A method and system for evaluating the waterlogging tolerance of corn varieties based on data analysis

Abstract

The present invention discloses a method and system for evaluating the waterlogging tolerance of corn varieties based on data analysis, which relates to the technical field of evaluating the waterlogging tolerance of corn varieties, and includes the following steps: setting an initial monitoring frequency for collecting corn leaf information according to existing monitoring requirements; after setting the initial monitoring frequency, regularly collecting corn leaf information according to the initial monitoring frequency; as the collection of corn leaf data progresses, comparing the currently collected corn leaf data with the previously collected corn leaf data. The present invention accurately predicts the change trend of corn leaves through a pre-trained convolutional neural network, automatically adjusts the monitoring frequency, significantly improves the accuracy and real-time performance of waterlogging tolerance evaluation. This dynamic adjustment mechanism optimizes resource utilization, avoids ineffective high-frequency collection, improves monitoring efficiency, and at the same time responds quickly in a sudden environment, ensuring the stability of agricultural production, enhancing the risk resistance ability of corn, and helping to improve the stability of crop yields.

Description

Technical Field

[0001] The present invention relates to the technical field of waterlogging tolerance evaluation of corn varieties, and particularly relates to a method and system for evaluating waterlogging tolerance of corn varieties based on data analysis. Background Art

[0002] The waterlogging tolerance evaluation system for corn varieties is a comprehensive evaluation tool based on data analysis. By collecting and analyzing key index data during the growth process of corn, it evaluates the waterlogging tolerance of different corn varieties. The system collects crop physiological characteristics such as corn leaves, plant height, yield, and disease resistance, as well as meteorological data and soil moisture data, and uses algorithms to generate a waterlogging tolerance index. The system can identify corn varieties with strong waterlogging tolerance in waterlogging-prone areas to guide farmers to select excellent varieties suitable for planting in waterlogged environments, thereby improving yield stability and the agricultural risk resistance ability.

[0003] In the process of evaluating the waterlogging tolerance of corn varieties, collecting corn leaf information plays an important role because leaves are the main organs for corn photosynthesis, and their health status directly reflects the plant's ability to respond to environmental stress. In a waterlogging environment, leaves will show phenomena such as wilting, yellowing, and dry leaf margins. These changes are the result of the plant suffering from excessive water and insufficient oxygen. Therefore, by monitoring parameters such as the color, morphology, size, and water content of leaf tissues, the physiological response of corn plants to waterlogging can be accurately evaluated, providing key data support for waterlogging tolerance evaluation. This information can help predict the growth performance of corn varieties under waterlogging conditions, thereby judging their waterlogging tolerance.

[0004] The existing technology has the following deficiencies:

[0005] The existing technology usually collects corn leaf information at fixed monitoring frequencies, regularly collecting data based on a preset time interval. This helps to simplify the monitoring process and ensure the stability and consistency of data collection. However, although this fixed-frequency method can provide continuous monitoring data, in the face of different environmental conditions, especially sudden stress environments such as waterlogging, the physiological changes of corn leaves are often relatively rapid. The fixed collection interval may not be able to capture these key changes in real time, resulting in missing important response time points, thus affecting the accuracy and timeliness of waterlogging tolerance evaluation.

[0006] The above information disclosed in the background art section is only used to enhance the understanding of the background of the present disclosure. Therefore, it may include information that does not constitute the prior art known to those of ordinary skill in the art. Summary of the Invention

[0007] The object of the present invention is to provide a method and system for evaluating the waterlogging tolerance of corn varieties based on data analysis. By using a pre-trained convolutional neural network to accurately predict and classify the change trends of corn leaves, it can effectively distinguish rapid changes from regular changes. When rapid changes are detected, the system automatically adjusts the monitoring frequency, significantly increasing the acquisition frequency to ensure that all key physiological change data can be captured in a timely manner, thereby significantly improving the accuracy and real-time performance of the waterlogging tolerance evaluation. At the same time, dynamically adjusting the acquisition frequency optimizes resource utilization, avoids ineffective acquisitions, and improves monitoring efficiency. It not only solves the problem that a fixed monitoring frequency cannot reflect the rapid changes of corn leaves in real time, but also ensures the efficient operation of the system under both regular and emergency conditions, enhancing the risk resistance ability and yield stability of corn, so as to solve the problems in the above-mentioned background technology.

[0008] To achieve the above object, the present invention provides the following technical solution: A method for evaluating the waterlogging tolerance of corn varieties based on data analysis, comprising the following steps:

[0009] According to the existing monitoring requirements, set an initial monitoring frequency for the collection of corn leaf information;

[0010] After setting the initial monitoring frequency, regularly collect corn leaf information according to the initial monitoring frequency;

[0011] As the collection of corn leaf data progresses, compare the currently collected corn leaf data with the previously collected corn leaf data to form a comparison set, and track the changes of the current corn leaves;

[0012] After establishing the comparison set, compare and analyze the adjacent two monitoring data in the comparison set, and input the data after comparison and analysis into a pre-trained convolutional neural network to predict the changes of the current corn leaves;

[0013] Based on the prediction results of the convolutional neural network, divide the changes of the current corn leaves into two categories: rapid changes and regular changes;

[0014] If the change of the current corn leaves is a regular change, continue to collect according to the initial monitoring frequency;

[0015] If it is detected that the change of the current corn leaves is a rapid change, automatically adjust the actual monitoring frequency according to the change of the corn leaves, specifically by significantly increasing the acquisition frequency to ensure that all key data during the change process can be captured.

[0016] Preferably, after establishing the comparison set, the adjacent two monitoring data in the comparison set are compared and analyzed to extract the change in the degree of withering at the edges of corn leaves and the change in the chlorophyll content in corn leaves. After comparing and analyzing the change in the degree of withering at the edges of corn leaves and the change in the chlorophyll content in corn leaves, a leaf margin withering index and a chlorophyll degradation index are generated respectively. The change in the degree of withering at the edges of corn leaves is measured by the leaf margin withering index, reflecting the damage of the leaves under environmental stress conditions. The change in the chlorophyll content in corn leaves is measured by the chlorophyll degradation index, reflecting the decline in photosynthesis efficiency and the deterioration of leaf health status.

[0017] Preferably, after obtaining the leaf margin withering index and the chlorophyll degradation index generated by comparing and analyzing the adjacent two monitoring data in the comparison set, the leaf margin withering index and the chlorophyll degradation index are input into a pre-trained convolutional neural network to generate a change trend evaluation coefficient, and the change situation of the current corn leaves is predicted through the change trend evaluation coefficient.

[0018] Preferably, the change trend evaluation coefficient generated by predicting the adjacent two monitoring data in the comparison set through the convolutional neural network is compared and analyzed with a pre-set change trend evaluation coefficient reference threshold to classify the change situation of the current corn leaves. The specific classification steps are as follows:

[0019] If the change trend evaluation coefficient is greater than or equal to the change trend evaluation coefficient reference threshold, the change situation of the current corn leaves is classified as a rapid change, and the rapid change indicates that the current corn leaves have experienced drastic physiological changes in a short period of time;

[0020] If the change trend evaluation coefficient is less than the change trend evaluation coefficient reference threshold, the change situation of the current corn leaves is classified as a normal change, and the normal change means that the growth state of the current corn leaves is within the normal range, and the physiological characteristics change gradually with the growth cycle.

[0021] Preferably, the adjacent two monitoring data in the comparison set are compared and analyzed to extract the change in the degree of withering at the edges of corn leaves. The specific steps for generating the leaf margin withering index after comparing and analyzing the change in the degree of withering at the edges of corn leaves are as follows:

[0022] Based on the adjacent two monitoring data, the corn leaf image is processed to extract the change information at the edges of the corn leaves. Through an image processing algorithm, the leaf margin contour at each monitoring time point is generated, and then the change rate of the withered part at the edges of the corn leaves between two adjacent monitorings is calculated. The calculation expression is:

[0023]

[0024] , where ΔE t represents the edge withering change rate, Ct and C t-1 are the contour areas of the maize leaf margins for the current and previous monitoring respectively, and A t is the overall area of the maize leaf, avoiding interference from the overall size of the maize leaf on the calculation results;

[0025] On the basis of extracting the dry area, further analyze the changes in the details of the maize leaf margin. Take the local curvature of the maize leaf margin and the refinement degree of the dry margin as key parameters, and describe the severity of the dryness of the maize leaf margin through the point-by-point curvature change, quantifying the change degree of the details of the dry margin. The specific quantification expression is: In the formula, K t is the total amount of curvature change of the dry margin, and κ t (i) and κ t-1 (i) are the curvatures of the i-th marginal point in the current and previous monitoring respectively, and n is the total number of marginal points;

[0026] Generate the final leaf margin dryness index by introducing the edge dryness change rate and the total amount of curvature change of the dry margin, and comprehensively quantify the degree of dryness of the maize leaf margin. Then the expression for generating the leaf margin dryness index is: EDI t =α·ΔE t +β·(K t ) γ , in the formula, EDI t is the leaf margin dryness index, α and β are the weight coefficients of the edge dryness change rate and the total amount of curvature change respectively, and γ is a regulation factor to enhance the sensitivity to the curvature change.

[0027] Preferably, compare and analyze the adjacent two monitoring data in the comparison set, extract the change in the chlorophyll content in the maize leaf, and the specific steps for generating the chlorophyll degradation index after comparing and analyzing the change in the chlorophyll content in the maize leaf are as follows:

[0028] Extract the chlorophyll content from the maize leaf data collected in two adjacent times. Let the chlorophyll content at the t1-th collection be and the chlorophyll content at the t2-th collection be Calculate the chlorophyll change amount between the two monitorings. The calculation formula is: In the formula, ΔR represents the chlorophyll content change amount during the two monitorings;

[0029] Introduce an acceleration coefficient to reflect the acceleration of the chlorophyll content change rate, and amplify the chlorophyll content change amount through a non-linear formula. The specific amplification expression is:

[0030]

[0031] , where λ is the chlorophyll degradation acceleration coefficient, which measures the acceleration effect of chlorophyll loss, and τ 12 is the time interval between two adjacent monitoring times, and k is the adjustment index, which is used to amplify or reduce the acceleration degree of the change;

[0032] By combining the change amount of chlorophyll content with the chlorophyll degradation acceleration coefficient, a chlorophyll degradation index is generated. The specific expression is:

[0033]

[0034] , where δ chlor is the chlorophyll degradation index, and the final δ chlor value is used to quantify the degradation degree of chlorophyll.

[0035] Preferably, the actual monitoring frequency is automatically adjusted according to the change of corn leaves. The specific steps are as follows:

[0036] Calculate the deviation between the change trend evaluation coefficient and the reference threshold of the change trend evaluation coefficient to obtain the change trend deviation coefficient Δ δ , Δ δ = Chang trend -δ threshold , where Chang trend is the change trend evaluation coefficient generated based on the convolutional neural network, which reflects the change degree of corn leaves, and δ threshold is the preset reference threshold of the change trend evaluation coefficient, which is used to distinguish rapid changes and regular changes;

[0037] According to the change trend deviation coefficient Δ δ , calculate the monitoring frequency adjustment factor, and use a non-linear function to enhance the response to rapid changes. The calculation expression of the monitoring frequency adjustment factor is: In the formula, φ is the monitoring frequency adjustment factor, which is used to adjust the actual monitoring frequency, and w is the sensitivity coefficient, which controls the growth rate of the monitoring frequency adjustment factor;

[0038] Use the initial monitoring frequency and the monitoring frequency adjustment factor to calculate the adjusted actual monitoring frequency. The calculation expression is: f actual = f initial ·φ. In the formula, f initial is the initial monitoring frequency, and f actual is the adjusted actual monitoring frequency, which will be used for subsequent data collection;

[0039] In order to avoid excessive consumption of system resources caused by too high monitoring frequency, set the upper limit of the actual monitoring frequency and adjust the actual monitoring frequency. The expression of the adjusted actual monitoring frequency is: f actual = min(f actual , fmax ) where f max is the maximum allowable value of the actual monitoring frequency, and the min(·) function represents taking the smaller of the two values within the parentheses, ensuring that the actual monitoring frequency f actual does not exceed the maximum allowable value f of the actual monitoring frequency max ;

[0040] Based on the calculated actual monitoring frequency f actual , update the monitoring plan of the data acquisition system to perform high-frequency acquisition according to the new frequency, ensuring that key data during the change process of corn leaves can be captured. At the same time, continuously monitor the change trend evaluation coefficient. If the subsequent change trend evaluation coefficient drops below the reference threshold of the change trend evaluation coefficient, then restore to the initial monitoring frequency.

[0041] A waterlogging tolerance evaluation system for corn varieties based on data analysis, including an initial monitoring frequency setting module, a regular data acquisition module, a data comparison set generation module, a convolutional neural network prediction module, a change situation classification module, a regular change monitoring module, and a rapid change frequency adjustment module;

[0042] The initial monitoring frequency setting module sets an initial monitoring frequency for the acquisition of corn leaf information according to the existing monitoring requirements;

[0043] The regular data acquisition module regularly acquires corn leaf information according to the initial monitoring frequency after setting the initial monitoring frequency;

[0044] The data comparison set generation module compares the currently acquired corn leaf data with the previously acquired corn leaf data as the corn leaf data is acquired, forms a comparison set, and tracks the change situation of the current corn leaves;

[0045] The convolutional neural network prediction module, after establishing the comparison set, compares and analyzes the adjacent two monitoring data in the comparison set, and inputs the data after comparison and analysis into a pre-trained convolutional neural network to predict the change situation of the current corn leaves;

[0046] The change situation classification module divides the change situation of the current corn leaves into two categories based on the prediction result of the convolutional neural network: rapid change and regular change;

[0047] The regular change monitoring module, if the change of the current corn leaves is a regular change, continues to acquire data according to the initial monitoring frequency;

[0048] The rapid change frequency adjustment module, if it detects that the change of the current corn leaves is a rapid change, automatically adjusts the actual monitoring frequency according to the change situation of the corn leaves, specifically by significantly increasing the acquisition frequency to ensure that all key data during the change process can be captured.

[0049] In the above technical solution, the technical effects and advantages provided by the present invention are as follows:

[0050] The present invention accurately predicts and classifies the change trend of corn leaves through a pre-trained convolutional neural network, and can effectively distinguish rapid changes from conventional changes. When it detects rapid changes in corn leaves, the system automatically adjusts the monitoring frequency, significantly increasing the acquisition frequency to ensure that all key physiological change data can be captured in a timely manner, thereby significantly improving the accuracy and real-time nature of waterlogging tolerance assessment. The mechanism of dynamically adjusting the acquisition frequency optimizes resource utilization, avoids ineffective high-frequency acquisitions, and greatly improves the monitoring efficiency. It not only solves the problem that the traditional fixed monitoring frequency cannot reflect the rapid changes of corn leaves in real time, but also ensures the efficient operation of the system under normal conditions, and can quickly respond in a sudden environment to ensure the stability of agricultural production, providing a more scientific and reliable decision-making basis for the waterlogging tolerance assessment of corn varieties, significantly enhancing the risk resistance ability of corn, and helping to improve the stability of crop yields. Description of the Drawings

[0051] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those of ordinary skill in the art, other drawings can also be obtained based on these drawings.

[0052] Figure 1 It is a method flow chart of a method for evaluating the waterlogging tolerance of corn varieties based on data analysis according to the present invention.

[0053] Figure 2 It is a module schematic diagram of a system for evaluating the waterlogging tolerance of corn varieties based on data analysis according to the present invention. Detailed Embodiments

[0054] Now, the exemplary embodiments will be described more comprehensively with reference to the drawings. However, the exemplary embodiments can be implemented in various forms and should not be construed as limited to the examples set forth herein; on the contrary, these exemplary embodiments are provided so that the present disclosure will be more complete and comprehensive, and the concept of the exemplary embodiments will be fully conveyed to those skilled in the art.

[0055] The present invention provides a method for evaluating the waterlogging tolerance of corn varieties based on data analysis as Figure 1 shown, including the following steps:

[0056] Set an initial monitoring frequency for corn leaf information acquisition according to the existing monitoring requirements;

[0057] The initial monitoring frequency should consider the variation law of corn leaves under normal growth environment, and combine different growth stages of corn and external environmental conditions (such as temperature, humidity, etc.) to formulate a basic standard for the sampling interval of leaves. On this basis, the initial monitoring frequency can be optimized through long-term accumulated historical data to ensure that sufficient monitoring data can be provided under normal circumstances, laying a foundation for subsequent dynamic adjustment. At the same time, the setting of the initial frequency should take into account the consumption of system resources, ensuring both the effective collection of information and avoiding unnecessary system pressure caused by too high a frequency.

[0058] After setting the initial monitoring frequency, regularly collect corn leaf information according to the initial monitoring frequency;

[0059] Use sensors, image analysis equipment or unmanned aerial vehicle (UAV) monitoring systems to obtain relevant data of the leaves, such as physiological characteristics of the leaves like color, shape, size, water content, etc. To ensure the accuracy and consistency of the data, the collection equipment needs to have high precision and be able to adapt to the complex environmental conditions in the field. During the data collection process, factors such as weather and light changes that affect the data quality also need to be considered. Therefore, the integration of multiple sensing devices can be introduced to enhance the diversity and stability of data collection.

[0060] The collection equipment specifically includes spectral sensors, RGB and multispectral cameras, near-infrared sensors (for monitoring the water content of leaves), laser scanners (for measuring the shape and size of leaves), temperature and humidity sensors (for monitoring environmental conditions), as well as the high-resolution camera and sensor combination carried by UAVs. The combined use of these devices can collect physiological characteristics such as the color, morphology, size and water content of corn leaves, and at the same time, in the complex field environment, it has anti-interference ability to ensure high-precision data can still be provided under different lighting and weather conditions.

[0061] As the collection of corn leaf data progresses, compare the currently collected corn leaf data with the previously collected corn leaf data to form a comparison set and track the changes of the current corn leaves;

[0062] During the process of collecting corn leaf data, each newly collected data will be compared and analyzed with the previously collected old data. By comparing the current corn leaf data (such as color, shape, size, water content, etc.) with the data collected last time, the system can identify the changes that have occurred to the leaves within the time period. This comparison can help monitor the growth trend of corn leaves or the physiological responses when dealing with environmental changes, determine whether there are significant changes in the leaves, and then decide whether to adjust the monitoring frequency or take other management measures.

[0063] After establishing the comparison set, compare and analyze the adjacent two monitoring data in the comparison set, and input the data after comparison and analysis into a pre-trained convolutional neural network to predict the changes of the current corn leaves;

[0064] After establishing the comparison set, compare and analyze the adjacent two monitoring data in the comparison set, extract the changes in the withering degree of the corn leaf edges and the changes in the chlorophyll content in the corn leaves, generate a leaf edge withering index and a chlorophyll degradation index respectively after comparing and analyzing the changes in the withering degree of the corn leaf edges and the changes in the chlorophyll content in the corn leaves. Measure the changes in the withering degree of the corn leaf edges through the leaf edge withering index, reflect the damage of the leaves under environmental stress conditions, and measure the changes in the chlorophyll content in the corn leaves through the chlorophyll degradation index, reflecting the decline in photosynthesis efficiency and the deterioration of leaf health.

[0065] After obtaining the leaf edge withering index and the chlorophyll degradation index generated by comparing and analyzing the adjacent two monitoring data in the comparison set, input the leaf edge withering index and the chlorophyll degradation index into a pre-trained convolutional neural network to generate a change trend evaluation coefficient, and predict the changes of the current corn leaves through the change trend evaluation coefficient.

[0066] The pre-trained convolutional neural network refers to a neural network model that has been learned and optimized through a large amount of historical data, specifically used to analyze the input image or time series data to help identify the change trend of corn leaves. In this context, the input data of the convolutional neural network are the leaf edge withering index and the chlorophyll degradation index generated according to the leaf monitoring information. Through the input of these indices, the convolutional neural network can automatically extract and identify the hidden features and patterns in the leaf change process. Different from traditional analysis methods, the convolutional neural network can extract important local features from complex multi-dimensional data through adaptive learning, and gradually abstract through a hierarchical structure to discover the non-linear relationships among them. This makes it particularly suitable for dealing with the evaluation problem of the physiological changes of corn leaves.

[0067] In the prediction of the changing trends of corn leaves, the convolutional neural network has, through the training of a large amount of historical data of corn leaves, learned to recognize the typical features of leaf changes under various environmental stress conditions and summarized the changing patterns that can reflect the health status of the leaves. This pre-trained model not only has the ability to identify individual changes (such as chlorophyll decline and leaf margin withering), but also can comprehensively analyze the synergistic effects of multiple physiological changes through the correlation between data. Through layers of convolution and pooling operations, the model gradually simplifies and refines the input leaf margin withering index and chlorophyll degradation index, learns the changing trends of the leaves under waterlogging or other environmental stresses, and finally generates a changing trend evaluation coefficient. This changing trend evaluation coefficient can be used to quantify the current health status of the leaves, help predict whether the leaves will show an accelerating deterioration or a gradual recovery trend in the future, and thus provide accurate decision-making basis for agricultural managers.

[0068] If the change in the withering degree of the edge of the currently collected corn leaf suddenly accelerates, it indicates that the changing trend of the corn leaf under waterlogging environment is changing rapidly. This suddenly intensified withering phenomenon is usually the result of the corn plant being subjected to strong environmental stress, such as long-term waterlogging or soil hypoxia, resulting in water loss from the leaves and increased cell damage. Therefore, the rapid deterioration of the withering degree of the edge can be regarded as a direct manifestation of the corn leaf being affected by waterlogging, indicating that the current physiological state of the leaf is deteriorating rapidly and the plant may face greater growth pressure.

[0069] Compare and analyze the adjacent two monitoring data in the comparison set, extract the change in the withering degree of the edge of the corn leaf. The specific steps for comparing and analyzing the change in the withering degree of the edge of the corn leaf to generate the leaf margin withering index are as follows:

[0070] Based on the adjacent two monitoring data, process the corn leaf image, extract the change information of the edge of the corn leaf, generate the leaf edge contour at each monitoring time point through the image processing algorithm, and then calculate the change rate of the withered part of the corn leaf edge between the adjacent two monitorings. The calculation formula is:

[0071]

[0072] , where ΔE t represents the edge withering change rate, C t and C t-1 are the areas of the leaf edge contours of the corn leaf at the current monitoring and the previous monitoring respectively, and A t is the overall area of the corn leaf to avoid the interference of the overall size of the corn leaf on the calculation result;

[0073] The image processing algorithms that can be used include Canny edge detection, Sobel operator, Laplacian edge detection, and Active Contour model (snake algorithm). Canny edge detection is a classic and efficient edge detection method that can accurately locate the edges of corn leaves and is suitable for high-noise environments. The Sobel operator extracts edge information by calculating the brightness gradient in the image and is suitable for detecting relatively smooth leaf edges. Laplacian edge detection obtains the edge features with sharp changes in the image through the second derivative and is especially suitable for high-contrast regions. The Active Contour model is a contour-based segmentation method that can dynamically change with the leaf edge and adaptively adjust the contour boundary to more accurately capture the leaf edges with complex shapes. Combining these algorithms can generate high-precision leaf edge contours to meet the needs of comparison and analysis.

[0074] By calculating the change in the area of the edge contour, the direct change in the dryness degree of the corn leaf edge is captured, with a focus on the change in the dry edge relative to the overall area of the corn leaf.

[0075] On the basis of extracting the dry area, the changes in the corn leaf edge at the detail level are further analyzed. The local curvature of the corn leaf edge and the refinement degree of the dry edge are used as key parameters, and the severity of the dryness of the corn leaf edge is described by the change in curvature at each point to quantify the change degree of the details of the dry edge. The specific quantification expression is: In the formula, K t is the total change in the curvature of the dry edge, κ t (i) and κ t-1 (i) are the curvatures of the i-th edge point in the current and the previous monitoring (monitoring of corn leaf information) respectively, and n is the total number of edge points;

[0076] The change in curvature reflects the dryness degree of the leaf edge. Especially when drying rapidly, the bending and breakage of the edge will be more obvious. This step further quantifies the dryness degree of the edge by analyzing the geometric changes of the leaf edge.

[0077] By introducing the change rate of edge dryness and the total change in the curvature of the dry edge, the final leaf edge dryness index is generated to comprehensively quantify the dryness degree of the corn leaf edge. The expression for generating the leaf edge dryness index is: EDI t =α·ΔE t +β·(K t ) γ , where EDI t is the leaf edge dryness index, α and β are the weight coefficients of the change rate of edge dryness and the total change in curvature respectively, which are adjusted based on different experimental environments, and γ is a adjustment factor to enhance the sensitivity to the change in curvature;

[0078] This step synthesizes the edge drying rate and the total change in the curvature of the dried edge. The leaf edge drying index is generated through the non-linear combination of the two, which is used to reflect the degree of edge drying of corn leaves under environmental stress. The weight coefficients α, β and the parameter γ can be adjusted according to specific application scenarios, so that the index can flexibly adapt to the changes of corn leaves in different growth environments.

[0079] From the leaf edge drying index, it can be seen that in a waterlogging environment, the larger the performance value of the leaf edge drying index generated after comparing and analyzing the changes in the degree of edge drying of corn leaves indicates that the current corn leaves have changed significantly compared with the previous monitoring. This means that the degree of edge drying of corn leaves has increased, which may be due to serious damage to leaf tissues caused by continuous environmental stress. By quantifying the changes in the dried edge, the leaf edge drying index can intuitively reflect the health status of the leaves at different time points and the degree of waterlogging damage. Therefore, the larger the leaf edge drying index, the faster the physiological state of the leaves deteriorates, and the drying range and degree at the edge increase significantly.

[0080] If the chlorophyll content in the currently collected corn leaf data decreases rapidly, it usually indicates that the change trend of corn leaves in a waterlogging environment is occurring rapidly. Chlorophyll is a key substance for plants to carry out photosynthesis. Waterlogging causes hypoxia in the roots and hinders nutrient absorption, which will directly affect the generation and maintenance of chlorophyll, resulting in a rapid decrease in its content. This rapid decrease usually reflects the physiological stress response of the plant when coping with environmental stress, indicating that the corn leaves are in a relatively serious damaged state, and the overall health of the plant is deteriorating rapidly, belonging to a rapid change trend.

[0081] Compare and analyze the adjacent two monitoring data in the comparison set, extract the change in the chlorophyll content in the corn leaves, and the specific steps for generating the chlorophyll degradation index after comparing and analyzing the change in the chlorophyll content in the corn leaves are as follows:

[0082] Extract the chlorophyll content from the corn leaf data collected twice adjacent. Let the chlorophyll content at the t1th collection be The chlorophyll content at the t2th collection is Calculate the chlorophyll change amount between the two monitors. The calculation formula is: In the formula, ΔR represents the chlorophyll content change amount during the two monitors;

[0083] The greater the chlorophyll content change amount indicates the rapid loss of chlorophyll, which is usually related to the degree of environmental stress on the plant;

[0084] Chlorophyll content can be accurately measured by advanced spectral sensors or near-infrared (NIR) technology. Spectral sensors detect the chlorophyll content in leaves by analyzing the light reflection or absorption characteristics at different wavelengths, especially in the red and near-infrared light regions, because chlorophyll has significant absorption characteristics for these specific bands of light. Near-infrared technology indirectly estimates the chlorophyll concentration by capturing the reflection of near-infrared light by water and other components in the leaves. Both technologies can non-destructively monitor the health status of plants in real time and have high spatial and temporal resolutions, enabling researchers to accurately capture the dynamic changes of chlorophyll under complex field conditions. These technologies are not only applicable to the monitoring of individual leaves but can also be used to achieve large-scale assessment of the health of field crops through sensors carried by drones or satellites.

[0085] An acceleration coefficient is introduced to reflect the acceleration of the change rate of chlorophyll content, and the change amount of chlorophyll content is amplified through a non-linear formula. The specific amplification expression is:

[0086]

[0087] , where λ is the chlorophyll degradation acceleration coefficient, which measures the acceleration effect of chlorophyll loss. A larger λ value means a rapid decrease in chlorophyll content, indicating that the plant is experiencing severe stress, and τ 12 is the time interval between two adjacent monitors, representing the changes that occur in the leaves over a certain period of time. κ is the adjustment index, which is used to amplify or reduce the acceleration degree of the change and is usually adjusted according to experimental data, with typical values between 1.5 and 2.5;

[0088] By combining the change amount of chlorophyll content with the chlorophyll degradation acceleration coefficient, a chlorophyll degradation index is generated. The specific expression for generation is:

[0089]

[0090] , where δ chlor is the chlorophyll degradation index, and the final δ chlor value is used to quantify the degree of chlorophyll degradation. The larger the δ chlor , the more severe the chlorophyll loss and the higher the degree of environmental stress on the plant.

[0091] From the chlorophyll degradation index, it can be seen that in a waterlogging environment, the larger the performance value of the chlorophyll degradation index generated after comparing and analyzing the changes in the chlorophyll content in corn leaves means that the current corn leaves have changed significantly compared to the corn leaves in the previous monitoring. The larger the chlorophyll degradation index, the faster the degradation rate of chlorophyll, usually indicating that the corn plants are experiencing severe physiological stress responses, such as a large amount of chlorophyll loss and significant hindrance to photosynthesis. The increase in the chlorophyll degradation index can be used as an important indicator of the increasing degree of waterlogging damage to corn leaves, indicating a rapid deterioration of the plant health status.

[0092] The convolutional neural network is not specifically limited here, and it can be any convolutional neural network that can realize the comprehensive analysis of the edge dryness index EDI t and the chlorophyll degradation index δ chlor to generate the change trend evaluation coefficient Chang trend For the sake of implementing the technical solution of the present invention, the present invention provides a specific implementation method to implement the technical solution of the present invention; the calculation formula for generating the change trend evaluation coefficient Chang trend is: Chang trend = r1 * EDI t + r2 * δ chlor , where r1 and r2 are the preset proportionality coefficients of the edge dryness index EDI t and the chlorophyll degradation index δ chlor respectively, and both r1 and r2 are greater than 0.

[0093] From the change trend evaluation coefficient, it can be seen that in a waterlogging environment, the larger the performance value of the edge dryness index generated after comparing and analyzing the changes in the dryness degree of the corn leaf edges, and the larger the performance value of the chlorophyll degradation index generated after comparing and analyzing the changes in the chlorophyll content in the corn leaves, that is, the larger the performance value of the change trend evaluation coefficient generated by the convolutional neural network for predicting the adjacent two monitoring data in the comparison set, the greater the change of the current corn leaves compared to the corn leaves in the previous monitoring, and vice versa, indicating that the change of the current corn leaves compared to the corn leaves in the previous monitoring is smaller.

[0094] Based on the prediction results of the convolutional neural network, the change situation of the current corn leaves is divided into two categories: rapid change and normal change;

[0095] The change trend evaluation coefficient generated by the convolutional neural network for predicting the adjacent two monitoring data in the comparison set is compared and analyzed with the preset change trend evaluation coefficient reference threshold to divide the change situation of the current corn leaves. The specific division steps are as follows:

[0096] If the change trend evaluation coefficient is greater than or equal to the reference threshold of the change trend evaluation coefficient, the change situation of the current corn leaf is classified as a rapid change. A rapid change indicates that the current corn leaf has experienced drastic physiological changes in a short period of time, usually caused by sudden stress conditions in the external environment. A rapid change reflects that the corn plant is under relatively high physiological stress;

[0097] If the change trend evaluation coefficient is less than the reference threshold of the change trend evaluation coefficient, the change situation of the current corn leaf is classified as a normal change. A normal change means that the growth state of the current corn leaf is within the normal range, and its physiological characteristics gradually change with the growth cycle. Such changes are usually relatively stable, and within the expected growth stage, there are no obvious signs of stress response. A normal change indicates that the corn plant is in a good growth environment, can carry out normal photosynthesis and growth, and the overall health state of the plant is stable.

[0098] If the change of the current corn leaf is a normal change, continue to collect data according to the initial monitoring frequency;

[0099] If it is detected that the change of the current corn leaf is a rapid change, automatically adjust the actual monitoring frequency according to the change situation of the corn leaf. Specifically, significantly increase the collection frequency to ensure that all key data during the change process can be captured;

[0100] Automatically adjust the actual monitoring frequency according to the change situation of the corn leaf. The specific steps are as follows:

[0101] Calculate the deviation between the change trend evaluation coefficient and the reference threshold of the change trend evaluation coefficient to obtain the change trend deviation coefficient Δ δ , Δ δ =Chang trend -δ threshold , where Chang trend is the change trend evaluation coefficient generated based on the convolutional neural network, reflecting the degree of change of the corn leaf, and δ threshold is the preset reference threshold of the change trend evaluation coefficient, used to distinguish rapid changes and normal changes;

[0102] According to the change trend deviation coefficient Δ δ , calculate the monitoring frequency adjustment factor, and use a non-linear function to enhance the response to rapid changes. The calculation expression of the monitoring frequency adjustment factor is: In the formula, φ is the monitoring frequency adjustment factor, used to adjust the actual monitoring frequency, w is the sensitivity coefficient, which controls the growth rate of the monitoring frequency adjustment factor. The typical value is a positive number, and the larger the value, the higher the sensitivity to changes.

[0103] Use the initial monitoring frequency and the monitoring frequency adjustment factor to calculate the adjusted actual monitoring frequency. The calculation expression is: factual = f initial ·φ, where f initial is the initial monitoring frequency, i.e., the set basic acquisition frequency, and f actual is the adjusted actual monitoring frequency, which will be used for subsequent data acquisition;

[0104] To avoid excessive consumption of system resources caused by too high monitoring frequency, an upper limit of the actual monitoring frequency is set, and the actual monitoring frequency is adjusted. The expression of the adjusted actual monitoring frequency is: f actual = min(f actual , f max ), where f max is the maximum allowable value of the actual monitoring frequency, which is preset according to the system capabilities and resource limitations. The min(·) function represents taking the smaller of the two values in the parentheses, ensuring that the actual monitoring frequency f actual does not exceed the maximum allowable value f max of the actual monitoring frequency;

[0105] According to the calculated actual monitoring frequency f actual , update the monitoring plan of the data acquisition system to perform high-frequency acquisition according to the new frequency, ensuring to capture the key data during the change process of corn leaves. At the same time, continuously monitor the change trend evaluation coefficient. If the subsequent change trend evaluation coefficient drops below the reference threshold of the change trend evaluation coefficient, then restore to the initial monitoring frequency.

[0106] The present invention can accurately predict and classify the change trend of corn leaves through a pre-trained convolutional neural network, and can effectively distinguish rapid changes and regular changes. When it detects a rapid change in corn leaves, the system automatically adjusts the monitoring frequency, significantly increasing the acquisition frequency to ensure timely capture of all key physiological change data, thereby significantly improving the accuracy and real-time performance of waterlogging tolerance evaluation. The mechanism of dynamically adjusting the acquisition frequency optimizes resource utilization, avoids ineffective high-frequency acquisition, and greatly improves the monitoring efficiency. It not only solves the problem that the traditional fixed monitoring frequency cannot reflect the rapid changes of corn leaves in real time, but also ensures the efficient operation of the system under normal conditions, and can quickly respond in a sudden environment to ensure the stability of agricultural production, providing a more scientific and reliable decision-making basis for the waterlogging tolerance evaluation of corn varieties, significantly enhancing the risk resistance ability of corn, and helping to improve the stability of crop yields.

[0107] The present invention provides a waterlogging tolerance evaluation system for corn varieties based on data analysis as Figure 2 shown, including an initial monitoring frequency setting module, a regular data acquisition module, a data comparison set generation module, a convolutional neural network prediction module, a change situation classification module, a regular change monitoring module, and a rapid change frequency adjustment module;

[0108] An initial monitoring frequency setting module sets an initial monitoring frequency for corn leaf information collection according to existing monitoring requirements;

[0109] A regular data collection module regularly collects corn leaf information according to the initial monitoring frequency after setting the initial monitoring frequency;

[0110] A data comparison set generation module compares the currently collected corn leaf data with the previously collected corn leaf data as the corn leaf data is collected, forms a comparison set, and tracks the changes of the current corn leaf;

[0111] A convolutional neural network prediction module, after establishing the comparison set, compares and analyzes the adjacent two monitoring data in the comparison set, and inputs the data after comparison and analysis into a pre-trained convolutional neural network to predict the changes of the current corn leaf;

[0112] A change situation classification module divides the change situation of the current corn leaf into two categories based on the prediction result of the convolutional neural network: rapid change and regular change;

[0113] A regular change monitoring module, if the change of the current corn leaf is a regular change, continues to collect according to the initial monitoring frequency;

[0114] A rapid change frequency adjustment module, if it is detected that the change of the current corn leaf is a rapid change, automatically adjusts the actual monitoring frequency according to the change situation of the corn leaf, specifically by greatly increasing the collection frequency to ensure that all key data in the change process can be captured;

[0115] A method for evaluating the waterlogging tolerance of corn varieties based on data analysis provided by an embodiment of the present invention is implemented through the above-mentioned system for evaluating the waterlogging tolerance of corn varieties based on data analysis. The specific methods and processes of the system for evaluating the waterlogging tolerance of corn varieties based on data analysis are detailed in the embodiments of the above-mentioned method for evaluating the waterlogging tolerance of corn varieties based on data analysis, and will not be elaborated here.

[0116] The above formulas are all dimensionless and take their numerical values for calculation. The formula is obtained by collecting a large amount of data for software simulation to obtain a formula closest to the actual situation. The preset parameters in the formula are set by those skilled in the art according to the actual situation.

[0117] The above is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present application, and all should be covered by the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

[0118] Only some exemplary embodiments of the present invention have been described by way of illustration. Undoubtedly, for those of ordinary skill in the art, the described embodiments can be modified in various different ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and description are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A method for evaluating the waterlogging tolerance of corn varieties based on data analysis, characterized in that, It includes the following steps: According to the existing monitoring requirements, set an initial monitoring frequency for corn leaf information collection; After setting the initial monitoring frequency, regularly collect corn leaf information according to the initial monitoring frequency; As the collection of corn leaf data progresses, compare the currently collected corn leaf data with the previously collected corn leaf data to form a comparison set and track the changes in the current corn leaves; After establishing the comparison set, conduct a comparative analysis of two adjacent monitoring data in the comparison set, and input the data after the comparative analysis into a pre-trained convolutional neural network to predict the changes in the current corn leaves; Based on the prediction results of the convolutional neural network, divide the changes in the current corn leaves into two categories: rapid changes and regular changes; If the change in the current corn leaves is a regular change, continue to collect according to the initial monitoring frequency; If it is detected that the change in the current corn leaves is a rapid change, automatically adjust the actual monitoring frequency according to the change in the corn leaves. Specifically, significantly increase the collection frequency to ensure that all key data during the change process can be captured; After establishing the comparison set, conduct a comparative analysis of two adjacent monitoring data in the comparison set, extract the changes in the degree of edge drying of the corn leaves and the changes in the chlorophyll content in the corn leaves. After conducting a comparative analysis of the changes in the degree of edge drying of the corn leaves and the changes in the chlorophyll content in the corn leaves, generate a leaf edge drying index and a chlorophyll degradation index respectively. Measure the changes in the degree of edge drying of the corn leaves through the leaf edge drying index, reflecting the damage of the leaves under environmental stress conditions. Measure the changes in the chlorophyll content in the corn leaves through the chlorophyll degradation index, reflecting the decline in photosynthesis efficiency and the deterioration of leaf health; The specific steps for conducting a comparative analysis of two adjacent monitoring data in the comparison set, extracting the changes in the degree of edge drying of the corn leaves, and generating a leaf edge drying index after conducting a comparative analysis of the changes in the degree of edge drying of the corn leaves are as follows: Based on the data of two adjacent monitors, process the corn leaf image, extract the change information of the corn leaf edge, generate the leaf edge contour at each monitoring time point through an image processing algorithm, and then calculate the change rate of the dried part of the corn leaf edge between two adjacent monitors. The calculation formula is: , where ΔE t represents the edge drying rate, C t and C t-1 are the edge contour areas of the corn leaf for the current monitoring and the previous monitoring respectively, A t is the overall area of the corn leaf, to avoid interference of the overall size of the corn leaf on the calculation result; On the basis of extracting the dry area, further analyze the changes in the details of the edges of corn leaves. Take the local curvature of the edges of corn leaves and the refinement degree of the dry edges as key parameters, and describe the severity of the dryness of the edges of corn leaves through the point-by-point curvature change to quantify the degree of change in the details of the dry edges. The specific quantification expression is: In the formula, K t is the total amount of curvature change of the dry edge, and κ t (i) and κ t-1 (i) are the curvatures of the i-th edge point in the current and previous monitoring respectively, and n is the total number of edge points; The final leaf edge drying index is generated by introducing the edge drying change rate and the total curvature change of the drying edge, comprehensively quantifying the degree of leaf edge drying of corn leaves. The expression for generating the leaf edge drying index is: EDI t = α·ΔE t + β·(K t ) γ , where EDI t is the leaf edge drying index, α and β are the weight coefficients of the edge drying change rate and the total curvature change respectively, and γ is a regulation factor to enhance the sensitivity to curvature changes.

2. The method for evaluating the waterlogging tolerance of corn varieties based on data analysis according to claim 1, wherein, After obtaining the leaf edge drying index and the chlorophyll degradation index generated by conducting a comparative analysis of two adjacent monitoring data in the comparison set, input the leaf edge drying index and the chlorophyll degradation index into a pre-trained convolutional neural network to generate a change trend evaluation coefficient, and predict the changes in the current corn leaves through the change trend evaluation coefficient.

3. The method for evaluating waterlogging tolerance of maize varieties based on data analysis according to claim 2, characterized in that, Conduct a comparative analysis of the change trend evaluation coefficient generated by predicting two adjacent monitoring data in the comparison set through the convolutional neural network and the pre-set change trend evaluation coefficient reference threshold, and divide the changes in the current corn leaves. The specific division steps are as follows: If the change trend evaluation coefficient is greater than or equal to the reference threshold of the change trend evaluation coefficient, the change situation of the current corn leaf is classified as a rapid change, which indicates that the current corn leaf has experienced drastic physiological changes in a short period of time. If the change trend evaluation coefficient is less than the reference threshold of the change trend evaluation coefficient, the change situation of the current corn leaf is classified as a normal change, which means that the growth state of the current corn leaf is within the normal range, and its physiological characteristics gradually change with the growth cycle.

4. The method for evaluating the waterlogging tolerance of corn varieties based on data analysis according to claim 1, wherein, The specific steps for comparing and analyzing the adjacent two monitoring data in the comparison set, extracting the change in chlorophyll content in the corn leaf, and generating the chlorophyll degradation index after comparing and analyzing the change in chlorophyll content in the corn leaf are as follows: Extract the chlorophyll content from the maize leaf data collected twice adjacent to each other. Let the chlorophyll content at the time of the t1-th collection be and the chlorophyll content at the time of the t2-th collection be Calculate the change in chlorophyll between the two monitors. The calculation formula is as follows: In the formula, ΔR represents the change in chlorophyll content during the two monitors; An acceleration coefficient is introduced to reflect the acceleration of the change rate of chlorophyll content, and the change amount of chlorophyll content is amplified through a non-linear formula. The specific amplification expression is: where λ is the chlorophyll degradation acceleration coefficient, which measures the acceleration effect of chlorophyll loss, and τ 12 is the time interval between two adjacent monitoring times, and k is the adjustment index, which is used to amplify or reduce the acceleration degree of the change; By combining the change amount of chlorophyll content with the chlorophyll degradation acceleration coefficient, the chlorophyll degradation index is generated. The specific generated expression is: In the formula, δ chlor is the chlorophyll degradation index, and the final δ chlor value is used to quantify the degree of chlorophyll degradation.

5. The method for evaluating waterlogging tolerance of corn varieties based on data analysis according to claim 3, characterized in that, Automatically adjust the actual monitoring frequency according to the change situation of the corn leaf. The specific steps are as follows: Calculate the deviation between the change trend evaluation coefficient and the reference threshold of the change trend evaluation coefficient to obtain the change trend deviation coefficient Δ δ , Δ δ = Chang trend - δ threshold , where Chang trend is the change trend evaluation coefficient generated based on the convolutional neural network, reflecting the degree of change of the corn leaves, and δ threshold is the preset reference threshold of the change trend evaluation coefficient, used to distinguish rapid changes and regular changes; According to the change trend deviation coefficient Δ δ , calculate the monitoring frequency adjustment factor, and use a non-linear function to enhance the response to rapid changes. The calculation expression of the monitoring frequency adjustment factor is: In the formula, φ is the monitoring frequency adjustment factor, which is used to adjust the actual monitoring frequency, and w is the sensitivity coefficient, which controls the growth rate of the monitoring frequency adjustment factor; Using the initial monitoring frequency and the monitoring frequency adjustment factor, calculate the adjusted actual monitoring frequency. The calculation expression is: f actual = f initial ·φ, where f initial is the initial monitoring frequency, f actual is the adjusted actual monitoring frequency, which will be used for subsequent data collection; To avoid excessive consumption of system resources caused by too high monitoring frequency, an upper limit of the actual monitoring frequency is set, and the actual monitoring frequency is adjusted. The expression of the adjusted actual monitoring frequency is: f actual = min(f actual , f max ), where f max is the maximum allowable value of the actual monitoring frequency, and the min(·) function represents taking the smaller of the two values in the parentheses, ensuring that the actual monitoring frequency f actual does not exceed the maximum allowable value f max of the actual monitoring frequency; According to the actually calculated monitoring frequency f actual , update the monitoring plan of the data acquisition system to perform high-frequency acquisition according to the new frequency, ensuring that key data during the change process of corn leaves is captured. At the same time, continuously monitor the change trend evaluation coefficient. If the subsequent change trend evaluation coefficient drops below the reference threshold of the change trend evaluation coefficient, then restore to the initial monitoring frequency.

6. A waterlogging tolerance evaluation system for corn varieties based on data analysis, which is used to implement the waterlogging tolerance evaluation method for corn varieties based on data analysis described in any one of the above claims 1-5, and is characterized in that, It includes an initial monitoring frequency setting module, a regular data collection module, a data comparison set generation module, a convolutional neural network prediction module, a change situation classification module, a normal change monitoring module, and a rapid change frequency adjustment module. The initial monitoring frequency setting module sets an initial monitoring frequency for the collection of corn leaf information according to the existing monitoring requirements. The regular data collection module regularly collects corn leaf information according to the initial monitoring frequency after setting the initial monitoring frequency. The data comparison set generation module, as the corn leaf data is collected, compares the currently collected corn leaf data with the previously collected corn leaf data to form a comparison set and track the change situation of the current corn leaf. The convolutional neural network prediction module, after establishing the comparison set, compares and analyzes the adjacent two monitoring data in the comparison set, and inputs the data after comparison and analysis into a pre-trained convolutional neural network to predict the change situation of the current corn leaf. The change situation classification module, based on the prediction result of the convolutional neural network, classifies the change situation of the current corn leaf into two categories: rapid change and normal change. The normal change monitoring module, if the change of the current corn leaf is a normal change, continues to collect data according to the initial monitoring frequency. The rapid change frequency adjustment module, if it detects that the change of the current corn leaf is a rapid change, automatically adjusts the actual monitoring frequency according to the change situation of the corn leaf, specifically by significantly increasing the collection frequency to ensure that all key data during the change process can be captured.

Images