Sorghum waterlogging monitoring and evaluating method and system

Abstract

The invention relates to the technical field of sorghum waterlogging monitoring, in particular to a sorghum waterlogging monitoring and evaluating method and system, and the method comprises the following steps: obtaining the influence index of waterlogging on sorghum; determining the leaf temperature difference as a sorghum waterlogging monitoring index through the correlation between the leaf temperature difference and the influence index; and performing deep learning training on the correlation between the leaf temperature difference and the influence index, and constructing a sorghum waterlogging evaluation model for quantifying the influence degree of waterlogging on sorghum according to the leaf temperature difference. According to the method, the leaf temperature difference of an important sorghum waterlogging monitoring index is monitored by using an infrared imaging technology, rapid and non-destructive monitoring of sorghum waterlogging is realized, deep learning training is performed on the correlation between the leaf temperature difference and an influence index, a sorghum waterlogging evaluation model is constructed, and the influence degree of waterlogging on sorghum is quantified according to the leaf temperature difference, so that the sorghum waterlogging evaluation accuracy is improved. Quantitative evaluation of sorghum waterlogging is realized, and accurate monitoring of sorghum waterlogging is realized.

Description

Technical Field

[0001] The invention relates to the technical field of sorghum waterlogging monitoring, and in particular to a sorghum waterlogging monitoring and evaluation method and system. Background Art

[0002] Sorghum waterlogging will affect the moisture content and physiological processes of sorghum plants. The sorghum waterlogging situation can be determined by testing the data of sorghum moisture content and physiological processes, thereby realizing the monitoring of sorghum waterlogging situation.

[0003] Existing technologies require the detection of moisture content and physiological processes one by one when monitoring waterlogging. The monitoring process involves a large number of biological treatment technologies, which are complicated and time-consuming. As a result, the monitoring results of sorghum waterlogging lag behind the occurrence stage of sorghum waterlogging, making it difficult to monitor the waterlogging at the first time it occurs, which also leads to the inability to provide timely rescue and treatment for sorghum waterlogging. In addition, the biological treatment technology during the monitoring process will cause damage to the sorghum plants, which is not safe enough. Summary of the invention

[0004] The purpose of the present invention is to provide a monitoring and evaluation method and system for sorghum waterlogging to solve the technical problems in the prior art that the moisture content and physiological processes need to be detected one by one, the monitoring process involves a large number of biological treatment technologies, the process is complicated and time-consuming, and it may also cause damage to sorghum plants and lack safety.

[0005] In order to solve the above technical problems, the present invention specifically provides the following technical solutions:

[0006] A monitoring and evaluation method for sorghum waterlogging comprises the following steps:

[0007] Obtain indicators of the impact of waterlogging on sorghum;

[0008] Through the correlation between the leaf temperature difference and the influencing index, the leaf temperature difference is determined as a monitoring index of sorghum waterlogging;

[0009] The correlation between leaf temperature difference and influencing indicators was trained through deep learning, and a sorghum waterlogging assessment model was constructed to quantify the impact of waterlogging on sorghum based on leaf temperature difference.

[0010] As a preferred embodiment of the present invention, the influencing indicators include: relative water content, net photosynthetic rate, and transpiration rate of sorghum leaves.

[0011] As a preferred solution of the present invention, a method for determining the correlation between the leaf temperature difference and the influencing index includes:

[0012] The experiment collected the relative water content, net photosynthetic rate and transpiration rate of sorghum leaves after waterlogging, as well as the leaf temperature difference of sorghum leaves;

[0013] Linear fitting was performed on the relative water content, net photosynthetic rate and transpiration rate of sorghum leaves and the leaf temperature difference of sorghum leaves, and the linear correlation between the relative water content, net photosynthetic rate and transpiration rate of sorghum leaves and the leaf temperature difference of sorghum leaves was determined;

[0014] Among them, the linear correlation between the relative water content, net photosynthetic rate and transpiration rate of the sorghum leaves and the leaf temperature difference of the sorghum leaves is: there is a significant linear negative correlation between the relative water content, net photosynthetic rate, transpiration rate and leaf temperature difference of the sorghum leaves.

[0015] As a preferred embodiment of the present invention, the experimental collection method of the relative water content of sorghum leaves comprises:

[0016] Prepare 50 pots of sorghum plants and apply waterlogging stress at the five-leaf stage of the sorghum plants;

[0017] After two weeks of waterlogging stress, 10 pots of sorghum plants with uniform growth were selected as the test objects. Samples were taken from the latest fully expanded leaves using a borer with a diameter of 8 mm, avoiding the leaf veins, and 30 samples were taken from each leaf;

[0018] 50 leaves were randomly selected from the sample and weighed to obtain the fresh weight (FW) of sorghum leaves;

[0019] 50 samples were soaked in distilled water for 5 h, the surface water was absorbed with absorbent paper, and the saturated weight TW of sorghum leaves was obtained by weighing;

[0020] 50 samples were dried and weighed to obtain the dry weight DW of sorghum leaves;

[0021] According to the saturated weight TW of sorghum leaves and the dry weight DW of sorghum leaves, the relative water content of sorghum leaves was obtained using the calculation formula of relative water content;

[0022] The relative water content is calculated as follows: RWC = (FW-DW) / (TW-DW) × 100%;

[0023] Where RWC is the relative water content, FW is the fresh weight of sorghum leaves, DW is the dry weight of sorghum leaves, and TW is the saturated weight of sorghum leaves.

[0024] As a preferred embodiment of the present invention, the method for collecting the net photosynthetic rate and transpiration rate of sorghum leaves comprises:

[0025] Between 9:00 and 11:00 in the morning, the gas exchange parameters of the latest fully expanded leaves of 50 pots of sorghum plants were measured using a LI-6400 photosynthetic instrument to obtain the net photosynthetic rate Pn and transpiration rate E of the sorghum leaves.

[0026] As a preferred embodiment of the present invention, the leaf temperature difference experimental collection method of sorghum leaves comprises:

[0027] The leaf temperature of 50 pots of sorghum plants was measured using the ThermaCAMSC3000 infrared imager produced by FLIR Corporation of the United States.

[0028] After the waterlogging stress was over, 5 pots of sorghum plants were selected from 50 pots and infrared images of sorghum were taken from 9:00 to 11:30 in the morning. The infrared imager was placed 1 meter above the sorghum leaves to take infrared images of the sorghum, and the air temperature T at the time of shooting was recorded. air ;

[0029] The sorghum leaf temperature was analyzed using FLIR thermal imaging software, and the average temperature of the center point of the three leaves at the top of the sorghum was taken as the sorghum leaf temperature T leaf ;

[0030] According to the air temperature T air and sorghum leaf temperature T leaf , using the leaf temperature difference calculation formula, the leaf temperature difference △T of sorghum leaves was obtained;

[0031] The calculation formula of leaf temperature difference is: △T=T leaf -T air ;

[0032] Where △T is the leaf temperature difference of sorghum leaves, T leaf is the leaf temperature of sorghum, T air is the air temperature.

[0033] As a preferred embodiment of the present invention, the method for constructing the sorghum waterlogging damage assessment model comprises:

[0034] The leaf temperature difference of sorghum leaves is used as the input item of the first BP neural network, and the relative water content, net photosynthetic rate and transpiration rate of sorghum leaves are used as the output items of the first BP neural network;

[0035] The relative water content of sorghum leaves is used as the input item of the second BP neural network, and the leaf temperature difference of sorghum leaves is used as the output item of the second BP neural network;

[0036] The net photosynthetic rate of sorghum leaves is used as the input item of the third BP neural network, and the leaf temperature difference of sorghum leaves is used as the output item of the third BP neural network;

[0037] The transpiration rate of sorghum leaves is used as the input item of the fourth BP neural network, and the leaf temperature difference of sorghum leaves is used as the output item of the fourth BP neural network;

[0038] Constructing a consistency loss function for the first BP neural network, the second BP neural network, the third BP neural network and the fourth BP neural network;

[0039] Based on the consistency loss function, the first BP neural network, the second BP neural network, the third BP neural network and the fourth BP neural network are trained to the optimum.

[0040] Using the first BP neural network trained to the optimum as the sorghum waterlogging assessment model;

[0041] Among them, the first BP neural network, the second BP neural network, the third BP neural network and the fourth BP neural network are respectively:

[0042] (RWC1, Pn1, E1) = BP1 (ΔT);

[0043] △T2=BP2(RWC);

[0044] △T3=BP3(Pn);

[0045] △T4=BP4(E);

[0046] Wherein, RWC1, Pn1, E1 are the relative water content, net photosynthetic rate, and transpiration rate output by the first BP neural network, respectively; △T is the leaf temperature difference input by the first BP neural network; △T2 is the leaf temperature difference output by the second BP neural network; RWC is the relative water content input by the second BP neural network; △T3 is the leaf temperature difference output by the third BP neural network; Pn is the net photosynthetic rate input by the second BP neural network; △T4 is the leaf temperature difference output by the fourth BP neural network; and E is the transpiration rate input by the fourth BP neural network.

[0047] As a preferred solution of the present invention, the consistency loss function is:

[0048] Loss1=||RWC1-RWC||2+||Pn1-Pn||2+||E1-E||2;

[0049] Loss2=||△T2-△T||2+||△T3-△T||2+||△T4-△T||2

[0050] Loss3=||BP1(△T2)-RWC||2+||BP1(△T3)-Pn||2+||BP1(△T4)-E||2;

[0051] Loss4=||BP2(RWC1)-△T||2+||BP3(Pn1)-△T||2+||BP4(E1)-△T||2;

[0052] Wherein, Loss1 is the loss function of the first BP neural network, Loss2 is the loss function of the second BP neural network, the third BP neural network and the fourth BP neural network, Loss3 is the conversion consistency loss function of relative water content, net photosynthetic rate and transpiration rate, and Loss4 is the conversion consistency loss function of leaf temperature difference.

[0053] As a preferred embodiment of the present invention, a method for monitoring and evaluating waterlogging of sorghum plants to be monitored using an infrared imager in combination with a sorghum waterlogging evaluation model includes:

[0054] The sorghum plants to be monitored were monitored between 9:00 and 11:30 in the morning. The infrared imager was placed 1 meter above the sorghum leaves to take infrared images of the sorghum, and the air temperature T at the time of shooting was recorded. air ;

[0055] The FLIR thermal imaging software was used to analyze the leaf temperature of the sorghum plant to be monitored, and the average temperature of the center point of the three leaves on the top of the sorghum was taken as the leaf temperature T of the sorghum plant to be monitored. leaf ;

[0056] According to the air temperature T air and the leaf temperature T of the sorghum plant to be monitored leaf , using the leaf temperature difference calculation formula, the leaf temperature difference △T of the sorghum leaves of the sorghum plants to be monitored is obtained;

[0057] The leaf temperature difference △T of the sorghum leaves of the sorghum plants to be monitored is input into the sorghum waterlogging assessment model to obtain the relative water content, net photosynthetic rate and transpiration rate of the sorghum leaves of the sorghum plants to be monitored.

[0058] As a preferred embodiment of the present invention, the present invention provides a monitoring and evaluation system for sorghum waterlogging. According to the monitoring and evaluation method for sorghum waterlogging, the system comprises:

[0059] Data collection unit to obtain indicators of the impact of waterlogging on sorghum;

[0060] A data processing unit, used for determining the leaf temperature difference as a monitoring index of sorghum waterlogging through the correlation between the leaf temperature difference and the influencing index;

[0061] And the correlation between leaf temperature difference and influencing indicators was trained through deep learning, and a sorghum waterlogging assessment model was constructed to quantify the impact of waterlogging on sorghum based on leaf temperature difference;

[0062] The monitoring and evaluation unit is used to monitor and evaluate the waterlogging damage of the monitored sorghum plants using an infrared imager combined with a sorghum waterlogging damage evaluation model.

[0063] Compared with the prior art, the present invention has the following beneficial effects:

[0064] The present invention uses infrared imaging technology to monitor the leaf temperature difference, an important indicator for monitoring sorghum waterlogging, to achieve rapid and damage-free monitoring of sorghum waterlogging. In addition, the correlation between the leaf temperature difference and the influencing indicators is trained through deep learning to construct a sorghum waterlogging assessment model. The impact of waterlogging on sorghum is quantified according to the leaf temperature difference, thereby achieving quantitative assessment of sorghum waterlogging and accurate monitoring of sorghum waterlogging. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] In order to more clearly illustrate the implementation methods of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for the implementation methods or the description of the prior art. Obviously, the drawings in the following description are only exemplary, and for ordinary technicians in this field, other implementation drawings can be derived from the provided drawings without creative work.

[0066] Figure 1 A flow chart of a monitoring and evaluation method for sorghum waterlogging provided by an embodiment of the present invention;

[0067] Figure 2 A block diagram of a monitoring and evaluation system for sorghum waterlogging provided by an embodiment of the present invention;

[0068] Figure 3 A correlation analysis diagram between relative water content and leaf temperature difference provided by an embodiment of the present invention;

[0069] Figure 4 A correlation analysis diagram between transpiration rate and leaf temperature difference provided in an embodiment of the present invention;

[0070] Figure 5 A correlation analysis diagram between the net photosynthetic rate and the leaf temperature difference provided in an embodiment of the present invention;

[0071] Figure 6 This is an infrared image of sorghum provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0072] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions 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.

[0073] like Figure 1 As shown, the present invention provides a monitoring and evaluation method for sorghum waterlogging, comprising the following steps:

[0074] Obtain indicators of the impact of waterlogging on sorghum;

[0075] Through the correlation between leaf temperature difference and influencing indexes, leaf temperature difference was determined as a monitoring index for sorghum waterlogging.

[0076] The correlation between leaf temperature difference and influencing indicators was trained through deep learning, and a sorghum waterlogging assessment model was constructed to quantify the impact of waterlogging on sorghum based on leaf temperature difference.

[0077] The present invention explores the correlation between the leaf temperature difference of sorghum leaves and the relative water content, net photosynthetic rate and transpiration rate of the sorghum leaves. Since sorghum waterlogging can be monitored by the relative water content, net photosynthetic rate and transpiration rate of the sorghum leaves, sorghum waterlogging can be monitored by the leaf temperature difference of the sorghum leaves.

[0078] The present invention converts the monitoring of sorghum waterlogging from relative water content, net photosynthetic rate and transpiration rate into leaf temperature difference, thereby realizing the conversion from a complicated and time-consuming biological treatment process to a convenient and time-saving infrared imaging technology, thereby realizing rapid monitoring of sorghum waterlogging. At the same time, the infrared imaging technology will not damage the sorghum plants during the monitoring process, thereby realizing damage-free monitoring of sorghum waterlogged plants.

[0079] The present invention explores the correlation between the leaf temperature difference of sorghum leaves and the relative water content, net photosynthetic rate and transpiration rate of the sorghum leaves. The data of leaf temperature difference and relative water content, net photosynthetic rate and transpiration rate are collected through potted plant experiments, and the collected data are analyzed and processed (linear fitting) to determine that the relative water content, net photosynthetic rate, transpiration rate and leaf temperature difference of sorghum leaves are significantly linearly negatively correlated.

[0080] The present invention monitors sorghum waterlogging through infrared imaging technology, and realizes qualitative monitoring of sorghum waterlogging through leaf temperature difference quickly and non-destructively. Although the correlation between leaf temperature difference and influencing index can be linearly fitted, the fitting is an approximate processing, and the accuracy is difficult to guarantee. Therefore, it cannot be used for quantitative measurement (i.e., the influencing index is calculated by leaf temperature difference), but it is sufficient to show the correlation characteristics. Therefore, the present invention adopts linear fitting to analyze the correlation between leaf temperature difference and influencing index. On the basis of qualitative monitoring of sorghum waterlogging, a sorghum waterlogging assessment model is further constructed. The influence of waterlogging on sorghum can be quantified through monitoring of leaf temperature difference, thereby realizing quantitative monitoring of sorghum waterlogging, and comprehensively realizing rapid, accurate and non-destructive monitoring of sorghum waterlogging.

[0081] When constructing a sorghum waterlogging assessment model, the present invention performs forward measurement training and reverse measurement training through multiple neural network structures in order to ensure accurate measurement of the impact of leaf temperature difference on sorghum waterlogging, that is, to ensure the accuracy of the model. The first BP neural network implements a forward measurement process of calculating relative water content, net photosynthetic rate, and transpiration rate from leaf temperature difference, and the second BP neural network to the fourth BP neural network implement a reverse measurement process of calculating leaf temperature difference from relative water content, net photosynthetic rate, and transpiration rate. The correlation between leaf temperature difference and relative water content, net photosynthetic rate, and transpiration rate is bidirectionally trained, and the bidirectional mapping of the correlation is mutually limited, so as to avoid the randomness of data mapping caused by unidirectional training and improve the accuracy of the model.

[0082] The present invention mutually limits the relative water content, net photosynthetic rate and transpiration rate calculated by the leaf temperature difference through bidirectional mapping of correlations, so that the relative water content, net photosynthetic rate and transpiration rate can also be restored to the measured leaf temperature difference, and the relative water content, net photosynthetic rate and transpiration rate can also be restored to the relative water content, net photosynthetic rate and transpiration rate through the leaf temperature difference calculated by the relative water content, net photosynthetic rate and transpiration rate, thereby ensuring the consistency of the correlation between the leaf temperature difference and the relative water content, net photosynthetic rate and transpiration rate in forward calculation and reverse calculation, that is, the model maintains the only correlation between the leaf temperature difference and the relative water content, net photosynthetic rate and transpiration rate during the calculation process, thereby ensuring the accurate performance of the model.

[0083] In order to maintain the consistency of the correlation between leaf temperature difference and relative water content, net photosynthetic rate and transpiration rate, the present invention uses the conversion consistency loss function of relative water content, net photosynthetic rate and transpiration rate, and the conversion consistency loss function of leaf temperature difference when training the neural network structure, to ensure that the relative water content, net photosynthetic rate and transpiration rate calculated by leaf temperature difference can also be restored to the calculated leaf temperature difference, and that the leaf temperature difference calculated by the relative water content, net photosynthetic rate and transpiration rate can also be restored to the relative water content, net photosynthetic rate and transpiration rate.

[0084] Similarly, the present invention also executes the forward measurement training process and the reverse measurement training process through the loss function of the network structure in the forward measurement training process (the loss between the predicted value and the true value), and the loss function of the network structure in the reverse measurement training process (the loss between the predicted value and the true value), so that the relative water content, net photosynthetic rate, and transpiration rate calculated by the leaf temperature difference are close to the true value, and the accuracy of the forward measurement is guaranteed, and the leaf temperature difference calculated by the relative water content, net photosynthetic rate, and transpiration rate are close to the true value, and the accuracy of the reverse measurement is guaranteed, thereby ensuring that the correlation between the leaf temperature difference and the relative water content, net photosynthetic rate, and transpiration rate in the forward measurement and reverse measurement is accurately mapped, and the combination ensures the consistency of the correlation between the leaf temperature difference and the relative water content, net photosynthetic rate, and transpiration rate in the forward measurement and reverse measurement, so that the model maintains the only accurate correlation between the leaf temperature difference and the relative water content, net photosynthetic rate, and transpiration rate during the measurement process, thereby further ensuring the accurate performance of the model.

[0085] After constructing the sorghum waterlogging assessment model, we used infrared imaging technology to obtain sorghum monitoring indicators, and combined the sorghum waterlogging assessment model to assess the degree of sorghum waterlogging, thus achieving real-time monitoring and accurate assessment of sorghum waterlogging.

[0086] The influencing indicators include: relative water content of sorghum leaves, net photosynthetic rate, and transpiration rate.

[0087] The present invention explores the correlation between the leaf temperature difference of sorghum leaves and the relative water content, net photosynthetic rate and transpiration rate of sorghum leaves. The data of leaf temperature difference and relative water content, net photosynthetic rate and transpiration rate are collected by potted plant experiment. The collected data are analyzed and processed (linear fitting). It is determined that the relative water content, net photosynthetic rate, transpiration rate and leaf temperature difference of sorghum leaves are significantly linearly negatively correlated, as follows:

[0088] Methods for determining the correlation between leaf temperature difference and influencing indexes include:

[0089] The experiment collected the relative water content, net photosynthetic rate and transpiration rate of sorghum leaves after waterlogging, as well as the leaf temperature difference of sorghum leaves;

[0090] Linear fitting was performed on the relative water content, net photosynthetic rate and transpiration rate of sorghum leaves and the leaf temperature difference of sorghum leaves, and the linear correlation between the relative water content, net photosynthetic rate and transpiration rate of sorghum leaves and the leaf temperature difference of sorghum leaves was determined;

[0091] Among them, the linear correlation between the relative water content, net photosynthetic rate and transpiration rate of sorghum leaves and the leaf temperature difference of sorghum leaves is: there is a significant linear negative correlation between the relative water content, net photosynthetic rate, transpiration rate and leaf temperature difference of sorghum leaves.

[0092] There was a significant linear correlation between the relative water content, net photosynthetic rate, transpiration rate and leaf temperature difference of sorghum leaves. Figure 3 There is a very significant negative correlation between relative water content and leaf temperature difference, and the fitting equation is y = -93.28x + 89.95 (R 2 =0.8355, P<0.01). Figure 4 As shown in Figure 2, there is a very significant negative correlation between transpiration rate and leaf temperature difference, and the fitting equation is y = -0.24x + 5.62 (R 2 =0.7844, P<0.01). Figure 5 As shown in Figure 2, there is a very significant negative correlation between the net photosynthetic rate and the leaf temperature difference, and the fitting equation is y = -1.32x + 25.00 (R 2 =0.7886, P<0.01). Correlation analysis showed that leaf temperature difference was significantly correlated with leaf water content, transpiration rate, and net photosynthetic rate. Leaf temperature difference can reflect the impact of waterlogging on sorghum leaf water content and gas exchange parameters (net photosynthetic rate and transpiration rate).

[0093] The experimental method for collecting relative water content of sorghum leaves includes:

[0094] Prepare 50 pots of sorghum plants and apply waterlogging stress at the five-leaf stage of the sorghum plants;

[0095] After two weeks of waterlogging stress, 10 pots of sorghum plants with uniform growth were selected as the test objects. Samples were taken from the latest fully expanded leaves using a borer with a diameter of 8 mm, avoiding the leaf veins, and 30 samples were taken from each leaf;

[0096] 50 leaves were randomly selected from the sample and weighed to obtain the fresh weight (FW) of sorghum leaves;

[0097] Soak 50 samples in distilled water for 5 hours, absorb the surface water with absorbent paper, and weigh them to obtain the saturated weight TW of sorghum leaves;

[0098] 50 samples were dried and weighed to obtain the dry weight DW of sorghum leaves;

[0099] According to the saturated weight TW of sorghum leaves and the dry weight DW of sorghum leaves, the relative water content of sorghum leaves was obtained using the calculation formula of relative water content;

[0100] The relative water content is calculated as follows: RWC = (FW-DW) / (TW-DW) × 100%;

[0101] Where RWC is the relative water content, FW is the fresh weight of sorghum leaves, DW is the dry weight of sorghum leaves, and TW is the saturated weight of sorghum leaves.

[0102] The experimental collection methods of the net photosynthetic rate and transpiration rate of sorghum leaves include:

[0103] Between 9:00 and 11:00 in the morning, the gas exchange parameters of the latest fully expanded leaves of 50 pots of sorghum plants were measured using a LI-6400 photosynthetic instrument to obtain the net photosynthetic rate Pn and transpiration rate E of the sorghum leaves.

[0104] The leaf temperature difference experiment collection method for sorghum leaves includes:

[0105] The leaf temperature of 50 pots of sorghum plants was measured using the ThermaCAMSC3000 infrared imager produced by FLIR Corporation of the United States.

[0106] After the waterlogging stress was over, 5 pots of sorghum plants were selected from 50 pots and infrared images of sorghum were taken from 9:00 to 11:30 in the morning. The infrared imager was placed 1 meter above the sorghum leaves to take infrared images of sorghum. Figure 6 As shown, the air temperature T at the time of shooting is recorded air ;

[0107] The sorghum leaf temperature was analyzed using FLIR thermal imaging software, and the average temperature of the center point of the three leaves at the top of the sorghum was taken as the sorghum leaf temperature T leaf ;

[0108] According to the air temperature T air and sorghum leaf temperature T leaf , using the leaf temperature difference calculation formula, the leaf temperature difference △T of sorghum leaves was obtained;

[0109] The calculation formula of leaf temperature difference is: △T=T leaf -T air ;

[0110] Where △T is the leaf temperature difference of sorghum leaves, T leaf is the leaf temperature of sorghum, T air is the air temperature.

[0111] The present invention monitors sorghum waterlogging through infrared imaging technology, and realizes qualitative monitoring of sorghum waterlogging through leaf temperature difference quickly and non-destructively. On the basis of qualitative monitoring of sorghum waterlogging, a sorghum waterlogging assessment model is constructed, which can quantify the impact of waterlogging on sorghum through monitoring leaf temperature difference, thereby realizing quantitative monitoring of sorghum waterlogging and comprehensively realizing rapid, accurate and non-destructive monitoring of sorghum waterlogging.

[0112] The construction method of sorghum waterlogging assessment model includes:

[0113] The leaf temperature difference of sorghum leaves is used as the input item of the first BP neural network, and the relative water content, net photosynthetic rate and transpiration rate of sorghum leaves are used as the output items of the first BP neural network;

[0114] The relative water content of sorghum leaves is used as the input item of the second BP neural network, and the leaf temperature difference of sorghum leaves is used as the output item of the second BP neural network;

[0115] The net photosynthetic rate of sorghum leaves is used as the input item of the third BP neural network, and the leaf temperature difference of sorghum leaves is used as the output item of the third BP neural network;

[0116] The transpiration rate of sorghum leaves is used as the input item of the fourth BP neural network, and the leaf temperature difference of sorghum leaves is used as the output item of the fourth BP neural network;

[0117] Constructing a consistency loss function for the first BP neural network, the second BP neural network, the third BP neural network and the fourth BP neural network;

[0118] Based on the consistency loss function, the first BP neural network, the second BP neural network, the third BP neural network and the fourth BP neural network are trained to the optimal state.

[0119] The first BP neural network trained to the best is used as the sorghum waterlogging assessment model;

[0120] Among them, the first BP neural network, the second BP neural network, the third BP neural network and the fourth BP neural network are respectively:

[0121] (RWC1, Pn1, E1) = BP1 (ΔT);

[0122] △T2=BP2(RWC);

[0123] △T3=BP3(Pn);

[0124] △T4=BP4(E);

[0125] Wherein, RWC1, Pn1, E1 are the relative water content, net photosynthetic rate, and transpiration rate output by the first BP neural network, respectively; △T is the leaf temperature difference input by the first BP neural network; △T2 is the leaf temperature difference output by the second BP neural network; RWC is the relative water content input by the second BP neural network; △T3 is the leaf temperature difference output by the third BP neural network; Pn is the net photosynthetic rate input by the second BP neural network; △T4 is the leaf temperature difference output by the fourth BP neural network; and E is the transpiration rate input by the fourth BP neural network.

[0126] When constructing a sorghum waterlogging assessment model, the present invention performs forward measurement training and reverse measurement training through multiple neural network structures in order to ensure accurate measurement of the impact of leaf temperature difference on sorghum waterlogging, that is, to ensure the accuracy of the model. The first BP neural network implements a forward measurement process of calculating relative water content, net photosynthetic rate, and transpiration rate from leaf temperature difference, and the second BP neural network to the fourth BP neural network implement a reverse measurement process of calculating leaf temperature difference from relative water content, net photosynthetic rate, and transpiration rate. The correlation between leaf temperature difference and relative water content, net photosynthetic rate, and transpiration rate is bidirectionally trained, and the bidirectional mapping of the correlation is mutually limited, so as to avoid the randomness of data mapping caused by unidirectional training and improve the accuracy of the model.

[0127] The present invention mutually limits the relative water content, net photosynthetic rate and transpiration rate calculated by the leaf temperature difference through bidirectional mapping of correlations, so that the relative water content, net photosynthetic rate and transpiration rate can also be restored to the measured leaf temperature difference, and the relative water content, net photosynthetic rate and transpiration rate can also be restored to the relative water content, net photosynthetic rate and transpiration rate through the leaf temperature difference calculated by the relative water content, net photosynthetic rate and transpiration rate, thereby ensuring the consistency of the correlation between the leaf temperature difference and the relative water content, net photosynthetic rate and transpiration rate in forward calculation and reverse calculation, that is, the model maintains the only correlation between the leaf temperature difference and the relative water content, net photosynthetic rate and transpiration rate during the calculation process, thereby ensuring the accurate performance of the model.

[0128] The consistency loss function is:

[0129] Loss1=||RWC1-RWC||2+||Pn1-Pn||2+||E1-E||2;

[0130] Loss2=||△T2-△T||2+||△T3-△T||2+||△T4-△T||2

[0131] Loss3=||BP1(△T2)-RWC||2+||BP1(△T3)-Pn||2+||BP1(△T4)-E||2;

[0132] Loss4=||BP2(RWC1)-△T||2+||BP3(Pn1)-△T||2+||BP4(E1)-△T||2;

[0133] Wherein, Loss1 is the loss function of the first BP neural network, Loss2 is the loss function of the second BP neural network, the third BP neural network and the fourth BP neural network, Loss3 is the conversion consistency loss function of relative water content, net photosynthetic rate and transpiration rate, and Loss4 is the conversion consistency loss function of leaf temperature difference.

[0134] In order to maintain the consistency of the correlation between leaf temperature difference and relative water content, net photosynthetic rate and transpiration rate, the present invention uses the conversion consistency loss function of relative water content, net photosynthetic rate and transpiration rate, and the conversion consistency loss function of leaf temperature difference when training the neural network structure, to ensure that the relative water content, net photosynthetic rate and transpiration rate calculated by leaf temperature difference can also be restored to the calculated leaf temperature difference, and that the leaf temperature difference calculated by the relative water content, net photosynthetic rate and transpiration rate can also be restored to the relative water content, net photosynthetic rate and transpiration rate.

[0135] Similarly, the present invention also executes the forward measurement training process and the reverse measurement training process through the loss function of the network structure in the forward measurement training process (the loss between the predicted value and the true value), and the loss function of the network structure in the reverse measurement training process (the loss between the predicted value and the true value), so that the relative water content, net photosynthetic rate, and transpiration rate calculated by the leaf temperature difference are close to the true value, and the accuracy of the forward measurement is guaranteed, and the leaf temperature difference calculated by the relative water content, net photosynthetic rate, and transpiration rate are close to the true value, and the accuracy of the reverse measurement is guaranteed, thereby ensuring that the correlation between the leaf temperature difference and the relative water content, net photosynthetic rate, and transpiration rate in the forward measurement and reverse measurement is accurately mapped, and the combination ensures the consistency of the correlation between the leaf temperature difference and the relative water content, net photosynthetic rate, and transpiration rate in the forward measurement and reverse measurement, so that the model maintains the only accurate correlation between the leaf temperature difference and the relative water content, net photosynthetic rate, and transpiration rate during the measurement process, thereby further ensuring the accurate performance of the model.

[0136] After the sorghum waterlogging assessment model is built, the monitoring indicators of sorghum are obtained through infrared imaging technology, and the degree of sorghum waterlogging is assessed in combination with the sorghum waterlogging assessment model to achieve real-time monitoring and accurate assessment of sorghum waterlogging, as follows:

[0137] The monitoring and evaluation methods of waterlogging damage to sorghum plants using infrared imagers combined with sorghum waterlogging damage assessment models include:

[0138] The sorghum plants to be monitored were monitored between 9:00 and 11:30 in the morning. The infrared imager was placed 1 meter above the sorghum leaves to take infrared images of the sorghum, and the air temperature T at the time of shooting was recorded. air ;

[0139] The FLIR thermal imaging software was used to analyze the leaf temperature of the sorghum plant to be monitored, and the average temperature of the center point of the three leaves on the top of the sorghum was taken as the leaf temperature T of the sorghum plant to be monitored. leaf ;

[0140] According to the air temperature T air and the leaf temperature T of the sorghum plant to be monitored leaf, using the leaf temperature difference calculation formula, the leaf temperature difference △T of the sorghum leaves of the sorghum plants to be monitored is obtained;

[0141] The leaf temperature difference △T of the sorghum leaves of the sorghum plants to be monitored is input into the sorghum waterlogging assessment model to obtain the relative water content, net photosynthetic rate and transpiration rate of the sorghum leaves of the sorghum plants to be monitored.

[0142] BP (back propagation) neural network is a concept proposed by scientists led by Rumelhart and McClelland. It is a multi-layer feedforward neural network trained according to the error back propagation algorithm and is one of the most widely used neural network models.

[0143] BP neural network has the ability to classify patterns of any complexity and excellent multi-dimensional function mapping ability. It learns certain rules through its own training and obtains the result closest to the expected output value when the input value is given, that is, it determines the mapping relationship between input and output. The BP algorithm includes two processes: forward propagation of signals and reverse propagation of errors. That is, when calculating the error output, it is done from input to output, while adjusting the weights and thresholds from output to input. During forward propagation, the input signal acts on the output node through the hidden layer, and after nonlinear transformation, the output signal is generated. If the actual output does not match the expected output, the error reverse propagation process is entered. Error back propagation is to propagate the output error through the hidden layer to the input layer layer by layer, and distribute the error to all units in each layer, and use the error signal obtained from each layer as the basis for adjusting the weights of each unit. By adjusting the connection strength between the input node and the hidden layer node, the connection strength between the hidden layer node and the output node, and the threshold, the error is reduced along the gradient direction. After repeated learning and training, the network parameters (weights and thresholds) corresponding to the minimum error are determined, and the training stops. At this point, the trained neural network can process the input information of similar samples and obtain the output information by itself after nonlinear transformation with the smallest output error.

[0144] Utilizing the function mapping properties of the BP neural network, the present invention applies the BP neural network to the measurement and evaluation process, and trains the BP neural network to learn the mapping relationship between the leaf temperature difference and the relative water content, net photosynthetic rate, and transpiration rate of sorghum leaves.

[0145] like Figure 2 As shown, the present invention provides a monitoring and evaluation system for sorghum waterlogging, according to a monitoring and evaluation method for sorghum waterlogging, the system comprises:

[0146] Data collection unit to obtain indicators of the impact of waterlogging on sorghum;

[0147] A data processing unit, used for determining the leaf temperature difference as a monitoring index of sorghum waterlogging through the correlation between the leaf temperature difference and the influencing index;

[0148] And the correlation between leaf temperature difference and influencing indicators was trained through deep learning, and a sorghum waterlogging assessment model was constructed to quantify the impact of waterlogging on sorghum based on leaf temperature difference;

[0149] The monitoring and evaluation unit is used to monitor and evaluate the waterlogging damage of the monitored sorghum plants using an infrared imager combined with a sorghum waterlogging damage evaluation model.

[0150] The present invention uses infrared imaging technology to monitor the leaf temperature difference, an important indicator for monitoring sorghum waterlogging, to achieve rapid and damage-free monitoring of sorghum waterlogging. In addition, the correlation between the leaf temperature difference and the influencing indicators is trained through deep learning to construct a sorghum waterlogging assessment model. The impact of waterlogging on sorghum is quantified according to the leaf temperature difference, thereby achieving quantitative assessment of sorghum waterlogging and accurate monitoring of sorghum waterlogging.

[0151] The above embodiments are only exemplary embodiments of the present application and are not intended to limit the present application. The protection scope of the present application is defined by the claims. Those skilled in the art may make various modifications or equivalent substitutions to the present application within the essence and protection scope of the present application, and such modifications or equivalent substitutions shall also be deemed to fall within the protection scope of the present application.

Claims

1. A monitoring and evaluation method for sorghum waterlogging, characterized in that: The following steps are involved: Obtain indicators of the impact of waterlogging on sorghum; Through the correlation between the leaf temperature difference and the influencing index, the leaf temperature difference is determined as a monitoring index of sorghum waterlogging; The correlation between leaf temperature difference and influencing indicators was trained through deep learning, and a sorghum waterlogging assessment model was constructed to quantify the impact of waterlogging on sorghum based on leaf temperature difference.

2. The method for monitoring and evaluating sorghum waterlogging according to claim 1, characterized in that: The influencing indicators include: relative water content of sorghum leaves, net photosynthetic rate, and transpiration rate.

3. The monitoring and evaluation method for sorghum waterlogging according to claim 2, characterized in that: The method for determining the correlation between the leaf temperature difference and the influencing index includes: The experiment collected the relative water content, net photosynthetic rate and transpiration rate of sorghum leaves after waterlogging, as well as the leaf temperature difference of sorghum leaves; Linear fitting was performed on the relative water content, net photosynthetic rate and transpiration rate of sorghum leaves and the leaf temperature difference of sorghum leaves, and the linear correlation between the relative water content, net photosynthetic rate and transpiration rate of sorghum leaves and the leaf temperature difference of sorghum leaves was determined; Among them, the linear correlation between the relative water content, net photosynthetic rate and transpiration rate of the sorghum leaves and the leaf temperature difference of the sorghum leaves is: there is a significant linear negative correlation between the relative water content, net photosynthetic rate, transpiration rate and leaf temperature difference of the sorghum leaves.

4. The method for monitoring and evaluating sorghum waterlogging according to claim 3, characterized in that: The experimental collection method of the relative water content of sorghum leaves comprises: Prepare 50 pots of sorghum plants and apply waterlogging stress at the five-leaf stage of the sorghum plants; After two weeks of waterlogging stress, 10 pots of sorghum plants with uniform growth were selected as the test objects. Samples were taken from the latest fully expanded leaves using a hole puncher with a diameter of 8 mm, avoiding the leaf veins, and 30 samples were taken from each leaf; 50 leaves were randomly selected from the sample and weighed to obtain the fresh weight (FW) of sorghum leaves; 50 samples were soaked in distilled water for 5 h, the surface water was absorbed with absorbent paper, and the saturated weight TW of sorghum leaves was obtained by weighing; 50 samples were dried and weighed to obtain the dry weight DW of sorghum leaves; According to the saturated weight TW of sorghum leaves and the dry weight DW of sorghum leaves, the relative water content of sorghum leaves was obtained using the calculation formula of relative water content; The relative water content is calculated as follows: RWC = (FW-DW) / (TW-DW) × 100%; Where RWC is the relative water content, FW is the fresh weight of sorghum leaves, DW is the dry weight of sorghum leaves, and TW is the saturated weight of sorghum leaves.

5. The method for monitoring and evaluating sorghum waterlogging according to claim 3, characterized in that: The method for collecting the net photosynthetic rate and transpiration rate of sorghum leaves experimentally comprises: Between 9:00 and 11:00 in the morning, the gas exchange parameters of the latest fully expanded leaves of 50 pots of sorghum plants were measured using a LI-6400 photosynthetic instrument to obtain the net photosynthetic rate Pn and transpiration rate E of the sorghum leaves.

6. The method for monitoring and evaluating sorghum waterlogging according to claim 3, characterized in that: The leaf temperature difference experimental collection method of sorghum leaves comprises: The leaf temperature of 50 pots of sorghum plants was measured using the ThermaCAMSC3000 infrared imager produced by FLIR Corporation of the United States. After the waterlogging stress was over, 5 pots of sorghum plants were selected from 50 pots and infrared images of sorghum were taken from 9:00 to 11:30 in the morning. The infrared imager was placed 1 meter above the sorghum leaves to take infrared images of sorghum, and the air temperature T at the time of shooting was recorded. air ; The sorghum leaf temperature was analyzed using FLIR thermal imaging software, and the average temperature of the center point of the three leaves at the top of the sorghum was taken as the sorghum leaf temperature T leaf ; According to the air temperature T air and sorghum leaf temperature T leaf , using the leaf temperature difference calculation formula, the leaf temperature difference △T of sorghum leaves was obtained; The calculation formula of leaf temperature difference is: △T=T leaf -T air ; Where △T is the leaf temperature difference of sorghum leaves, T leaf is the leaf temperature of sorghum, T air is the air temperature.

7. The method for monitoring and evaluating sorghum waterlogging according to claim 2, characterized in that: The method for constructing the sorghum waterlogging assessment model comprises: The leaf temperature difference of sorghum leaves is used as the input item of the first BP neural network, and the relative water content, net photosynthetic rate and transpiration rate of sorghum leaves are used as the output items of the first BP neural network; The relative water content of sorghum leaves is used as the input item of the second BP neural network, and the leaf temperature difference of sorghum leaves is used as the output item of the second BP neural network; The net photosynthetic rate of sorghum leaves is used as the input item of the third BP neural network, and the leaf temperature difference of sorghum leaves is used as the output item of the third BP neural network; The transpiration rate of sorghum leaves is used as the input item of the fourth BP neural network, and the leaf temperature difference of sorghum leaves is used as the output item of the fourth BP neural network; Constructing a consistency loss function for the first BP neural network, the second BP neural network, the third BP neural network and the fourth BP neural network; Based on the consistency loss function, the first BP neural network, the second BP neural network, the third BP neural network and the fourth BP neural network are trained to the optimum. Using the first BP neural network trained to the optimum as the sorghum waterlogging assessment model; Among them, the first BP neural network, the second BP neural network, the third BP neural network and the fourth BP neural network are respectively: (RWC1, Pn1, E1) = BP1 (ΔT); △T2=BP2(RWC); △T3=BP3(Pn); △T4=BP4(E); Wherein, RWC1, Pn1, E1 are the relative water content, net photosynthetic rate, and transpiration rate output by the first BP neural network, respectively; △T is the leaf temperature difference input by the first BP neural network; △T2 is the leaf temperature difference output by the second BP neural network; RWC is the relative water content input by the second BP neural network; △T3 is the leaf temperature difference output by the third BP neural network; Pn is the net photosynthetic rate input by the second BP neural network; △T4 is the leaf temperature difference output by the fourth BP neural network; and E is the transpiration rate input by the fourth BP neural network.

8. The method for monitoring and evaluating sorghum waterlogging according to claim 7, characterized in that: The consistency loss function is: Loss1=||RWC1-RWC||2+||Pn1-Pn||2+||E1-E||2; Loss2=||△T2-△T||2+||△T3-△T||2+||△T4-△T||2 Loss3=||BP1(△T2)-RWC||2+||BP1(△T3)-Pn||2+||BP1(△T4)-E||2; Loss4=||BP2(RWC1)-△T||2+||BP3(Pn1)-△T||2+||BP4(E1)-△T||2; Wherein, Loss1 is the loss function of the first BP neural network, Loss2 is the loss function of the second BP neural network, the third BP neural network and the fourth BP neural network, Loss3 is the conversion consistency loss function of relative water content, net photosynthetic rate and transpiration rate, and Loss4 is the conversion consistency loss function of leaf temperature difference.

9. The method for monitoring and evaluating sorghum waterlogging according to claim 2, characterized in that: The monitoring and evaluation methods of waterlogging damage to sorghum plants using infrared imagers combined with sorghum waterlogging damage assessment models include: The sorghum plants to be monitored were monitored between 9:00 and 11:30 in the morning. The infrared imager was placed 1 meter above the sorghum leaves to take infrared images of the sorghum, and the air temperature T at the time of shooting was recorded. air ; The FLIR thermal imaging software was used to analyze the leaf temperature of the sorghum plant to be monitored, and the average temperature of the center point of the three leaves on the top of the sorghum was taken as the leaf temperature T of the sorghum plant to be monitored. leaf ; According to the air temperature T air and the leaf temperature T of the sorghum plant to be monitored leaf , using the leaf temperature difference calculation formula, the leaf temperature difference △T of the sorghum leaves of the sorghum plants to be monitored is obtained; The leaf temperature difference △T of the sorghum leaves of the sorghum plants to be monitored is input into the sorghum waterlogging assessment model to obtain the relative water content, net photosynthetic rate and transpiration rate of the sorghum leaves of the sorghum plants to be monitored.

10. A monitoring and evaluation system for sorghum waterlogging, characterized in that: A monitoring and evaluation method for sorghum waterlogging as described in any one of claims 1 to 9, the system comprising: Data collection unit to obtain indicators of the impact of waterlogging on sorghum; A data processing unit, used for determining the leaf temperature difference as a monitoring index of sorghum waterlogging through the correlation between the leaf temperature difference and the influencing index; And the correlation between leaf temperature difference and influencing indicators was trained through deep learning, and a sorghum waterlogging assessment model was constructed to quantify the impact of waterlogging on sorghum based on leaf temperature difference; The monitoring and evaluation unit is used to monitor and evaluate the waterlogging damage of the monitored sorghum plants using an infrared imager combined with a sorghum waterlogging damage evaluation model.

Images