Adaptive weight adjustment corn water deficit diagnosis method and device

Through the adaptive weight adjustment method, combined with crop, soil and environmental characteristics, and using deep convolutional neural network, the problem of inaccurate diagnosis of crop moisture deficiency in the existing technology is solved, and the accurate assessment of corn moisture deficiency is achieved. .

CN118711069BActive Publication Date: 2024-11-12INTELLIGENT EQUIPMENT RESEARCH CENTER BEIJING ACADEMY OF AGRICULTURE AND FORESTRY SCIENCES
View PDF 7 Cites 0 Cited by

Patent Information

Application Number
CN202411163382.X
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-08-23
Publication Date
2024-11-12
Estimated Expiration
2044-08-23

AI Technical Summary

Technical Problem

The diagnosis method of crop water deficiency in the prior art has the problem of inaccurate judgment, which is difficult to accurately reflect the crop water deficiency. It is mainly due to the dependence of a single indicator and neglecting the comprehensive impact of multiple factors of soil, crops and environment.

Method used

Adaptive weight-adjusted corn moisture deficiency diagnosis method is used to obtain crop phenotype characteristics, soil characteristics and environmental characteristics, and feature extraction is carried out according to the growth period of corn, key feature parameters are determined, and pre-trained deep convolutional neural network is used. Combined with information entropy, the characteristic parameter weight is adjusted adaptively, and a comprehensive analysis of the water deficiency of corn is carried out.

Benefits of technology

It improves the accurate diagnosis of corn moisture deficiency, can more accurately evaluate the impact of each characteristic parameter on moisture deficiency, and enhances the accuracy and reliability of the diagnosis.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118711069B_ABST
    Figure CN118711069B_ABST
Patent Text Reader

Abstract

The present invention provides a corn water deficit diagnosis method and device with adaptive weight adjustment, which are applied to the field of crop water deficit judgment. The method comprises: obtaining crop phenotypic characteristics, soil characteristics and environmental characteristics; performing feature extraction according to the growth period of corn, and determining key feature parameters corresponding to each growth period, wherein the growth period includes seedling stage, jointing stage, tasseling and silking stage, and filling and maturity stage; determining the influence index of the key feature parameters on corn water deficit; determining a target feature parameter whose influence index is greater than an influence index threshold; inputting the target feature parameter into a pre-trained deep convolutional neural network, and obtaining the degree of corn water deficit output by the pre-trained deep convolutional neural network, wherein the weight of the target feature parameter is adaptively adjusted based on the information entropy of the target feature parameter; the present invention can accurately diagnose the corn water deficit.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of crop water deficit diagnosis, and in particular to a corn water deficit diagnosis method and device with adaptive weight adjustment. Background Art

[0002] Crop water deficit refers to the phenomenon that the rate at which crops absorb water is lower than the rate at which they lose water, resulting in insufficient water in the crops and hindering their normal physiological activities. The indicators used to reflect the degree of crop water deficit include leaf water potential and saturation deficit.

[0003] In related technologies, water deficit diagnosis is mainly based on soil data, environmental data, and crop infrared image data. However, the crop water consumption process is affected by the combined influence of three indicators: soil, crop, and environment. It is difficult to accurately reflect the crop water deficit situation by relying on a single indicator.

[0004] It can be seen that the crop water deficit diagnosis method in the related art has the technical problem of inaccurate judgment of the crop water deficit situation. Summary of the invention

[0005] The present invention provides a corn water deficit diagnosis method and device with adaptive weight adjustment, which are used to solve the defect of inaccurate judgment of crop water deficit in the prior art crop water deficit diagnosis method, and realize the joint diagnosis of corn water deficit by analyzing crop data, environmental data and soil data under the time series changes of the entire growth cycle of corn crops.

[0006] The present invention provides a corn water deficit diagnosis method with adaptive weight adjustment, comprising the following steps: obtaining crop phenotypic characteristics, soil characteristics and environmental characteristics, wherein the crop phenotypic characteristics are collected by an image sensor, the soil characteristics are collected by a soil sensor, and the environmental characteristics are collected by an environmental sensor; extracting characteristics according to the growth period of corn, and determining key characteristic parameters corresponding to each growth period, wherein the growth period includes the seedling stage, the jointing stage, the tasseling and silking stage, and the filling and maturity stage; determining the influence index of the key characteristic parameters on corn water deficit; determining the target characteristic parameter whose influence index is greater than the influence index threshold, wherein the influence index threshold is set according to the growth period; inputting the target characteristic parameter into a pre-trained deep convolutional neural network, and obtaining the degree of water deficit of the corn output by the pre-trained deep convolutional neural network, wherein the weight of the target characteristic parameter is adaptively adjusted based on the information entropy of the target characteristic parameter.

[0007] According to a method for diagnosing corn water deficit with adaptive weight adjustment provided by the present invention, the method for obtaining crop phenotypic characteristics, soil characteristics and environmental characteristics includes: installing an image sensor, a soil sensor and an environmental sensor at a preset position; performing timed acquisition by the image sensor to obtain a corn visible light image and a corn thermal infrared image; performing classification based on the corn visible light image by a preset image classification model to obtain leaf color, leaf morphology and tassel morphology; performing segmentation based on the corn visible light image by a preset image segmentation model to obtain plant height; performing extraction based on the corn thermal infrared image to obtain canopy temperature; using the leaf color, leaf morphology, tassel morphology, plant height and canopy temperature as crop phenotypic characteristics; performing timed acquisition by the soil sensor to obtain soil moisture content and soil temperature, and using the soil moisture content and soil temperature as soil characteristics, wherein the soil moisture content includes soil moisture content at multiple different preset depths; performing timed acquisition by the environmental sensor to obtain air temperature, air humidity, precipitation and solar radiation, and using the air temperature, air humidity, precipitation and solar radiation as environmental characteristics.

[0008] According to a method for diagnosing corn water deficit with adaptive weight adjustment provided by the present invention, feature extraction is performed according to the growth period of corn to determine key feature parameters corresponding to each growth period, including: when the corn is in the seedling stage, the key feature parameters of the corn in the seedling stage are determined, wherein the key feature parameters of the seedling stage include crop phenotypic characteristics of the seedling stage, soil characteristics of the seedling stage and environmental characteristics of the seedling stage, and the crop phenotypic characteristics of the seedling stage include leaf color, canopy coverage and canopy temperature; the soil characteristics of the seedling stage include soil moisture content from 0 to 20 cm and soil temperature; the environmental characteristics of the seedling stage include air temperature, air humidity and precipitation; when the corn is in the jointing stage, the key feature parameters of the corn in the jointing stage are determined, wherein the key feature parameters of the jointing stage include crop phenotypic characteristics of the jointing stage, soil characteristics of the jointing stage and environmental characteristics of the jointing stage; the crop phenotypic characteristics of the jointing stage include leaf morphology and the canopy coverage; the soil characteristics of the jointing stage include soil moisture content from 0 to 40 cm and the soil temperature; the environmental characteristics of the jointing stage include air temperature, air humidity and precipitation. the precipitation, solar radiation; when the corn is in the tasseling and silking stage, determining the key characteristic parameters of the corn in the tasseling and silking stage, wherein the key characteristic parameters of the tasseling and silking stage include crop phenotypic characteristics of the tasseling and silking stage, soil characteristics of the tasseling and silking stage, and environmental characteristics of the tasseling and silking stage; the crop phenotypic characteristics of the tasseling and silking stage include male spike morphology and the canopy temperature; the soil characteristics of the tasseling and silking stage include soil moisture content from 0 to 60 cm and the soil temperature; the environmental characteristics of the tasseling and silking stage include the air temperature, the air humidity, the solar radiation; when the corn is in the grain filling maturity stage, determining the key characteristic parameters of the corn in the grain filling maturity stage, wherein the key characteristic parameters of the grain filling maturity stage include crop phenotypic characteristics of the grain filling maturity stage, soil characteristics of the grain filling maturity stage and environmental characteristics of the grain filling maturity stage; the crop phenotypic characteristics of the grain filling maturity stage include the leaf color and the canopy temperature; the soil characteristics of the grain filling maturity stage include the soil moisture content from 0 to 80 cm and the soil temperature; the environmental characteristics of the grain filling maturity stage include the air temperature, the air humidity and the solar radiation.

[0009] According to a corn water deficit diagnosis method with adaptive weight adjustment provided by the present invention, before the target characteristic parameters are input into a pre-trained deep convolutional neural network to obtain the degree of water deficit of the corn output by the pre-trained deep convolutional neural network, the method further includes: sorting the target characteristic parameters of each growth period in chronological order to construct a time series characteristic matrix corresponding to each growth period; determining the characteristic parameter proportion of each characteristic parameter in the time series characteristic matrix; determining the information entropy of each characteristic parameter according to the characteristic parameter proportion; taking the difference between a constant 1 and the information entropy as a coefficient of variation; determining the sum of the coefficients of variation of each characteristic parameter, and taking the ratio of the coefficient of variation of each characteristic parameter to the sum as the weight of each characteristic parameter; performing time series analysis on the weight of each characteristic parameter to determine a time series analysis result, wherein the time series analysis result is used to represent the changing trend of the weight over time; setting a weight threshold based on the time series analysis result, and adjusting the weight of the characteristic parameter when the weight of the characteristic parameter exceeds the weight threshold.

[0010] According to a corn water deficit diagnosis method with adaptive weight adjustment provided by the present invention, before sorting the target characteristic parameters of each growth period in chronological order and constructing a time series characteristic matrix corresponding to each growth period, the method also includes: standardizing the target characteristic parameters to map the numerical values ​​of the target characteristic parameters to a preset numerical range.

[0011] According to a corn water deficit diagnosis method with adaptive weight adjustment provided by the present invention, determining the influence index of the key characteristic parameter on corn water deficit includes: determining the correlation coefficient between the key characteristic parameter and corn water deficit, and the contribution rate of the key characteristic parameter to the corn water deficit; and taking the product of the contribution rate and the correlation coefficient as the influence index of the corn water deficit.

[0012] The present invention also provides a corn water deficit diagnosis device with adaptive weight adjustment, comprising the following modules:

[0013] An acquisition module is used to acquire crop phenotypic characteristics, soil characteristics and environmental characteristics, wherein the crop phenotypic characteristics are collected by image sensors, the soil characteristics are collected by soil sensors, and the environmental characteristics are collected by environmental sensors; an extraction module is used to perform feature extraction according to the growth period of corn and determine the key feature parameters corresponding to each growth period, wherein the growth period includes seedling stage, jointing stage, tasseling and silking stage, and filling and maturity stage; a first determination module is used to determine the influence index of the key feature parameters on corn water deficit; a second determination module is used to determine the target feature parameter whose influence index is greater than the influence index threshold, wherein the influence index threshold is set according to the growth period; an output module is used to input the target feature parameter into a pre-trained deep convolutional neural network to obtain the degree of water deficit of the corn output by the pre-trained deep convolutional neural network, wherein the weight of the target feature parameter is adaptively adjusted based on the information entropy of the target feature parameter.

[0014] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the corn water deficit diagnosis method with adaptive weight adjustment as described in any one of the above is implemented.

[0015] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the method for diagnosing corn water deficiency with adaptive weight adjustment as described in any one of the above is implemented.

[0016] The present invention also provides a computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the method for diagnosing corn water deficiency with adaptive weight adjustment as described above is implemented.

[0017] The adaptive weight-adjusted corn water deficit diagnosis method and device provided by the present invention obtain crop phenotypic characteristics, soil characteristics and environmental characteristics, extract characteristics according to the growth period of corn, and determine the key characteristic parameters corresponding to each growth period. Therefore, considering that the crop water consumption process is interactively affected by multiple factors such as the environment, soil, and the crop itself, it is difficult to accurately reflect the crop water deficit situation by relying on a single indicator. The present invention aims to jointly diagnose the corn water deficit situation by analyzing the crop data, environmental data and soil data under the time series changes of the entire growth cycle of the corn crop; determine the target characteristic parameter whose influence index is greater than the influence index threshold, thereby, the influence of each characteristic parameter and the corn water deficit can be comprehensively analyzed and evaluated, and the target characteristic parameter is input into a pre-trained deep convolutional neural network to obtain the corn water deficit degree output by the pre-trained deep convolutional neural network, thereby solving the technical problem of inaccurate judgment of the crop water deficit situation in the crop water deficit diagnosis method in the related art. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced one by one below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0019] Figure 1 It is a flow chart of the corn water deficit diagnosis method with adaptive weight adjustment provided by the present invention.

[0020] Figure 2 It is a schematic diagram of a process for obtaining crop phenotypic characteristics, soil characteristics and environmental characteristics provided by an embodiment of the present invention.

[0021] Figure 3 It is a schematic diagram of the process of adaptive dynamic weight adjustment provided by the present invention.

[0022] Figure 4 It is a schematic diagram of a process for determining the influence of corn water deficit provided by the present invention.

[0023] Figure 5 It is an overall flow chart of the corn water deficit diagnosis method with adaptive weight adjustment provided by the present invention.

[0024] Figure 6 It is a structural schematic diagram of a corn water deficit diagnostic device with adaptive weight adjustment provided by the present invention.

[0025] Figure 7 It is a schematic diagram of the physical structure of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0026] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are 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.

[0027] Crop water deficit refers to the phenomenon that the rate at which crops absorb water is lower than the rate at which they lose water, resulting in insufficient water in the crops and hindering their normal physiological activities. The indicators used to reflect the degree of crop water deficit include leaf water potential and saturation deficit.

[0028] Moderate water deficit will cause stomatal closure and restrict crop growth; severe water deficit will lead to cessation of photosynthesis, metabolic disorders and even death of the entire plant. Therefore, it is urgent to strengthen water-saving agriculture, achieve precision irrigation and improve water-saving efficiency. The key issue lies in the grasp of irrigation timing, so crop water deficit diagnosis is necessary. At present, crop water deficit diagnosis mainly obtains crop water deficit status indirectly or directly through moisture monitoring of three types of indicators: soil, crops and environment. Diagnosis of crop water demand from a single type of indicator has formed a relatively mature research method. However, the crop water consumption process is affected by multiple factors such as the environment, soil, and crops themselves, showing a complex change characteristic of nonlinearity, transientness, and strong mutual coupling. It is difficult to accurately reflect the crop water deficit situation by relying on a single indicator.

[0029] Therefore, the embodiment of the present invention proposes a corn water deficit diagnosis method based on adaptive weight adjustment of multimodal data under time series changes, so as to comprehensively and accurately diagnose the water deficit of crops.

[0030] In the related technology, water deficit diagnosis is mainly based on a certain type of data, including soil data, environmental data, and crop infrared image data. However, the crop water consumption process is affected by the comprehensive influence of three types of indicators, namely soil, crop, and environment. It is difficult to accurately reflect the crop water deficit situation by relying on a single indicator. Therefore, it is necessary to consider the mutual coupling of multimodal data such as soil data, environmental data, and crop image data. Secondly, the existing research does not consider the temporal changes in the crop growth process. The key characteristics of processes in different growth periods are different, and the weights of multimodal data need to be dynamically and adaptively adjusted. Therefore, a corn water deficit diagnosis method based on adaptive weight adjustment of multimodal data under temporal changes is needed, which integrates soil data, environmental data, and corn visible light and thermal infrared image data to diagnose the water deficit status of corn crops.

[0031] Considering that the crop water consumption process is interactively affected by multiple factors such as the environment, soil, and the crop itself, it presents a complex change characteristic of nonlinearity, transientness, and strong mutual coupling. It is difficult to accurately reflect the crop water deficit situation by relying on a single indicator. The present invention aims to jointly diagnose the water deficit of corn by analyzing the crop data, environmental data, and soil data of the corn crop under the time series changes of the entire growth cycle.

[0032] The embodiment of the present invention uses environmental sensors, soil sensors and image sensors to collect data of the entire growth cycle of corn growth, and preliminarily screens the key characteristics of corn in different growth stages by analyzing the influencing factors of water deficit in different growth stages. Calculate the influence of each characteristic parameter on corn water deficit and screen the key characteristic parameters. Construct a corn water deficit diagnosis model to judge the degree of corn water deficit, that is, no water deficit, mild water deficit, moderate water deficit, and severe water deficit.

[0033] refer to Figure 1 , Figure 1 : is a flow chart of the corn water deficit diagnosis method with adaptive weight adjustment provided by the present invention, such as Figure 1 As shown, the method includes the following steps.

[0034] Step 101, acquiring crop phenotypic characteristics, soil characteristics and environmental characteristics, wherein the crop phenotypic characteristics are acquired by an image sensor, the soil characteristics are acquired by a soil sensor, and the environmental characteristics are acquired by an environmental sensor.

[0035] In the embodiment of the present invention, environmental sensors, soil sensors and image sensors are installed at fixed positions to automatically collect data at regular intervals.

[0036] Among them, environmental sensors collect air temperature, air humidity, precipitation, and solar radiation, soil sensors collect soil moisture, and image sensors collect visible light images and thermal infrared images. After data collection is completed, the raw data is preprocessed, including data cleaning and processing.

[0037] Step 102, extracting features according to the growth period of corn, and determining key feature parameters corresponding to each growth period, wherein the growth period includes the seedling stage, the jointing stage, the tasseling and silking stage, and the grain filling and maturity stage.

[0038] Since the characteristic parameters of corn at different growth stages are not exactly the same, feature extraction is performed according to the growth stage. The influencing factors of water deficit at different growth stages of corn are shown in Table 1 (and Table 1 continuation), which is the influencing factors of water deficit at different growth stages of corn provided by the present invention, including growth stage, characteristic type, key characteristics and description.

[0039] Table 1

[0040]

[0041] Table 1 continued

[0042]

[0043] Since the crop characteristics, soil characteristics, and environmental characteristics during the harvest period have little effect on the yield and quality of the current corn crop, and mainly affect the growth conditions of the next crop, the impact of various factors during the harvest period on crop water deficit is not considered.

[0044] Step 103, determining the influence index of the key characteristic parameters on corn water deficit.

[0045] In the embodiment of the present invention, the correlation coefficient R between each characteristic parameter and corn water deficit is calculated using the correlation analysis method, and the contribution rate analysis method is used to calculate the contribution rate C of each characteristic parameter to corn water deficit. The correlation index and the contribution rate index are integrated to form a new evaluation index, namely, the influence index I.

[0046] Step 104, determining a target characteristic parameter whose influence index is greater than an influence index threshold, wherein the influence index threshold is set according to the growth period.

[0047] To determine the target characteristic parameters whose influence index is greater than the influence index threshold, it is first necessary to set the corresponding influence index threshold according to the characteristics and importance of each growth period.

[0048] The influence index can comprehensively analyze and evaluate the impact of each characteristic parameter on corn water deficit. The influence index is used to screen key characteristic parameters. The larger the influence index, the greater the impact of the characteristic parameter on corn water deficit.

[0049] Step 105, inputting the target feature parameter into a pre-trained deep convolutional neural network to obtain the degree of water deficit of corn output by the pre-trained deep convolutional neural network, wherein the weight of the target feature parameter is adaptively adjusted based on the information entropy of the target feature parameter.

[0050] In this embodiment of the present invention, the environmental characteristics, soil characteristics, and crop phenotypic characteristics that have a greater impact on corn water deficit screened in step 104 are used as input to construct a deep convolutional neural network to diagnose the degree of corn water deficit, i.e., no water deficit, mild water deficit, moderate water deficit, and severe water deficit.

[0051] Since the data of various characteristic parameters in the growth process of corn show time series changes, it is necessary to combine the characteristic parameter data of each growth period in the growth process of corn when training the deep convolutional neural network, use the entropy weight method combined with time series analysis, and adaptively and dynamically adjust the weights of different characteristic parameters to improve the accurate diagnosis of corn water deficit.

[0052] Through the above-mentioned embodiments provided by the present invention, crop phenotypic characteristics, soil characteristics and environmental characteristics are obtained, wherein the crop phenotypic characteristics are collected by image sensors, the soil characteristics are collected by soil sensors, and the environmental characteristics are collected by environmental sensors; feature extraction is performed according to the growth period of corn to determine the key feature parameters corresponding to each growth period, wherein the growth period includes the seedling stage, the jointing stage, the tasseling and silking stage, and the filling and maturity stage; the influence index of the key feature parameters on the water deficit of corn is determined; the target feature parameter whose influence index is greater than the influence index threshold is determined, wherein the influence index threshold is set according to the growth period; the target feature parameter is input into a pre-trained deep convolutional neural network to obtain the degree of water deficit of corn output by the pre-trained deep convolutional neural network, wherein the weight of the target feature parameter is adaptively adjusted based on the information entropy of the target feature parameter; thereby solving the technical problem of inaccurate judgment of the water deficit situation of crops in the crop water deficit diagnosis method in the related art.

[0053] refer to Figure 2 , Figure 2 It is a schematic diagram of a process for obtaining crop phenotypic characteristics, soil characteristics and environmental characteristics provided by an embodiment of the present invention.

[0054] According to a method for diagnosing corn water deficit with adaptive weight adjustment provided by the present invention, crop phenotypic characteristics, soil characteristics and environmental characteristics are obtained, including the following steps.

[0055] Step 201: Install an image sensor, a soil sensor, and an environmental sensor at a preset position.

[0056] Step 202: performing timed acquisition through an image sensor to obtain a visible light image of corn and a thermal infrared image of corn.

[0057] Step 203, classifying the corn visible light image using a preset image classification model to obtain leaf color, leaf shape, and tassel shape.

[0058] Step 204 , performing segmentation based on the corn visible light image using a preset image segmentation model to obtain plant height.

[0059] Step 205: extract the corn thermal infrared image to obtain the canopy temperature.

[0060] Step 206, taking leaf color, leaf morphology, tassel morphology, plant height and canopy temperature as crop phenotypic characteristics.

[0061] Step 207, performing regular data collection through a soil sensor to obtain soil moisture content and soil temperature, and using the soil moisture content and soil temperature as soil characteristics, wherein the soil moisture content includes soil moisture contents at multiple different preset depths.

[0062] Step 208, performing regular data collection through environmental sensors to obtain air temperature, air humidity, precipitation, and solar radiation, and using the air temperature, air humidity, precipitation, and solar radiation as environmental characteristics.

[0063] In the embodiment of the present invention, environmental sensors, soil sensors and image sensors are installed at fixed positions to automatically collect data at regular intervals. The environmental sensors collect air temperature, air humidity, precipitation, and solar radiation, the soil sensors collect soil moisture, and the image sensors collect visible light images and thermal infrared images. The raw data is preprocessed, including data cleaning and processing.

[0064] In an embodiment of the present invention, an appropriate installation location is selected for the environmental sensor, usually a place that is open, unobstructed and representative of the average environmental conditions of the monitoring area, and the environmental sensor is installed to monitor air temperature, air humidity, precipitation (for example, a rain collector), and solar radiation (for example, a light intensity sensor or a radiometer).

[0065] According to the soil depth and crop root distribution, install soil moisture sensors at different soil depths. Ensure that the sensor is in good contact with the soil to accurately reflect the soil moisture status.

[0066] Mount the image sensor at a suitable height and angle to capture visible light and thermal infrared images of the monitored area. Thermal infrared image sensors (such as infrared cameras) need to be calibrated to ensure accurate temperature measurement.

[0067] In the embodiment of the present invention, the timed collection can be to configure the data collection system or software to automatically read data from the sensor at a set time interval (such as every hour, every day, etc.), and store the collected data in a local server or cloud database to ensure data security and accessibility.

[0068] Among them, the above-mentioned data preprocessing of the original data can be to check whether there are missing values ​​in the data set. For missing values, interpolation methods (such as linear interpolation, mean value filling, etc.) can be used or records with missing values ​​can be deleted according to specific circumstances. For image data, especially thermal infrared images, filtering processing is required to reduce noise interference, such as using median filtering, Gaussian filtering, etc. For environmental data and soil data, standardization or normalization processing is performed as needed so that each feature is at the same level in subsequent analysis.

[0069] According to a method for diagnosing corn water deficit with adaptive weight adjustment provided by the present invention, feature extraction is performed according to the growth period of corn to determine key feature parameters corresponding to each growth period, including the following steps.

[0070] When corn is in the seedling stage, the key characteristic parameters of the corn seedling stage are determined, wherein the key characteristic parameters of the seedling stage include seedling crop phenotypic characteristics, seedling soil characteristics and seedling environmental characteristics. The seedling crop phenotypic characteristics include leaf color, canopy coverage, and canopy temperature; the seedling soil characteristics include 0 to 20 cm soil moisture content and soil temperature; the seedling environmental characteristics include air temperature, air humidity, and precipitation.

[0071] When corn is in the jointing stage, the key characteristic parameters of the jointing stage of corn are determined, wherein the key characteristic parameters of the jointing stage include crop phenotypic characteristics of the jointing stage, soil characteristics of the jointing stage and environmental characteristics of the jointing stage; crop phenotypic characteristics of the jointing stage include leaf morphology and canopy coverage; soil characteristics of the jointing stage include soil moisture content from 0 to 40 cm and soil temperature; environmental characteristics of the jointing stage include air temperature, air humidity, precipitation and solar radiation.

[0072] When corn is in the tasseling and silking stage, the key characteristic parameters of the tasseling and silking stage are determined, wherein the key characteristic parameters of the tasseling and silking stage include crop phenotypic characteristics of the tasseling and silking stage, soil characteristics of the tasseling and silking stage, and environmental characteristics of the tasseling and silking stage; crop phenotypic characteristics of the tasseling and silking stage include male ear morphology and canopy temperature; soil characteristics of the tasseling and silking stage include soil moisture content from 0 to 60 cm and soil temperature; environmental characteristics of the tasseling and silking stage include air temperature, air humidity, and solar radiation.

[0073] When corn is in the grain filling maturity stage, the key characteristic parameters of the grain filling maturity stage of the corn are determined, wherein the key characteristic parameters of the grain filling maturity stage include crop phenotypic characteristics of the grain filling maturity stage, soil characteristics of the grain filling maturity stage and environmental characteristics of the grain filling maturity stage; crop phenotypic characteristics of the grain filling maturity stage include leaf color and canopy temperature; soil characteristics of the grain filling maturity stage include soil moisture content from 0 to 80 cm and soil temperature; environmental characteristics of the grain filling maturity stage include air temperature, air humidity and solar radiation.

[0074] In the embodiment of the present invention, since the characteristic parameters concerned in different growth stages of corn are not exactly the same, feature extraction is performed according to the growth stage, with reference to Table 2, which is the key features of corn in different growth stages provided by the present invention.

[0075] Table 2

[0076]

[0077] The above environmental features are collected by environmental sensors, soil features are collected by soil sensors, and crop phenotypic features are extracted from images. In particular, the image classification model is used to obtain leaf color, leaf morphology, and tassel morphology in corn visible light images, and the image segmentation model is used to obtain plant height. Canopy temperature is extracted from corn thermal infrared images.

[0078] refer to Figure 3 , Figure 3 It is a schematic diagram of the process of adaptive dynamic weight adjustment provided by the present invention.

[0079] According to a corn water deficit diagnosis method with adaptive weight adjustment provided by the present invention, before inputting the target feature parameter into a pre-trained deep convolutional neural network to obtain the corn water deficit degree output by the pre-trained deep convolutional neural network, the method further includes:

[0080] Step 301, sorting the target characteristic parameters of each growth period in chronological order, and constructing a time series characteristic matrix corresponding to each growth period.

[0081] Step 302: determine the characteristic parameter weight of each characteristic parameter in the time series characteristic matrix.

[0082] Step 303: Determine the information entropy of each characteristic parameter according to the weight of each characteristic parameter.

[0083] Step 304: The difference between the constant 1 and the information entropy is used as the coefficient of variation.

[0084] Step 305: determine the sum of the coefficients of variation of each characteristic parameter, and use the ratio of the coefficient of variation of each characteristic parameter to the sum as the weight of each characteristic parameter.

[0085] Step 306, performing a time series analysis on the weight of each characteristic parameter to determine a time series analysis result, wherein the time series analysis result is used to represent a change trend of the weight over time.

[0086] Step 307: setting a weight threshold based on the timing analysis result, and adjusting the weight of the feature parameter when the weight of the feature parameter exceeds the weight threshold.

[0087] Since the data of various characteristic parameters in the growth process of corn show time series changes, it is necessary to combine the characteristic parameter data of each growth period in the growth process of corn when training the deep convolutional neural network, use the entropy weight method combined with time series analysis, and adaptively and dynamically adjust the weights of different characteristic parameters to improve the accurate diagnosis of corn water deficit.

[0088] Among them, the steps of adaptive dynamic weight adjustment are as follows:

[0089] Step 1, constructing a time series feature matrix: According to the time series of corn growth, the characteristic parameter data of each growth period are arranged in chronological order to construct a time series feature matrix for each growth period. The rows of the matrix represent different dates, and the columns represent different characteristic parameters.

[0090] It should be noted that before the above step 1, the data of each feature parameter is standardized and the sample value of each feature parameter is compressed within the interval [0-1] to eliminate the influence of unit difference on the calculation result, so that each feature parameter can be compared at the same scale.

[0091] Before the above step 1, according to the growth characteristics of corn, the entire growth process is divided into seedling stage, jointing stage, tasseling and silking stage, and filling and maturity stage.

[0092] Step 2, calculate the proportion of characteristic parameters: for each growth period, calculate the proportion of each sample value under each characteristic parameter to the characteristic parameter, which can be referred to the following formula (1).

[0093] (1)

[0094] in, ; Indicates The first sample of (corn) The proportion of characteristic parameters, represents the standardized data matrix (i.e. the time series feature matrix mentioned above), represents the total number of samples, Indicates the total number of feature parameters.

[0095] Step 3, calculate information entropy: for each growth period, calculate the information entropy of each characteristic parameter according to the characteristic parameter proportion. The information entropy can refer to the following formula (2).

[0096] (2)

[0097] in, Indicates The information entropy of the characteristic parameters is represents the total number of samples, Indicates The first sample of (corn) The proportion of characteristic parameters.

[0098] Step 4, calculate the coefficient of variation: In order to reflect the degree of variation between the characteristic parameters, it is necessary to calculate the redundancy of information entropy. The redundancy of information entropy (i.e. the above-mentioned coefficient of variation) can refer to the following formula (3).

[0099] (3)

[0100] Among them, Indicates The coefficient of variation of the characteristic parameters, Indicates The information entropy of the feature parameters.

[0101] In the embodiment of the present invention, the coefficient of variation is calculated for each growth period, and the larger the coefficient of variation is, the greater the importance of the indicator in the comprehensive evaluation is.

[0102] Step 5, determine the weight: for each growth period, determine the weight of each indicator according to the coefficient of difference. The weight can refer to the following formula (4).

[0103] (4)

[0104] in, Indicates The weight of the feature parameters, Indicates The coefficient of variation of the characteristic parameters, Indicates the total number of feature parameters.

[0105] Step 6, time series analysis: Perform time series analysis on the weights calculated for each growth period to determine the changing trend of the weights over time.

[0106] Among them, time series analysis is a method used to understand and predict the trend of time series data over time; using charts (such as line charts, bar charts, or area charts) to visualize the changes in weights over time. It helps to intuitively observe the overall change trend of weights and possible cyclical fluctuations. Through visualization, you can preliminarily identify characteristics such as trends (rising, falling, or stable), seasonal changes, cyclic fluctuations, and irregular fluctuations.

[0107] Step 7, adaptive dynamic adjustment: set a threshold based on the results of the timing analysis, and automatically adjust the weight of a feature parameter when the weight exceeds the threshold.

[0108] Adaptive dynamic weight adjustment is an intelligent decision-making method based on data analysis results, which aims to automatically optimize system performance according to changes in real-time or historical data.

[0109] In the embodiment of the present invention, appropriate methods (such as trend analysis, seasonal decomposition, cyclic fluctuation analysis, etc.) are used to perform time series analysis on the weights of characteristic parameters to understand their changing patterns over time. According to the results of the time series analysis, one or more thresholds are set for each characteristic parameter for each growth period in combination with business requirements and system goals. These thresholds can be absolute values ​​or percentage changes relative to historical data.

[0110] Through the embodiments of the present invention, on the basis of considering multiple factors of soil-crop-environment, combined with the actual changes in the growth process of corn, based on the time series characteristics of each characteristic parameter, combined with the entropy weight method and time series analysis, adaptive dynamic adjustment of the characteristic weights is achieved.

[0111] According to a method for diagnosing corn water deficit with adaptive weight adjustment provided by the present invention, before sorting the target characteristic parameters of each growth period in chronological order and constructing a time series characteristic matrix corresponding to each growth period, the method further includes:

[0112] The target characteristic parameters are standardized to map the values ​​of the target characteristic parameters to a preset value range.

[0113] In the embodiment of the present invention, the data of each characteristic parameter (target characteristic parameter) is standardized, and the sample value of each characteristic parameter is compressed within the interval [0-1] to eliminate the influence of unit difference on the calculation result, so that each characteristic parameter can be compared at the same scale. The standardized formula can refer to the following formula (5) to formula (6):

[0114] (5)

[0115] (6)

[0116] in, is the standardized data matrix; is the original data matrix; For the samples, For the characteristic parameters, is the number of samples, is the number of feature parameters.

[0117] Among them, the above formula (5) is a standardized formula for positive characteristic parameters, that is, the higher the value, the better the characteristic parameter; the above formula (6) is a standardized formula for negative characteristic parameters, that is, the lower the value, the better the characteristic parameter.

[0118] refer to Figure 4 , Figure 4It is a schematic diagram of a process for determining the influence of corn water deficit provided by the present invention.

[0119] According to a corn water deficit diagnosis method with adaptive weight adjustment provided by the present invention, determining the influence index of key characteristic parameters on corn water deficit includes the following steps:

[0120] Step 401, determining the correlation coefficient between the key characteristic parameter and the corn water deficit, and the contribution rate of the key characteristic parameter to the corn water deficit;

[0121] Step 402: taking the product of the contribution rate and the correlation coefficient as the influence index of corn water deficit.

[0122] In the embodiment of the present invention, the correlation coefficient R between each characteristic parameter and corn water deficit is calculated using a correlation analysis method, and the contribution rate analysis method is used to calculate the contribution rate C of each characteristic parameter to corn water deficit. The correlation index and the contribution rate index are integrated to form a new evaluation index, namely, the influence index I. The calculation formula of the influence can refer to the following formula (7).

[0123] (7)

[0124] in, Indicates The influence of characteristic parameters (indicators), Indicates The correlation coefficient of the characteristic parameters is Indicates The contribution rate of the characteristic parameters.

[0125] In the embodiment of the present invention, the influence index can comprehensively analyze and evaluate the influence of each characteristic parameter on corn water deficit. The influence index is used to screen key characteristic parameters. The larger the influence index, the greater the influence of the characteristic parameter on corn water deficit.

[0126] Through the above-mentioned embodiments of the present invention, the soil-crop-environment multi-factors are comprehensively considered to comprehensively diagnose corn water deficit, and numerical and image multi-modal data are integrated to construct a corn water deficit diagnosis method based on adaptive weight adjustment of multi-modal data under time series changes. On the basis of considering the soil-crop-environment multi-factors, combined with the actual changes in the corn growth process, based on the time series characteristics of each characteristic parameter, combined with the entropy weight method and time series analysis, adaptive dynamic adjustment of the feature weight is achieved. By integrating the correlation index and the contribution rate index, a new evaluation index, the influence index, is proposed, which can comprehensively analyze and evaluate the impact of each characteristic parameter on corn water deficit.

[0127] Based on the different key characteristic parameters of soil-crop-environment in different growth periods, a corn water deficit diagnosis method based on adaptive weight adjustment of multimodal data under time series changes is constructed to achieve adaptive dynamic adjustment of feature weights.

[0128] The following describes an example of an embodiment of the present invention in a practical application scenario.

[0129] refer to Figure 5 , Figure 5 It is an overall flow chart of the corn water deficit diagnosis method with adaptive weight adjustment provided by the present invention, which includes data collection and processing, feature parameter extraction, feature parameter optimization and corn water deficit diagnosis model.

[0130] Data collection and processing include: environmental data, soil data, crop thermal infrared images, and crop visible light images.

[0131] The characteristic parameter extraction includes: leaf color, plant height, canopy temperature, soil moisture content, air temperature, air humidity, and precipitation in the seedling stage; leaf morphology, leaf color, soil moisture content, air temperature, air humidity, precipitation, and solar radiation in the jointing stage; male spike morphology, canopy temperature, soil moisture content, air temperature, air humidity, and solar radiation in the tasseling and silking stage; leaf color, canopy temperature, soil moisture content, air humidity, precipitation, and solar radiation in the grain filling and maturity stage.

[0132] Feature parameter optimization includes: calculating the correlation coefficient using the correlation analysis method, calculating the contribution rate using the contribution rate analysis method, and calculating the influence of each feature parameter based on the correlation coefficient and contribution rate.

[0133] The corn water deficit diagnosis model includes: a corn water deficit diagnosis model based on adaptive dynamic weight adjustment to obtain the degree of corn water deficit.

[0134] First, the environmental sensor, soil sensor, and image sensor are used to obtain the environmental data, soil data, crop visible light, and thermal infrared image data of corn. The image classification model is used to obtain the leaf color, leaf shape, and tassel shape in the visible light image of corn, and the image segmentation model is used to obtain the plant height. The canopy temperature is extracted from the thermal infrared image of corn.

[0135] When corn is in the seedling stage, the correlation coefficient and contribution rate between each characteristic parameter and corn water deficit are calculated by using the correlation analysis method and contribution rate analysis method for seven characteristics, namely leaf color, plant height, canopy temperature, soil moisture content, air temperature, air humidity, and precipitation. The influence of each characteristic parameter on corn water deficit is analyzed by the influence index. The key characteristic parameters are selected according to the size of the influence index, and the selected characteristic parameters are input into the corn water deficit diagnosis model for training. The weights of the characteristic parameters are dynamically adjusted using the entropy weight method and time series analysis. The final water deficit degree result is obtained using the trained corn water deficit diagnosis model.

[0136] When corn is in the jointing stage, the correlation coefficient and contribution rate between each characteristic parameter and corn water deficit are calculated by using the correlation analysis method and contribution rate analysis method for seven characteristics, including leaf morphology, leaf color, soil moisture content, air temperature, air humidity, precipitation, and solar radiation. The influence of each characteristic parameter on corn water deficit is analyzed by the influence index. The key characteristic parameters are selected according to the size of the influence index, and the selected characteristic parameters are input into the corn water deficit diagnosis model for training. The entropy weight method and time series analysis are used to dynamically adjust the weights of the characteristic parameters. The trained corn water deficit diagnosis model is used to obtain the final water deficit degree result.

[0137] When corn is in the tasseling and silking stage, the correlation coefficient and contribution rate between each characteristic parameter and corn water deficit are calculated by using the correlation analysis method and contribution rate analysis method for the six characteristics of tassel morphology, canopy temperature, soil moisture content, air temperature, air humidity, and solar radiation. The influence of each characteristic parameter on corn water deficit is analyzed by the influence index. The key characteristic parameters are selected according to the size of the influence index, and the selected characteristic parameters are input into the corn water deficit diagnosis model for training. The entropy weight method and time series analysis are used to dynamically adjust the weights of the characteristic parameters. The trained corn water deficit diagnosis model is used to obtain the final water deficit degree result.

[0138] When corn is in the grain filling maturity stage, the correlation coefficient and contribution rate between each characteristic parameter and corn water deficit are calculated by using the correlation analysis method and contribution rate analysis method for the six characteristics of leaf color, canopy temperature, soil moisture content, air humidity, precipitation, and solar radiation. The influence of each characteristic parameter on corn water deficit is analyzed by the influence index. The key characteristic parameters are selected according to the size of the influence index, and the selected characteristic parameters are input into the corn water deficit diagnosis model for training. The weights of the characteristic parameters are dynamically adjusted using the entropy weight method and time series analysis. The final water deficit degree result is obtained using the trained corn water deficit diagnosis model.

[0139] The corn water deficit diagnosis device with adaptive weight adjustment provided by the present invention is described below. The corn water deficit diagnosis device with adaptive weight adjustment described below and the corn water deficit diagnosis method with adaptive weight adjustment described above can be referenced to each other.

[0140] refer to Figure 6 , Figure 6 It is a structural schematic diagram of the corn water deficit diagnostic device with adaptive weight adjustment provided by the present invention, which includes an acquisition module 601, an extraction module 602, a first determination module 603, a second determination module 604 and an output module 605.

[0141] An acquisition module 601 is used to acquire crop phenotypic characteristics, soil characteristics and environmental characteristics, wherein the crop phenotypic characteristics are acquired by an image sensor, the soil characteristics are acquired by a soil sensor, and the environmental characteristics are acquired by an environmental sensor;

[0142] Extraction module 602, for extracting features according to the growth period of corn, and determining key feature parameters corresponding to each growth period, wherein the growth period includes seedling stage, jointing stage, tasseling and silking stage, and grain filling and maturity stage;

[0143] The first determination module 603 is used to determine the influence index of the key characteristic parameter on the water deficit of corn;

[0144] A second determination module 604 is used to determine a target characteristic parameter whose influence index is greater than an influence index threshold, wherein the influence index threshold is set according to the growth period;

[0145] The output module 605 is used to input the target feature parameters into the pre-trained deep convolutional neural network to obtain the degree of water deficit of corn output by the pre-trained deep convolutional neural network, wherein the weight of the target feature parameters is adaptively adjusted based on the information entropy of the target feature parameters.

[0146] Specifically, the above-mentioned corn water deficit diagnosis device with adaptive weight adjustment provided by the present invention can implement all the method steps implemented by the above-mentioned corn water deficit diagnosis method embodiment with adaptive weight adjustment, and can achieve the same technical effect. The parts and beneficial effects that are the same as those in the method embodiment will not be described in detail here.

[0147] Figure 7 is a schematic diagram of the physical structure of the electronic device provided by the present invention, such as Figure 7As shown, the electronic device may include: a processor (processor) 710 , a communication interface (Communications Interface) 720 , a memory (memory) 730 and a communication bus 740 , wherein the processor 710 , the communication interface 720 , and the memory 730 communicate with each other through the communication bus 740 . The processor 710 can call the logic instructions in the memory 730 to execute the corn water deficit diagnosis method with adaptive weight adjustment, which includes: obtaining crop phenotypic characteristics, soil characteristics and environmental characteristics, wherein the crop phenotypic characteristics are collected by image sensors, the soil characteristics are collected by soil sensors, and the environmental characteristics are collected by environmental sensors; extracting features according to the growth period of corn to determine the key feature parameters corresponding to each growth period, wherein the growth period includes the seedling stage, the jointing stage, the tasseling and silking stage, and the filling and maturity stage; determining the influence index of the key feature parameters on corn water deficit; determining the target feature parameter whose influence index is greater than the influence index threshold, wherein the influence index threshold is set according to the growth period; inputting the target feature parameter into a pre-trained deep convolutional neural network to obtain the degree of corn water deficit output by the pre-trained deep convolutional neural network, wherein the weight of the target feature parameter is adaptively adjusted based on the information entropy of the target feature parameter.

[0148] In addition, the logic instructions in the above-mentioned memory 730 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on this understanding, the technical solution of the present invention can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0149] On the other hand, the present invention also provides a computer program product, which includes a computer program, which can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the corn water deficit diagnosis method with adaptive weight adjustment provided by the above methods, the method including: obtaining crop phenotypic characteristics, soil characteristics and environmental characteristics, wherein the crop phenotypic characteristics are collected by an image sensor, the soil characteristics are collected by a soil sensor, and the environmental characteristics are collected by an environmental sensor; performing feature extraction according to the growth period of corn, and determining key feature parameters corresponding to each growth period, wherein the growth period includes seedling stage, jointing stage, tasseling and silking stage, and filling and maturity stage; determining the influence index of the key feature parameter on corn water deficit; determining a target feature parameter whose influence index is greater than an influence index threshold, wherein the influence index threshold is set according to the growth period; inputting the target feature parameter into a pre-trained deep convolutional neural network to obtain the degree of corn water deficit output by the pre-trained deep convolutional neural network, wherein the weight of the target feature parameter is adaptively adjusted based on the information entropy of the target feature parameter.

[0150] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to execute the corn water deficit diagnosis method with adaptive weight adjustment provided by the above-mentioned methods, the method comprising: obtaining crop phenotypic characteristics, soil characteristics and environmental characteristics, wherein the crop phenotypic characteristics are collected by an image sensor, the soil characteristics are collected by a soil sensor, and the environmental characteristics are collected by an environmental sensor; performing feature extraction according to the growth period of corn, and determining key feature parameters corresponding to each growth period, wherein the growth period includes the seedling stage, the jointing stage, the tasseling and silking stage, and the filling and maturity stage; determining the influence index of the key feature parameters on corn water deficit; determining a target feature parameter whose influence index is greater than an influence index threshold, wherein the influence index threshold is set according to the growth period; inputting the target feature parameter into a pre-trained deep convolutional neural network to obtain the degree of corn water deficit output by the pre-trained deep convolutional neural network, wherein the weight of the target feature parameter is adaptively adjusted based on the information entropy of the target feature parameter.

[0151] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Those of ordinary skill in the art may understand and implement it without creative effort.

[0152] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0153] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A corn water deficit diagnosis method with adaptive weight adjustment, characterized in that: include: Acquiring crop phenotypic characteristics, soil characteristics and environmental characteristics, wherein the crop phenotypic characteristics are acquired by an image sensor, the soil characteristics are acquired by a soil sensor, and the environmental characteristics are acquired by an environmental sensor; Extracting features according to the growth period of corn to determine key feature parameters corresponding to each growth period, wherein the growth period includes seedling stage, jointing stage, tasseling and silking stage, and grain filling and maturity stage; Determine the influence index of the key characteristic parameter on corn water deficit; Determining a target characteristic parameter whose influence index is greater than an influence index threshold, wherein the influence index threshold is set according to the growth period; Inputting the target feature parameter into a pre-trained deep convolutional neural network to obtain the water deficit degree of the corn output by the pre-trained deep convolutional neural network, wherein the weight of the target feature parameter is adaptively adjusted based on the information entropy of the target feature parameter; Before inputting the target characteristic parameters into the pre-trained deep convolutional neural network to obtain the water deficit degree of the corn output by the pre-trained deep convolutional neural network, the method also includes: sorting the target characteristic parameters of each growth period in chronological order to construct a time series characteristic matrix corresponding to each growth period; determining the characteristic parameter proportion of each characteristic parameter in the time series characteristic matrix; determining the information entropy of each characteristic parameter according to the characteristic parameter proportion; taking the difference between a constant 1 and the information entropy as a coefficient of variation; determining the sum of the coefficients of variation of each characteristic parameter, and taking the ratio of the coefficient of variation of each characteristic parameter to the sum as the weight of each characteristic parameter; performing time series analysis on the weight of each characteristic parameter to determine a time series analysis result, wherein the time series analysis result is used to represent the changing trend of the weight over time; setting a weight threshold based on the time series analysis result, and adjusting the weight of the characteristic parameter when the weight of the characteristic parameter exceeds the weight threshold.

2. The corn water deficit diagnosis method with adaptive weight adjustment according to claim 1, characterized in that: The obtaining of crop phenotypic characteristics, soil characteristics and environmental characteristics includes: Install image sensors, soil sensors, and environmental sensors at pre-set locations; The image sensor is used to perform timed acquisition to obtain a visible light image of corn and a thermal infrared image of corn; Classifying the corn visible light image by a preset image classification model to obtain leaf color, leaf shape, and tassel shape; Segmenting the corn visible light image using a preset image segmentation model to obtain plant height; Extracting based on the corn thermal infrared image to obtain the canopy temperature; using the leaf color, the leaf morphology, the tassel morphology, the plant height and the canopy temperature as crop phenotypic characteristics; The soil sensor is used to collect data at regular intervals to obtain soil moisture content and soil temperature, and the soil moisture content and soil temperature are used as soil characteristics, wherein the soil moisture content includes soil moisture contents at multiple different preset depths; The environmental sensor is used to collect data at regular intervals to obtain air temperature, air humidity, precipitation, and solar radiation, and the air temperature, air humidity, precipitation, and solar radiation are used as environmental characteristics.

3. The corn water deficit diagnosis method with adaptive weight adjustment according to claim 1, characterized in that: The feature extraction is performed according to the growth period of corn to determine the key feature parameters corresponding to each growth period, including: When the corn is in the seedling stage, determining the key characteristic parameters of the corn in the seedling stage, wherein the key characteristic parameters of the seedling stage include seedling crop phenotypic characteristics, seedling soil characteristics and seedling environmental characteristics, the seedling crop phenotypic characteristics include leaf color, canopy coverage, and canopy temperature; the seedling soil characteristics include 0 to 20 cm soil moisture content and soil temperature; the seedling environmental characteristics include air temperature, air humidity, and precipitation; When the corn is in the jointing stage, the key characteristic parameters of the corn in the jointing stage are determined, wherein the key characteristic parameters of the jointing stage include crop phenotypic characteristics of the jointing stage, soil characteristics of the jointing stage, and environmental characteristics of the jointing stage; the crop phenotypic characteristics of the jointing stage include leaf morphology and canopy coverage; the soil characteristics of the jointing stage include soil moisture content from 0 to 40 cm and soil temperature; the environmental characteristics of the jointing stage include air temperature, air humidity, precipitation, and solar radiation; When the corn is in the tasseling and silking stage, the key characteristic parameters of the corn in the tasseling and silking stage are determined, wherein the key characteristic parameters of the tasseling and silking stage include crop phenotypic characteristics in the tasseling and silking stage, soil characteristics in the tasseling and silking stage, and environmental characteristics in the tasseling and silking stage; the crop phenotypic characteristics in the tasseling and silking stage include tassel morphology and the canopy temperature; the soil characteristics in the tasseling and silking stage include soil moisture content from 0 to 60 cm and the soil temperature; the environmental characteristics in the tasseling and silking stage include the air temperature, the air humidity, and the solar radiation; When the corn is in the grain filling maturity stage, the key characteristic parameters of the corn in the grain filling maturity stage are determined, wherein the key characteristic parameters of the grain filling maturity stage include crop phenotypic characteristics of the grain filling maturity stage, soil characteristics of the grain filling maturity stage and environmental characteristics of the grain filling maturity stage; the crop phenotypic characteristics of the grain filling maturity stage include the leaf color and the canopy temperature; the soil characteristics of the grain filling maturity stage include the soil moisture content of 0 to 80 cm and the soil temperature; the environmental characteristics of the grain filling maturity stage include the air temperature, the air humidity and the solar radiation.

4. The corn water deficit diagnosis method with adaptive weight adjustment according to claim 1, characterized in that: Before sorting the target characteristic parameters of each growth period in chronological order and constructing a time series characteristic matrix corresponding to each growth period, the method further includes: The target characteristic parameter is standardized to map the value of the target characteristic parameter to a preset value range.

5. The corn water deficit diagnosis method with adaptive weight adjustment according to claim 1, characterized in that: The determining of the influence index of the key characteristic parameter on corn water deficit includes: Determining the correlation coefficient between the key characteristic parameter and corn water deficit, and the contribution rate of the key characteristic parameter to the corn water deficit; The product of the contribution rate and the correlation coefficient is used as the influence indicator of the corn water deficit.

6. A corn water deficit diagnosis device with adaptive weight adjustment, characterized in that: include: An acquisition module, used to acquire crop phenotypic characteristics, soil characteristics and environmental characteristics, wherein the crop phenotypic characteristics are acquired by an image sensor, the soil characteristics are acquired by a soil sensor, and the environmental characteristics are acquired by an environmental sensor; An extraction module is used to extract features according to the growth period of corn and determine key feature parameters corresponding to each growth period, wherein the growth period includes seedling stage, jointing stage, tasseling and silking stage, and grain filling and maturity stage; A first determination module is used to determine the influence index of the key characteristic parameter on corn water deficit; A second determination module is used to determine a target characteristic parameter whose influence index is greater than an influence index threshold, wherein the influence index threshold is set according to the growth period; An output module, used for inputting the target characteristic parameter into a pre-trained deep convolutional neural network to obtain the water deficit degree of the corn output by the pre-trained deep convolutional neural network, wherein the weight of the target characteristic parameter is adaptively adjusted based on the information entropy of the target characteristic parameter; The output module is also used to sort the target characteristic parameters of each growth period in chronological order to construct a time series characteristic matrix corresponding to each growth period; determine the characteristic parameter proportion of each characteristic parameter in the time series characteristic matrix; determine the information entropy of each characteristic parameter according to the characteristic parameter proportion; take the difference between the constant 1 and the information entropy as the coefficient of variation; determine the sum of the coefficients of variation of each characteristic parameter, and take the ratio of the coefficient of variation of each characteristic parameter to the sum as the weight of each characteristic parameter; perform time series analysis on the weight of each characteristic parameter to determine the time series analysis result, wherein the time series analysis result is used to represent the changing trend of the weight over time; set a weight threshold based on the time series analysis result, and adjust the weight of the characteristic parameter when the weight of the characteristic parameter exceeds the weight threshold.

7. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the corn water deficit diagnosis method with adaptive weight adjustment as described in any one of claims 1 to 5 is implemented.

8. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the corn water deficit diagnosis method with adaptive weight adjustment as described in any one of claims 1 to 5 is implemented.

9. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the corn water deficit diagnosis method with adaptive weight adjustment as described in any one of claims 1 to 5 is implemented.

Citation Information

Patent Citations

  • Summer corn water deficit index data comprehensive display system

    CN112540992A

  • Crop moisture deficit diagnosis method based on soil moisture and meteorological monitoring

    CN116773744A

  • Moisture deficit detection model training method, moisture deficit detection method, moisture deficit detection device and moisture deficit detection equipment

    CN117541887A

  • Drought monitoring analysis method based on vegetation index and drought coupling relation

    CN118070120A

  • Corn waterlogging risk assessment method

    CN118313646A