growth abnormality score is determined based on the predicted pest probability data and the growth height data, including: fitting the growth height data and multiple moments of the current monitoring period to obtain a growth height function; determining a growth height derivative function based on the growth height function; determining a growth height change rate based on the growth height derivative function; and determining a growth abnormality score based on the predicted pest probability data and the growth height change rate.

[0011] Further, determining a growth abnormality score based on the predicted pest probability data and the growth height change rate includes: according to the formula and Determine the growth dysregulation score A1 , where H′ i，k H′ is the growth height change rate of the i-th sample licorice at the k-th moment in the current monitoring period, p is the growth height change rate threshold, F 1 is a piecewise function that compares the growth height change rate of the i-th sample licorice at the k-th moment in the current monitoring period with the growth height change rate threshold, G i，k，p is the predicted pest probability data of the i-th sample licorice at the k-th moment in the current monitoring period, α 1 and α 2 is the preset weight, and α 1 +α 2 =1,n i is the number of sample licorice, n k is the number of monitoring cycle times, i≤n i , k≤n k , and i, k, n i and n k All are positive integers.

[0012] Furthermore, according to the climate parameters, a climate anomaly score is determined, including: obtaining suitable climate parameters for the licorice growing area, wherein the suitable climate parameters include suitable temperature parameters, suitable humidity parameters and suitable light intensity parameters; obtaining historical climate parameters for the licorice growing area at multiple moments in the last historical monitoring cycle, wherein the historical climate parameters include historical temperature parameters, historical humidity parameters and historical light intensity parameters; determining a climate anomaly score according to the suitable climate parameters, the historical climate parameters and the climate parameters.

[0013] Further, determining a climate anomaly score according to the suitable climate parameter, the historical climate parameter and the climate parameter includes: according to the formula Determine the climate anomaly score A 2 , where t k is the temperature parameter at the kth moment of the current monitoring period, t p is the suitable temperature parameter, w k is the humidity parameter at the kth moment of the current monitoring period, w p is the suitable humidity parameter, m k is the light intensity parameter at the kth moment of the current monitoring period, m p is the appropriate light intensity parameter, t k，h is the historical temperature parameter at the kth moment of the last historical monitoring period, w k，h is the historical humidity parameter at the kth moment of the last historical monitoring period, m k，h is the historical light intensity parameter at the kth moment of the previous historical monitoring period, β 1 and β 2is the preset weight, and β 1 +β 2 =1,n k is the number of monitoring cycle times, k≤n k , and k and n k All are positive integers, and max is the maximum value function.

[0014] Furthermore, determining a soil anomaly score according to the nutrient content information and the moisture content information includes: obtaining a plurality of standard nutrient content information and standard moisture content information in licorice soil; determining a standard soil state vector of licorice soil according to the standard nutrient content information and the standard moisture content information; determining a real-time soil state vector of licorice soil according to the nutrient content information and the moisture content information; and determining a soil anomaly score according to the standard soil state vector and the real-time soil state vector.

[0015] Further, determining a soil anomaly score according to the standard soil state vector and the real-time soil state vector includes: according to the formula Determine the soil anomaly score A 3 , where C 1，k , C 2，k ,…,C N，k is the content information of the 1st, 2nd, ..., Nth nutrient element at the kth moment in the current monitoring period, C 1，T , C 2，T ,…,C N,T is the content information of the 1st, 2nd, ..., Nth standard nutrient element, S k is the moisture content information at the kth moment of the current monitoring period, S T is the standard moisture content information, γ 1 , γ 2 , …, γ N and γ N+1 is the preset weight, is the real-time soil state vector at the kth moment in the current monitoring period, is the standard soil state vector, N is the number of nutrient elements, n k is the number of monitoring cycle times, k≤n k , and k, N and n k All are positive integers.

[0016] Furthermore, a licorice growth environment monitoring report is generated according to the growth anomaly score, the climate anomaly score and the soil anomaly score, including: when the growth anomaly score is greater than the growth anomaly score threshold, a growth monitoring report is generated; when the climate anomaly score is greater than the climate anomaly score threshold, a climate monitoring report is generated; when the soil anomaly score is greater than the soil anomaly score threshold, a soil monitoring report is generated. According to a second aspect of the present invention, a licorice growth environment monitoring system is provided, including: a sample licorice module, for sampling licorice in a licorice growing area to obtain a plurality of sample licorice; a high-definition sample licorice image and growth height data module, for taking high-definition photos of the sample licorice at multiple moments in the current monitoring cycle to obtain a plurality of high-definition sample licorice images and growth height data of the sample licorice; a predicted pest probability model module, for inputting the high-definition sample licorice image into a trained predicted pest probability model to obtain predicted pest probability data of a plurality of sample licorice at multiple moments in the current monitoring cycle; a growth anomaly score module, for determining a growth anomaly score based on the predicted pest probability data and the growth height data; and a climate parameter prediction model module. A data module is used to obtain the climate parameters of the licorice growing area at multiple moments in the current monitoring period, wherein the climate parameters include temperature parameters, humidity parameters and light intensity parameters; a climate anomaly scoring module is used to determine the climate anomaly score according to the climate parameters; a soil sample module is used to collect soil samples in the licorice growing area at multiple moments in the current monitoring period and perform spectral component analysis to obtain nutrient element content information and water content information in the soil samples; a soil anomaly scoring module is used to determine the soil anomaly score according to the nutrient element content information and the water content information; a licorice growth environment monitoring report module is used to generate a licorice growth environment monitoring report according to the growth anomaly score, the climate anomaly score and the soil anomaly score.

[0017] Technical effect: According to the present invention, by predicting the pest probability model, the probability of licorice being infested with pests can be predicted, thereby reducing the impact of pests on licorice growth. By performing spectral component analysis on soil samples to obtain nutrient element content information and water content information, the impact of soil conditions on licorice growth can be identified, and targeted soil improvement can be carried out. Through the comprehensive evaluation of growth anomaly scores, climate anomaly scores and soil anomaly scores, the advantages and disadvantages of the licorice growth environment can be reflected from multiple dimensions, and the comprehensiveness of monitoring the licorice growth environment can be improved. When determining the training loss function of the pest probability prediction model, the influence of the historical root and stem feature vectors, the historical leaf feature vectors and the historical fruit feature vectors on the probability of licorice being infested can be used to determine the influence of the above data on the error of the predicted pest probability, so as to train the loss function based on the influence and the relative error of the historical predicted pest probability, and based on the characteristics that the shorter the time interval, the shorter the cycle interval, and the higher the prediction accuracy, the weights are set to improve the objectivity and accuracy of the loss function, so as to reduce the training loss function during the training process, improve the training intensity and training efficiency, and improve the accuracy of the pest probability prediction model. When determining the growth anomaly score, the growth anomaly score can be determined by the average predicted pest probability of licorice and the average abnormal degree of growth height. The growth state of licorice can be more comprehensively evaluated by the two aspects of pest probability and growth height change rate, which is helpful to timely discover growth anomalies and improve the comprehensiveness and accuracy of growth anomaly score. When determining the climate anomaly score, the climate anomaly score can be determined by suitable climate parameters, historical climate parameters and current climate parameters, so as to effectively evaluate the climate anomaly in the licorice growing area. By comparing the three climate parameters of temperature, humidity and light intensity, the deviation between the current climate parameters and the suitable climate parameters is comprehensively evaluated, which is helpful to fully understand the impact of climate on licorice growth. The current climate parameters can be compared with the historical climate parameters, which is helpful to identify the trend of climate change and improve the comprehensiveness, scientificity and objectivity of the climate anomaly score. When determining the soil anomaly score, the soil anomaly score can be determined based on the cosine similarity of the weighted real-time soil state vector and the weighted standard soil state vector, and the weights are set based on the relative importance of different nutrients and water to licorice growth, so that the soil state assessment is more accurate and helpful to monitor and improve the soil environment.

[0018] It should be understood that the above general description and the following detailed description are exemplary and explanatory only and do not limit the present invention. Other features and aspects of the present invention will become more apparent from the following detailed description of exemplary embodiments with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other embodiments can be obtained based on these drawings without creative work.

[0020] Figure 1 A schematic diagram of a process of monitoring a licorice growth environment according to an embodiment of the present invention is exemplarily shown;

[0021] Figure 2 A block diagram of a licorice growth environment monitoring system according to an embodiment of the present invention is exemplarily shown. DETAILED DESCRIPTION

[0022] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0023] The technical solution of the present invention is described in detail with specific embodiments below. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments.

[0024] Figure 1The flow chart of the method for monitoring the licorice growth environment according to an embodiment of the present invention is exemplarily shown, and the method comprises: step S101, sampling licorice in a licorice growth area to obtain a plurality of sample licorice; step S102, taking high-definition photos of the sample licorice at multiple moments in the current monitoring cycle to obtain a plurality of high-definition sample licorice images and growth height data of the sample licorice; step S103, inputting the high-definition sample licorice images into a trained predicted pest probability model to obtain predicted pest probability data of a plurality of sample licorice at multiple moments in the current monitoring cycle; step S104, determining a growth abnormality score according to the predicted pest probability data and the growth height data; step S105, Step S105, obtaining the climate parameters of the licorice growing area at multiple moments in the current monitoring period, wherein the climate parameters include temperature parameters, humidity parameters and light intensity parameters; Step S106, determining the climate anomaly score according to the climate parameters; Step S107, collecting soil samples in the licorice growing area at multiple moments in the current monitoring period and performing spectral component analysis to obtain nutrient content information and moisture content information in the soil samples; Step S108, determining the soil anomaly score according to the nutrient content information and the moisture content information; Step S109, generating a licorice growth environment monitoring report according to the growth anomaly score, the climate anomaly score and the soil anomaly score.

[0025] According to the licorice growth environment monitoring method of an embodiment of the present invention, the probability of licorice being infested by insect pests can be predicted by predicting the insect pest probability model, thereby reducing the impact of insect pests on licorice growth. By performing spectral component analysis on soil samples to obtain nutrient element content information and water content information, the impact of soil conditions on licorice growth can be identified, and targeted soil improvement can be performed. Through the comprehensive evaluation of growth anomaly scores, climate anomaly scores, and soil anomaly scores, the advantages and disadvantages of the licorice growth environment can be reflected from multiple dimensions, thereby improving the comprehensiveness of monitoring the licorice growth environment.

[0026] According to an embodiment of the present invention, in step S101, a number of licorice plants are randomly selected for sampling in a licorice growing area, so that the selected sample licorice can represent the licorice growth conditions of the entire licorice growing area.

[0027] According to one embodiment of the present invention, in step S102, each monitoring period can be set to three days, seven days, etc., and the interval between adjacent moments can be set to 12 hours, 24 hours, etc., which is not limited by the present invention. The high-definition sample licorice image taken by a high-resolution camera device is clear for subsequent analysis and evaluation. While shooting, the growth height of each sample licorice is measured, and a ruler or a special measuring device can be used to record the actual height data of each sample licorice.

[0028] According to one embodiment of the present invention, in step S103, the high-definition sample licorice image is input into a trained pest probability prediction model, which can be a neural network model. Through machine learning or deep learning technology, it is trained based on a large amount of historical data, and can analyze the high-definition sample licorice image and predict the probability of occurrence of licorice pests.

[0029] According to one embodiment of the present invention, the training step of the pest prediction probability model includes: obtaining historical high-definition sample licorice images at multiple moments in multiple historical monitoring cycles, wherein the previous historical monitoring cycle of the current monitoring cycle is the first historical monitoring cycle; inputting the historical high-definition sample licorice images into the trained pest prediction probability model to obtain historical predicted pest probability data of multiple sample licorice at multiple moments in multiple historical monitoring cycles; performing feature extraction processing on the licorice rhizome area in the historical high-definition sample licorice images through the rhizome feature encoding level of the trained image recognition neural network model to obtain the historical rhizome area of ​​each historical sample licorice. stem feature vector; extracting features from the licorice leaf region in the historical high-definition sample licorice image through the leaf feature coding level of the trained image recognition neural network model to obtain the historical leaf feature vector of each historical sample licorice; extracting features from the licorice fruit region in the historical high-definition sample licorice image through the fruit feature coding level of the trained image recognition neural network model to obtain the historical fruit feature vector of each historical sample licorice; obtaining historical pest probability data of multiple sample licorice at multiple times in multiple historical monitoring cycles; determining the training loss function Loss of the pest probability prediction model according to formula (1),

[0030]

[0031] Among them, R i，s,k is the historical root feature vector of the i-th sample of licorice at the k-th moment in the s-th historical monitoring period, The nth historical monitoring period k The historical root feature vector of the i-th sample licorice at the moment, For R i，s，k The transposed vector, X i，s，k is the historical leaf feature vector of the i-th sample licorice at the k-th moment in the s-th historical monitoring period, The nth historical monitoring period k The historical leaf feature vector of the i-th sample licorice at the moment, For X i，s，k The transposed vector, Y i，s，k is the historical fruit feature vector of the i-th sample licorice at the k-th moment in the s-th historical monitoring period, The nth historical monitoring periodk The historical fruit feature vector of the i-th sample licorice at the moment, Y i，s，k The transposed vector, G i，s， k is the historical pest probability data of the i-th sample licorice at the k-th moment in the s-th historical monitoring period, G i，s，k，p is the historical predicted pest probability data of the i-th sample licorice at the k-th moment in the s-th historical monitoring period, n i is the number of sample licorice, n s is the number of historical monitoring cycles, n k is the number of monitoring cycle times, i≤n i , s≤n s , k≤n k , and i, s, k, n i 、n s and n k are all positive integers; according to the training loss function, the predicted pest probability model is trained to obtain the trained predicted pest probability model.

[0032] According to one embodiment of the present invention, the image recognition neural network model may be a convolutional neural network model. The rhizome feature coding level, the leaf feature coding level and the fruit feature coding level of the trained image recognition neural network model are used to extract the rhizome region, the leaf region and the fruit region of the licorice in the historical high-definition sample licorice image, respectively, to obtain the historical rhizome feature vector, the historical leaf feature vector and the historical fruit feature vector. The historical rhizome feature vector generally includes the appearance features of the rhizome, such as the length of the rhizome, the number of branches, the color, etc., which can be used to analyze and identify the health status of the rhizome. The historical leaf feature vector generally includes the appearance features of the leaf, such as the texture, area, edge shape, color, etc. of the leaf, which can be used to analyze and identify the health status of the leaf. The historical fruit feature vector generally includes the appearance features of the fruit, such as the size, shape, color, etc. of the fruit, which can be used to analyze and identify the health status of the fruit. Licorice has various pest manifestations, for example, after the short-haired grass weevil feeds, the leaf margin of the licorice is notched. Therefore, the historical rhizome feature vector, the historical leaf feature vector and the historical fruit feature vector will affect the probability of licorice being infested by insects. The historical pest probability data is based on historical monitoring records and reflects the occurrence of licorice pests at different time points. For each sample licorice, the historical pest probability data can be obtained from the recorded pest occurrence probability. For example, if the sample licorice has pests, the historical pest probability data of the sample licorice is 1; if the sample licorice has no pests, the historical pest probability data of the sample licorice is 0.

[0033] According to one embodiment of the present invention, as licorice grows, its root system gradually develops, its ability to absorb nutrients and water in the soil is enhanced, and its ability to resist insect pests is also improved accordingly. k The error in predicting the probability of pests at each moment is small, and the prediction result is more accurate. Therefore, in formula (1), is the historical root feature vector of the i-th sample of licorice at the k-th moment in the s-th historical monitoring period and the historical root feature vector of the n-th sample of licorice at the n-th moment in the s-th historical monitoring period. k The similarity of the historical rhizome feature vector of the i-th sample licorice at the moment is calculated. The closer the similarity is to 1, the smaller the impact of the change of licorice rhizome on the probability of insect pests is. Therefore, the change of the historical rhizome feature vector has a smaller impact on the error of predicting the probability of insect pests. is the historical leaf feature vector of the i-th sample licorice at the k-th moment in the s-th historical monitoring period and the historical leaf feature vector of the n-th sample licorice at the s-th historical monitoring period. k The similarity of the historical leaf feature vector of the i-th sample licorice at the moment is calculated. The closer the similarity is to 1, the smaller the impact of the change of licorice leaves on the probability of insect pests is. Therefore, the change of the historical leaf feature vector has a smaller impact on the error of predicting the probability of insect pests. is the historical fruit feature vector of the i-th sample licorice at the k-th moment in the s-th historical monitoring period and the historical fruit feature vector of the n-th sample licorice at the s-th historical monitoring period. k The similarity of the historical fruit feature vector of the i-th sample licorice at the moment. The closer the similarity is to 1, the smaller the impact of the change of licorice fruit on the probability of pests. Therefore, the change of the historical fruit feature vector has a smaller impact on the error of predicting the probability of pests. The historical root feature vector, historical leaf feature vector and historical fruit feature vector at the k-th moment of the s-th historical monitoring cycle and the n-th moment of the historical monitoring cycle are compared. k The more similar the feature vectors at each moment are, that is, and The larger the value of , the smaller the probability of pests, and the smaller the impact on the error of predicting the probability of pests. Therefore, the reciprocal of the product of the above three items is taken. The larger the reciprocal, the greater the probability of pests, and the greater the impact on the error of predicting the probability of pests. It is the relative error between the historical pest probability data of the i-th sample licorice at the k-th moment in the s-th historical monitoring period and the historical predicted pest probability data of the i-th sample licorice at the k-th moment in the s-th historical monitoring period. represents the weight at the kth moment, which is used to reasonably weight the relative errors at different moments in the loss function. For the sth historical monitoring period, the accuracy of the historical predicted pest probability data at the kth moment output by the predicted pest probability model is usually lower than the accuracy of the historical predicted pest probability data at the k+1th moment, that is, the accuracy of the historical predicted pest probability data at a certain moment in a historical monitoring period is different from that at the nth moment. kThe longer the time interval between moments, the less accurate the prediction result is, and therefore, the lower its weight is. Conversely, the more accurate the prediction result is, the higher its weight is.

[0034] It represents the weighted summation of the relative errors at each moment in the sth historical monitoring period. is the weight of the sth historical monitoring cycle, and the previous historical monitoring cycle of the current monitoring cycle is the first historical monitoring cycle. The longer the growth cycle of licorice, the lower the probability of being infested by insects. For example, when licorice is in the seedling stage, the root system is not fully developed, and the ability to absorb nutrients and water in the soil is limited. Therefore, the ability to resist insect pests is weak. That is, the longer the time interval between a historical monitoring cycle and the current monitoring cycle, the less accurate the prediction result, and therefore, the lower the weight. Conversely, the more accurate the prediction result, the higher the weight. Through the weight of the historical monitoring cycle, the weighted sum of the relative errors at each moment in the above sth historical monitoring cycle is weighted and summed to obtain the loss function of the pest prediction probability model. During the training process, the loss function is back-propagated, the model parameters are adjusted, and the loss function is minimized, thereby improving the accuracy of the pest prediction probability model and obtaining the trained pest prediction probability model.

[0035] In this way, the influence of historical root and stem feature vectors, historical leaf feature vectors and historical fruit feature vectors on the probability of licorice being affected by pests can be used to determine the influence of the above data on the error in predicting the probability of pests. Based on this influence and the relative error of the historical predicted probability of pests, and based on the characteristics that the shorter the moment interval and the shorter the cycle interval, the higher the prediction accuracy, weights are set to train the loss function to improve the objectivity and accuracy of the loss function, thereby reducing the training loss function during the training process, thereby improving the training intensity and efficiency, and improving the accuracy of the pest probability prediction model.

[0036] According to an embodiment of the present invention, in step S104, a growth abnormality score is determined based on the predicted pest probability data and the growth height data.

[0037] According to one embodiment of the present invention, step S104 includes: fitting the growth height data and multiple moments of the current monitoring period to obtain a growth height function; determining a growth height derivative function based on the growth height function; determining a growth height change rate based on the growth height derivative function; and determining a growth abnormality score based on the predicted pest probability data and the growth height change rate.

[0038] According to one embodiment of the present invention, the growth height data and the moments in the current monitoring cycle are fitted to obtain a growth height function for describing the changes in the growth height data in the current monitoring cycle over time, the growth height function is differentiated to obtain the growth height derivative function, and multiple growth height data are substituted into the growth height derivative function to determine the growth height change rate at multiple moments in the current monitoring cycle.

[0039] According to one embodiment of the present invention, determining a growth abnormality score according to the predicted pest probability data and the growth height change rate includes: determining a growth abnormality score A according to formulas (2) and (3): 1 ,

[0040]

[0041] Among them, H′ i，k H′ is the growth height change rate of the i-th sample licorice at the k-th moment in the current monitoring period, p is the growth height change rate threshold, F 1 is a piecewise function that compares the growth height change rate of the i-th sample licorice at the k-th moment in the current monitoring period with the growth height change rate threshold, G i，k，p is the predicted pest probability data of the i-th sample licorice at the k-th moment in the current monitoring period, α 1 and α 2 is the preset weight, and α 1 +α 2 =1,n i is the number of sample licorice, n k is the number of monitoring cycle times, i≤n i , k≤n k , and i, k, n i and n k All are positive integers.

[0042] According to one embodiment of the present invention, in formula (2), Indicates that when the growth height change rate of the i-th sample licorice at the k-th moment in the current monitoring period is less than the growth height change rate threshold, the value is When the growth height change rate of the i-th sample licorice at the k-th moment in the current monitoring period is greater than or equal to the growth height change rate threshold, the value is 0. is the relative difference between the growth height change rate of the i-th sample licorice at the k-th moment of the current monitoring period and the growth height change rate threshold, indicating the abnormality of the sample licorice growth height. The larger the relative difference, the more abnormal the sample licorice growth height, and the smaller the relative difference, the more normal the sample licorice growth height. In formula (3), It means that the predicted pest probability data of multiple licorice samples at multiple times in the current monitoring period are averaged to obtain the average predicted pest probability of multiple licorice samples in the current monitoring period. It means that the piecewise function values ​​corresponding to multiple licorice samples at multiple times in the current monitoring period are averaged to obtain the average abnormality of the growth height of multiple licorice samples in the current monitoring period. The average predicted pest probability and the average abnormality of the growth height of multiple licorice samples in the current monitoring period are weighted and summed to obtain the growth abnormality score. The higher the growth abnormality score, the more abnormal the growth condition of the licorice.

[0043] In this way, the growth abnormality score can be determined by the average predicted pest probability of licorice and the average abnormal degree of growth height. The growth status of licorice can be more comprehensively evaluated by the two aspects of pest probability and growth height change rate, which helps to detect growth abnormalities in time and improves the comprehensiveness and accuracy of growth abnormality scores.

[0044] According to one embodiment of the present invention, in step S105, the temperature, humidity and light intensity of the licorice growing area are measured by using a temperature sensor, a humidity sensor and a light intensity sensor, and the climate parameters are recorded for subsequent analysis.

[0045] According to an embodiment of the present invention, in step S106, a climate anomaly score is determined according to the climate parameter.

[0046] According to one embodiment of the present invention, step S106 includes: obtaining suitable climate parameters for the licorice growing area, wherein the suitable climate parameters include suitable temperature parameters, suitable humidity parameters and suitable light intensity parameters; obtaining historical climate parameters for the licorice growing area at multiple moments in the last historical monitoring cycle, wherein the historical climate parameters include historical temperature parameters, historical humidity parameters and historical light intensity parameters; determining a climate anomaly score based on the suitable climate parameters, the historical climate parameters and the climate parameters.

[0047] According to one embodiment of the present invention, the suitable temperature parameter is the optimal temperature for licorice growth, for example, the suitable temperature parameter is 20°C or 25°C, which is conducive to the germination of licorice seeds, the growth of seedlings and the development of roots. The suitable humidity parameter is the optimal humidity for licorice growth, for example, the suitable humidity parameter is 60% or 70%. It is conducive to the absorption of water and nutrients by the roots of licorice and promotes the growth and development of the plant. The suitable light intensity parameter is the best light intensity for promoting licorice photosynthesis, for example, the suitable light intensity parameter is 3000 or 4000 lux, which is conducive to the germination of licorice seeds and the growth of seedlings. By comparing the current climate parameters with the suitable climate parameters and the historical climate parameters, the gap between the current climate parameters and the suitable climate parameters for licorice growth can be identified.

[0048] According to one embodiment of the present invention, determining a climate anomaly score according to the suitable climate parameter, the historical climate parameter and the climate parameter includes: determining a climate anomaly score A according to formula (4): 2 ,

[0049]

[0050] Among them, t k is the temperature parameter at the kth moment of the current monitoring period, t p is the suitable temperature parameter, w k is the humidity parameter at the kth moment of the current monitoring period, w p is the suitable humidity parameter, m k is the light intensity parameter at the kth moment of the current monitoring period, m p is the appropriate light intensity parameter, t k，h is the historical temperature parameter at the kth moment of the last historical monitoring period, w k，h is the historical humidity parameter at the kth moment of the last historical monitoring period, m k，h is the historical light intensity parameter at the kth moment of the previous historical monitoring period, β 1 and β 2 is the preset weight, and β 1 +β 2 =1,n k is the number of monitoring cycle times, k≤n k , and k and n k All are positive integers, and max is the maximum value function.

[0051] According to one embodiment of the present invention, in formula (4), It is the relative difference between the temperature parameter at the kth moment of the current monitoring period and the suitable temperature parameter, indicating the degree of deviation of the current temperature. The larger the relative difference, the less suitable the current temperature is for licorice growth; the smaller the relative difference, the more suitable the current temperature is for licorice growth. It is the relative difference between the humidity parameter at the kth moment of the current monitoring period and the suitable humidity parameter, indicating the degree of deviation of the current humidity. The larger the relative difference, the less suitable the current humidity is for licorice growth; the smaller the relative difference, the more suitable the current humidity is for licorice growth. The relative difference between the light intensity parameter at the kth moment of the current monitoring period and the suitable light intensity parameter indicates the deviation degree of the current light intensity. The larger the relative difference, the less suitable the current light intensity is for licorice growth. The smaller the relative difference, the more suitable the current light intensity is for licorice growth. The deviation degree of the current climate parameter relative to the suitable climate parameter can be obtained by multiplying and averaging the above three items. The larger the deviation degree, the less suitable the current climate parameter is for licorice growth.

[0052] According to one embodiment of the present invention, in formula (4), It is the maximum value of the relative difference between the temperature parameters at multiple moments in the current monitoring cycle and the average value of the historical temperature parameters at multiple moments in the previous historical monitoring cycle. The larger the maximum value, the greater the degree of change in the temperature at multiple moments in the current monitoring cycle relative to the historical average temperature of the previous historical monitoring cycle, and the more abnormal the temperature in the licorice growing area. It is the maximum value of the relative difference between the humidity parameters at multiple moments in the current monitoring period and the average value of the historical humidity parameters at multiple moments in the previous historical monitoring period. The larger the maximum value, the greater the degree of change in the humidity at multiple moments in the current monitoring period relative to the historical average humidity of the previous historical monitoring period, and the more abnormal the humidity in the licorice growing area. It is the maximum value of the relative difference between the light intensity parameters at multiple moments in the current monitoring cycle and the average value of the historical light intensity parameters at multiple moments in the previous historical monitoring cycle. The larger the maximum value, the greater the degree of change of the light intensity at multiple moments in the current monitoring cycle relative to the historical average light intensity of the previous historical monitoring cycle, and the more abnormal the light intensity in the licorice growing area. The maximum value of the above three items is taken as the degree of change of the climate parameters in the licorice growing area.

[0053] According to one embodiment of the present invention, the degree of deviation of the current climate parameters relative to the suitable climate parameters and the degree of change of the climate parameters in the licorice growing area are weighted and summed to obtain a climate anomaly score. The higher the climate anomaly score, the more abnormal the climate in the licorice growing area.

[0054] In this way, the climate anomaly score can be determined through suitable climate parameters, historical climate parameters and current climate parameters, so as to effectively evaluate the climate anomaly in the licorice growing area. By comparing the three climate parameters of temperature, humidity and light intensity, the deviation between the current climate parameters and the suitable climate parameters is comprehensively evaluated, which helps to fully understand the impact of climate on licorice growth. The current climate parameters can be compared with the historical climate parameters, which helps to identify the trend of climate change and improve the comprehensiveness, scientificity and objectivity of the climate anomaly score.

[0055] According to one embodiment of the present invention, in step S107, soil samples are collected at multiple times in the current monitoring cycle, wherein the soil samples at multiple times are obtained at the same location and the same depth. Spectral component analysis is to identify substances and determine their chemical composition and relative content based on the spectrum of the substance. Spectral component analysis technology is used to detect and analyze soil samples at multiple times in the current monitoring cycle. Through spectral component analysis, nutrient element content information and water content information in the soil sample are obtained, for example, major nutrient elements such as nitrogen, phosphorus, and potassium.

[0056] According to an embodiment of the present invention, in step S108, a soil anomaly score is determined based on the nutrient element content information and the moisture content information.

[0057] According to one embodiment of the present invention, step S108 includes: obtaining multiple standard nutrient element content information and standard moisture content information in licorice soil; determining a standard soil state vector of the licorice soil based on the standard nutrient element content information and the standard moisture content information; determining a real-time soil state vector of the licorice soil based on the nutrient element content information and the moisture content information; and determining a soil anomaly score based on the standard soil state vector and the real-time soil state vector.

[0058] According to one embodiment of the present invention, the standard nutrient element content information and the standard moisture content information in the licorice soil need to be determined according to the growth requirements of the licorice. For example, the standard nutrient element content information of nitrogen is 3%, and the standard moisture content information of the soil required for licorice seed germination is 7.5%. The content and moisture content of each nutrient element are combined into a multidimensional data structure, namely, a standard soil state vector. The standard soil state vector reflects the ideal soil conditions for licorice growth and provides a comparative benchmark for subsequent real-time monitoring. The nutrient element content information and moisture content information in the licorice soil actually monitored are combined into a multidimensional data structure, namely, a real-time soil state vector. The real-time state vector reflects the actual nutritional status and moisture status of the soil at different times.

[0059] According to one embodiment of the present invention, determining a soil anomaly score according to the standard soil state vector and the real-time soil state vector includes: determining the soil anomaly score A according to formula (5): 3 ,

[0060]

[0061] Among them, C 1，k , C 2，k ,…,C N，k is the content information of the 1st, 2nd, ..., Nth nutrient element at the kth moment in the current monitoring period, C 1，T , C 2，T ,…,C N，T is the content information of the 1st, 2nd, ..., Nth standard nutrient element, S k is the moisture content information at the kth moment of the current monitoring period, S T is the standard moisture content information, γ 1 , γ 2 , …, γ N and γ N+1 is the preset weight, is the real-time soil state vector at the kth moment in the current monitoring period, is the standard soil state vector, N is the number of nutrient elements, n k is the number of monitoring cycle times, k≤n k , and k, N and n k All are positive integers.

[0062] According to one embodiment of the present invention, in formula (5), is the cosine similarity of the weighted real-time soil state vector and the weighted standard soil state vector. The closer the cosine similarity is to 1, the closer the weighted real-time soil state vector at the kth moment of the current monitoring cycle is to the weighted standard soil state vector, and the better the soil state at the kth moment of the current monitoring cycle is. Among them, the preset weights are multiplied by the corresponding nutrient element content information and moisture content information to represent the relative importance of different nutrient element content information and moisture content information in the soil state. The result of 1 minus the above cosine similarity can be obtained. The closer the result is to 1, the less similar the weighted real-time soil state vector at the kth moment of the current monitoring cycle is to the weighted standard soil state vector, and the worse the soil state at the kth moment of the current monitoring cycle is. The average value of the results at various moments of the current monitoring cycle is used as the soil anomaly score. The larger the soil anomaly score, the worse the soil state in the licorice growing area, and the soil environment needs to be improved, for example, a reasonable fertilization plan, timely drainage and irrigation, etc. The smaller the soil anomaly score, the better the soil state in the licorice growing area, and the more suitable it is for licorice growth.

[0063] In this way, the soil anomaly score can be determined based on the cosine similarity of the weighted real-time soil state vector and the weighted standard soil state vector, and the weights can be set based on the relative importance of different nutrients and water to licorice growth, making the soil state assessment more accurate and helpful for monitoring and improving the soil environment.

[0064] According to an embodiment of the present invention, in step S109, a licorice growth environment monitoring report is generated according to the growth anomaly score, the climate anomaly score and the soil anomaly score.

[0065] According to one embodiment of the present invention, step S109 includes: when the growth anomaly score is greater than the growth anomaly score threshold, generating a growth monitoring report; when the climate anomaly score is greater than the climate anomaly score threshold, generating a climate monitoring report; when the soil anomaly score is greater than the soil anomaly score threshold, generating a soil monitoring report.

[0066] According to one embodiment of the present invention, when the growth anomaly score is greater than the growth anomaly score threshold, it indicates that there is an anomaly in the current growth state, and a growth monitoring report needs to be generated. The growth monitoring report will record in detail the probability and height change of the current licorice being infested by insects, and provide corresponding solutions. When the climate anomaly score is greater than the climate anomaly score threshold, it indicates that there is an anomaly in the current climate parameters, and a climate monitoring report needs to be generated. The climate monitoring report will contain the current climate data, anomaly score, and analysis of the impact of possible abnormal climate on plant growth. When the soil anomaly score is greater than the soil anomaly score threshold, it indicates that there is an anomaly in the current soil state, and a soil monitoring report needs to be generated. The soil monitoring report will describe in detail the chemical properties of the soil and the state of the microbial community, analyze the cause of the anomaly, and make suggestions for improving soil quality.

[0067] According to the method for monitoring the growth environment of licorice of an embodiment of the present invention, the probability of licorice being infested with insect pests can be predicted by predicting the pest probability model, thereby reducing the impact of pests on the growth of licorice. By performing spectral component analysis on soil samples to obtain nutrient element content information and water content information, the impact of soil conditions on licorice growth can be identified, and targeted soil improvement can be performed. Through the comprehensive evaluation of growth anomaly scores, climate anomaly scores and soil anomaly scores, the advantages and disadvantages of the licorice growth environment can be reflected from multiple dimensions, and the comprehensiveness of monitoring the licorice growth environment can be improved. When determining the training loss function of the pest probability prediction model, the influence of the historical root and stem feature vectors, the historical leaf feature vectors and the historical fruit feature vectors on the probability of licorice being infested can be used to determine the influence of the above data on the error of the prediction of the pest probability, thereby based on the influence, the relative error of the historical prediction of the pest probability, and the characteristics of the shorter the time interval, the shorter the cycle interval, and the higher the prediction accuracy, the weights are set to train the loss function to improve the objectivity and accuracy of the loss function, thereby reducing the training loss function during the training process, improving the training intensity and training efficiency, and improving the accuracy of the pest probability prediction model. When determining the growth anomaly score, the growth anomaly score can be determined by the average predicted pest probability of licorice and the average abnormal degree of growth height. The growth state of licorice can be more comprehensively evaluated by the two aspects of pest probability and growth height change rate, which is helpful to timely discover growth anomalies and improve the comprehensiveness and accuracy of growth anomaly score. When determining the climate anomaly score, the climate anomaly score can be determined by suitable climate parameters, historical climate parameters and current climate parameters, so as to effectively evaluate the climate anomaly in the licorice growing area. By comparing the three climate parameters of temperature, humidity and light intensity, the deviation between the current climate parameters and the suitable climate parameters is comprehensively evaluated, which is helpful to fully understand the impact of climate on licorice growth. The current climate parameters can be compared with the historical climate parameters, which is helpful to identify the trend of climate change and improve the comprehensiveness, scientificity and objectivity of the climate anomaly score. When determining the soil anomaly score, the soil anomaly score can be determined based on the cosine similarity of the weighted real-time soil state vector and the weighted standard soil state vector, and the weights are set based on the relative importance of different nutrients and water to licorice growth, so that the soil state assessment is more accurate and helpful to monitor and improve the soil environment.

[0068] Figure 2A block diagram of a licorice growth environment monitoring system according to an embodiment of the present invention is exemplarily shown, the system comprising: a sample licorice module, used to sample licorice in a licorice growth area to obtain a plurality of sample licorice; a high-definition sample licorice image and growth height data module, used to perform high-definition photography of the sample licorice at multiple moments in the current monitoring cycle to obtain a plurality of high-definition sample licorice images and growth height data of the sample licorice; a predicted pest probability model module, used to input the high-definition sample licorice image into a trained predicted pest probability model to obtain predicted pest probability data of a plurality of sample licorice at multiple moments in the current monitoring cycle; a growth abnormality scoring module, used to determine a growth abnormality score according to the predicted pest probability data and the growth height data; A climate parameter module is used to obtain the climate parameters of the licorice growing area at multiple moments in the current monitoring period, wherein the climate parameters include temperature parameters, humidity parameters and light intensity parameters; a climate anomaly scoring module is used to determine the climate anomaly score based on the climate parameters; a soil sample module is used to collect soil samples in the licorice growing area at multiple moments in the current monitoring period and perform spectral component analysis to obtain nutrient element content information and moisture content information in the soil samples; a soil anomaly scoring module is used to determine the soil anomaly score based on the nutrient element content information and the moisture content information; a licorice growth environment monitoring report module is used to generate a licorice growth environment monitoring report based on the growth anomaly score, the climate anomaly score and the soil anomaly score.

[0069] It should be understood by those skilled in the art that the embodiments of the present invention described above and shown in the accompanying drawings are only examples and do not limit the present invention. The purpose of the present invention has been fully and effectively achieved. The functional and structural principles of the present invention have been demonstrated and explained in the embodiments, and the embodiments of the present invention may be deformed or modified in any way without departing from the principles.

Claims

1. A method for monitoring the growth environment of liquorice, characterized in that: include: Sampling licorice in a licorice growing area to obtain a plurality of sample licorice; taking high-definition photos of the sample licorice at multiple times in the current monitoring cycle to obtain a plurality of high-definition sample licorice images and growth height data of the sample licorice; inputting the high-definition sample licorice images into a trained prediction pest probability model to obtain prediction pest probability data of a plurality of sample licorice at multiple times in the current monitoring cycle; determining a growth anomaly score based on the predicted pest probability data and the growth height data; obtaining climate parameters of the licorice growing area at multiple times in the current monitoring cycle, wherein the climate parameters include temperature parameters, humidity parameters and light intensity parameters; determining a climate anomaly score based on the climate parameters; collecting soil samples in the licorice growing area and performing spectral component analysis at multiple times in the current monitoring cycle to obtain nutrient element content information and moisture content information in the soil samples; determining a soil anomaly score based on the nutrient element content information and the moisture content information; generating a licorice growth environment monitoring report based on the growth anomaly score, the climate anomaly score and the soil anomaly score; the training steps of the prediction pest probability model include: The method includes: obtaining historical high-definition sample licorice images at multiple moments in multiple historical monitoring cycles, wherein the previous historical monitoring cycle of the current monitoring cycle is the first historical monitoring cycle; inputting the historical high-definition sample licorice images into a trained predicted pest probability model to obtain historical predicted pest probability data of multiple sample licorice at multiple moments in multiple historical monitoring cycles; performing feature extraction processing on the licorice rhizome region in the historical high-definition sample licorice images through the rhizome feature coding level of the trained image recognition neural network model to obtain historical rhizome feature vectors of each historical sample licorice; performing feature extraction processing on the licorice leaf region in the historical high-definition sample licorice images through the leaf feature coding level of the trained image recognition neural network model to obtain historical leaf feature vectors of each historical sample licorice; performing feature extraction processing on the licorice fruit region in the historical high-definition sample licorice images through the fruit feature coding level of the trained image recognition neural network model to obtain historical fruit feature vectors of each historical sample licorice; obtaining historical pest probability data of multiple sample licorice at multiple moments in multiple historical monitoring cycles; and calculating the historical pest probability of multiple sample licorice at multiple moments in multiple historical monitoring cycles according to the formula Determine the training loss function Loss of the pest prediction probability model, where R i，s，k is the historical root feature vector of the i-th sample of licorice at the k-th moment in the s-th historical monitoring period, The nth historical monitoring period k The historical root feature vector of the i-th sample licorice at the moment, For R i，s，k The transposed vector, X i，s，k is the historical leaf feature vector of the i-th sample licorice at the k-th moment in the s-th historical monitoring period, The nth historical monitoring period k The historical leaf feature vector of the i-th sample licorice at the moment, For X i，s，k The transposed vector, Y i，s，k is the historical fruit feature vector of the i-th sample licorice at the k-th moment in the s-th historical monitoring period, The nth historical monitoring period k The historical fruit feature vector of the i-th sample licorice at the moment, Y i，s，k The transposed vector, G i，s，k is the historical pest probability data of the i-th sample licorice at the k-th moment in the s-th historical monitoring period, G i，s，k，p is the historical predicted pest probability data of the i-th sample licorice at the k-th moment in the s-th historical monitoring period, n i is the number of sample licorice, n s is the number of historical monitoring cycles, n k is the number of monitoring cycle times, i≤n i , s≤n s , k≤n k , and i, s, k, n i 、n s and n k are all positive integers; according to the training loss function, the predicted pest probability model is trained to obtain the trained predicted pest probability model.

2. The method for monitoring the growth environment of liquorice according to claim 1, characterized in that: Determine a growth abnormality score according to the predicted pest probability data and the growth height data, including: fitting the growth height data and multiple moments of the current monitoring period to obtain a growth height function; determine a growth height derivative function according to the growth height function; determine a growth height change rate according to the growth height derivative function; determine a growth abnormality score according to the predicted pest probability data and the growth height change rate.

3. The method for monitoring the growth environment of liquorice according to claim 2, characterized in that: Determining a growth abnormality score according to the predicted pest probability data and the growth height change rate includes: according to the formula and Determine the growth abnormality score A1, where H′ i，k H′ is the growth height change rate of the i-th sample licorice at the k-th moment in the current monitoring period, p is the growth height change rate threshold, F1 is the piecewise function that compares the growth height change rate of the i-th sample licorice at the k-th moment in the current monitoring period with the growth height change rate threshold, G i，k，p is the predicted pest probability data of the i-th sample licorice at the k-th moment in the current monitoring period, α1 and α2 are preset weights, and α1+α2=1, n i is the number of sample licorice, n k is the number of monitoring cycle times, i≤n i , k≤n k , and i, k, n i and n k All are positive integers.

4. The method for monitoring the growth environment of liquorice according to claim 1, characterized in that: Determining a climate anomaly score according to the climate parameters includes: obtaining suitable climate parameters for a licorice growing area, wherein the suitable climate parameters include suitable temperature parameters, suitable humidity parameters, and suitable light intensity parameters; obtaining historical climate parameters for the licorice growing area at multiple moments in the last historical monitoring cycle, wherein the historical climate parameters include historical temperature parameters, historical humidity parameters, and historical light intensity parameters; determining a climate anomaly score according to the suitable climate parameters, the historical climate parameters, and the climate parameters.

5. The method for monitoring the growth environment of liquorice according to claim 4, characterized in that: Determining a climate anomaly score according to the suitable climate parameter, the historical climate parameter and the climate parameter includes: according to the formula Determine the climate anomaly score A2, where t k is the temperature parameter at the kth moment of the current monitoring period, t p is the suitable temperature parameter, w k is the humidity parameter at the kth moment of the current monitoring period, w p is the suitable humidity parameter, m k is the light intensity parameter at the kth moment of the current monitoring period, m p is the appropriate light intensity parameter, t k，h is the historical temperature parameter at the kth moment of the last historical monitoring period, w k，h is the historical humidity parameter at the kth moment of the last historical monitoring period, m k，h is the historical light intensity parameter at the kth moment of the previous historical monitoring cycle, β1 and β2 are preset weights, and β1+β2=1, n k is the number of monitoring cycle times, k≤n k , and k and n k All are positive integers, and max is the maximum value function.

6. The method for monitoring the growth environment of liquorice according to claim 1, characterized in that: Determining a soil anomaly score according to the nutrient content information and the moisture content information includes: obtaining a plurality of standard nutrient content information and standard moisture content information in licorice soil; determining a standard soil state vector of the licorice soil according to the standard nutrient content information and the standard moisture content information; determining a real-time soil state vector of the licorice soil according to the nutrient content information and the moisture content information; and determining a soil anomaly score according to the standard soil state vector and the real-time soil state vector.

7. The method for monitoring the growth environment of liquorice according to claim 6, characterized in that: Determining a soil anomaly score according to the standard soil state vector and the real-time soil state vector includes: determining a soil anomaly score according to the formula Determine the soil anomaly score A3, where C 1，k , C 2，k ,…,C N，k is the content information of the 1st, 2nd, ..., Nth nutrient element at the kth moment in the current monitoring period, C 1，T , C 2，T ,…,C N，T is the content information of the 1st, 2nd, ..., Nth standard nutrient element, S k is the moisture content information at the kth moment of the current monitoring period, S T is the standard moisture content information, γ1, γ2, …, γ N and γ N+1 is the preset weight, is the real-time soil state vector at the kth moment in the current monitoring period, is the standard soil state vector, N is the number of nutrient elements, n k is the number of monitoring cycle times, k≤n k , and k, N and n k All are positive integers.

8. The method for monitoring the growth environment of liquorice according to claim 1, characterized in that: A licorice growth environment monitoring report is generated according to the growth anomaly score, the climate anomaly score and the soil anomaly score, including: when the growth anomaly score is greater than the growth anomaly score threshold, a growth monitoring report is generated; when the climate anomaly score is greater than the climate anomaly score threshold, a climate monitoring report is generated; when the soil anomaly score is greater than the soil anomaly score threshold, a soil monitoring report is generated.

9. A licorice growth environment monitoring system, characterized in that: include: A sample licorice module is used to sample licorice in a licorice growing area to obtain multiple sample licorice; A high-definition sample licorice image and growth height data module is used to take high-definition photos of the sample licorice at multiple moments in the current monitoring cycle to obtain multiple high-definition sample licorice images and growth height data of the sample licorice; a predicted pest probability model module is used to input the high-definition sample licorice image into a trained predicted pest probability model to obtain predicted pest probability data of multiple sample licorice at multiple moments in the current monitoring cycle; a growth anomaly scoring module is used to determine a growth anomaly score based on the predicted pest probability data and the growth height data; a climate parameter module is used to obtain the climate parameters of the licorice growth area at multiple moments in the current monitoring cycle, wherein the climate parameters include temperature parameters, humidity parameters and light intensity parameters; a climate anomaly scoring module is used to determine a climate anomaly score based on the climate parameters; The soil sample module is used to collect soil samples in the licorice growing area and perform spectral component analysis at multiple times in the current monitoring cycle to obtain nutrient element content information and moisture content information in the soil samples; the soil anomaly scoring module is used to determine the soil anomaly score based on the nutrient element content information and the moisture content information; the licorice growth environment monitoring report module is used to generate a licorice growth environment monitoring report based on the growth anomaly score, the climate anomaly score and the soil anomaly score.

Images