Method for monitoring crop drought based on leaf morphology identification and system thereof

Abstract

The application discloses a monitoring method and system for identifying crop drought based on leaf morphology, and relates to the field of agricultural drought monitoring, which comprises the following steps: selecting a fully expanded healthy leaf, cleaning the leaf with ultrapure water and fixing the leaf on a standard carrier after nitrogen blowing and drying, and obtaining a standardized leaf sample; obtaining a leaf front and back gray-scale image based on the standardized leaf sample through a high-resolution scanner, and obtaining a binary image with a clear vein network through image splicing and binarization processing; performing skeletonization processing on the binary image with the clear vein network, calculating the vein fractal dimension by using a box-counting method, and obtaining a fractal dimension change rate; and fixing a detection area of the standardized leaf sample. Through innovative vein fractal dimension analysis and dynamic contact angle measurement, the application constructs a crop drought monitoring system, and in the aspect of vein analysis, skeletonization processing and a box-counting method are adopted to realize accurate quantification of the vein network topological structure.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of agricultural drought monitoring, in particular to a monitoring method for identifying crop drought based on leaf morphology and a system thereof. BACKGROUND

[0002] In the field of agricultural drought monitoring, diagnosis techniques based on plant physiological characteristics have attracted widespread attention in recent years. In the prior art, the degree of water stress is evaluated by collecting two-dimensional images of leaves, calculating morphological parameters such as leaf area shrinkage rate and leaf edge curling degree, and the like. This technique uses a machine learning algorithm to establish a mapping model of morphological characteristics and soil water content, achieving non-destructive monitoring, and showing certain practicality in field application. Similarly, multi-spectral imaging technology is used to capture leaf water reflection characteristics, and vegetation index is used for drought early warning. This method has an efficiency advantage in large-scale farmland monitoring.

[0003] The prior art has the limitation of single monitoring dimension, and the two-dimensional morphological parameters relied on are difficult to capture the changes in the topology of the leaf vein network. As the key channel for water transport, the early response characteristics of the leaf vein often precede the macroscopic morphological changes. The microstructure reorganization of the leaf vein at the early stage of drought will lead to changes in its fractal characteristics. However, existing image analysis methods are limited by resolution and algorithm, and cannot quantify the morphological evolution at the sub-millimeter level. Leaf surface wettability, as an important indicator reflecting the state of the wax layer, is directly related to the stomatal regulation mechanism, but has not been included in the existing monitoring system. SUMMARY

[0004] In view of the above existing problems, the present application is proposed.

[0005] Therefore, the present application provides a monitoring method for identifying crop drought based on leaf morphology, which solves the monitoring blind area problem that the existing technology cannot simultaneously capture the changes in the micro-topology of the leaf and the evolution of the surface physical characteristics.

[0006] To solve the above technical problems, the present application provides the following technical solutions:

[0007] In a first aspect, the present application provides a monitoring method for identifying crop drought based on leaf morphology, which comprises,

[0008] Select a fully expanded healthy leaf, wash it with ultrapure water and dry it with nitrogen gas, and then fix it on a standard sample stage to obtain a standardized leaf sample;

[0009] Based on the standardized leaf sample, obtain the gray scale images of the front and back surfaces of the leaf by a high-resolution scanner, and obtain a binary image with clear leaf vein network through image stitching and binary processing;

[0010] Skeletonization processing was performed on the binarized image with a clear leaf vein network, and the fractal dimension of the leaf vein was calculated using the box counting method to obtain the rate of change of fractal dimension;

[0011] In a fixed testing area of ​​a standardized blade sample, the advance angle and retreat angle are continuously measured using a contact angle measuring instrument, and the contact angle hysteresis value is calculated to obtain contact angle hysteresis change data.

[0012] The fractal dimension change rate and contact angle hysteresis change data are input into the random forest model to calculate the drought composite index, obtain the drought level results, and generate a visualization report based on the drought level results.

[0013] As a preferred embodiment of the crop drought monitoring method based on leaf morphology identification described in this invention, the method includes the following steps: selecting fully expanded healthy leaves, washing them with ultrapure water and drying them with nitrogen, and then fixing them on a standard stage to obtain standardized leaf samples.

[0014] Using sterile sampling forceps, select fully expanded healthy leaves from the middle of the crop canopy. Place the collected healthy leaves into a pre-cooled constant temperature transport box, and place a temperature and humidity recorder inside the pre-cooled constant temperature transport box to monitor the environment.

[0015] Healthy leaves are immersed in ultrapure water, cleaned with an ultrasonic cleaner to remove surface contaminants, and then dried with high-purity nitrogen to obtain dried healthy leaves.

[0016] The dried, healthy leaves were laid flat on the stage, and the edges were fixed with stress-free transparent fixing rings to obtain standardized leaf samples with clean surfaces, stable shapes, and no stress deformation.

[0017] As a preferred embodiment of the crop drought monitoring method based on leaf morphology identification described in this invention, the method includes: acquiring grayscale images of the front and back of standardized leaf samples using a high-resolution scanner, and then performing image stitching and binarization to obtain a binarized image with a clear leaf vein network, comprising the following steps.

[0018] Standardized leaf samples were placed on the glass platform of a high-resolution scanner, and the parameters of the high-resolution scanner were adjusted to acquire grayscale images of the front and back of the leaf, and the grayscale images of the front and back of the leaf were registered.

[0019] The Laplacian pyramid algorithm is used to fuse the registered grayscale images of the front and back of the leaf into a complete panoramic image of the leaf. The dynamic Otsu algorithm is then used to convert the registered grayscale images of the front and back of the leaf into binary images. Morphological operations using 3×3 circular structuring elements are then performed to remove noise points, resulting in a binarized image with a clear leaf vein network.

[0020] As a preferred scheme of the monitoring method for identifying crop drought based on leaf morphology, wherein: the binarized image with clear vein network is skeletonized, the box counting method is used to calculate the fractal dimension of the vein, and the change rate of the fractal dimension is obtained, including the following steps,

[0021] The connectivity of the binarized image with clear vein network is checked, and the binarized image with clear vein network after checking is used, the Zhang-Suen thinning algorithm is used for iterative processing, and a vein skeleton image is generated;

[0022] The scale range of the vein skeleton image is analyzed using the box counting method, the minimum grid number covering the vein skeleton is counted, and the fractal dimension of the vein is calculated using the fitted linear regression curve;

[0023] The fractal dimension of the vein of the healthy leaf of the same variety of crop is collected under standard irrigation conditions, and statistical analysis is performed to obtain the baseline value of the healthy sample of the same variety;

[0024] The fractal dimension of the vein is compared with the baseline value of the healthy sample of the same variety, and the change rate of the fractal dimension is calculated.

[0025] As a preferred scheme of the monitoring method for identifying crop drought based on leaf morphology, wherein: in the fixed detection area of the standardized leaf sample, the advancing angle and the receding angle are continuously measured by the contact angle measuring instrument, the contact angle hysteresis value is calculated, and the contact angle hysteresis change data is obtained, including the following steps,

[0026] At one-third of the distance from the leaf tip of the standardized leaf sample, a detection point is marked with a laser positioner, the marked leaf is placed in a constant temperature and humidity chamber for balancing, a standard PDMS sheet is used to calibrate the contact angle measuring instrument, ultrapure water is injected at the marked detection point, the profile at the maximum expansion moment of the droplet is recorded to obtain the advancing angle, the profile at the moment when the droplet starts to shrink is captured to obtain the receding angle, and the contact angle hysteresis value is calculated;

[0027] Under standard environmental control conditions, the healthy leaves of the same growth period are measured and statistically processed after being uniformly standardized to obtain the baseline value of the healthy sample of the same variety;

[0028] The contact angle hysteresis value is compared with the baseline value of the healthy sample of the same variety to obtain the contact angle hysteresis change data.

[0029] As a preferred scheme of the monitoring method for identifying crop drought based on leaf morphology, wherein: the fractal dimension change rate and the contact angle hysteresis change data are input into a random forest model, a drought comprehensive index is calculated, and a drought grade result is obtained, including the following steps,

[0030] The fractal dimension change rate and the contact angle hysteresis change data are combined into a two-dimensional feature vector, a RandomForestClassifier of Scikit-learn is used for ten-fold cross-validation training, and a trained random forest model is obtained;

[0031] The two-dimensional feature vector is input into the pre-trained random forest model, and a drought comprehensive index is calculated;

[0032] Based on the ROC curve analysis combined with the crop physiological response inflection point of the field measured data, an optimized segmentation point is formed through the Youden index, and a drought threshold is obtained;

[0033] When the drought comprehensive index is less than the drought threshold, the crop water state is determined to be in a normal range;

[0034] When the drought comprehensive index is equal to the drought threshold, the crop water state is determined to be in a mild drought;

[0035] When the drought comprehensive index is greater than the drought threshold, the crop water state is determined to be in a severe drought.

[0036] As a preferred scheme of the monitoring method for identifying crop drought based on leaf morphology, the application comprises the following steps of generating a visual report according to the drought grade result,

[0037] Based on the fractal dimension change rate, a JET color scale mapping is performed on the binary image with clear vein network to obtain a color scale degradation thermodynamic map, and Matplotlib is used to draw a time sequence curve of the advancing angle and the receding angle;

[0038] Based on the drought comprehensive index, the drought threshold and the irrigation event mark, a vector trend chart is generated;

[0039] The color scale degradation thermodynamic map, the time sequence curve of the advancing angle and the receding angle, and the vector trend chart are inserted into an HTML5 template through Jinja2 to generate a visual report.

[0040] In a second aspect, the application provides a monitoring system for identifying crop drought based on leaf morphology, comprising,

[0041] The standardized module selects a fully expanded healthy leaf, which is cleaned by ultrapure water and dried by nitrogen, and then fixed on a standard carrier table to obtain a standardized leaf sample;

[0042] The quantification module obtains the gray scale images of the front and back surfaces of the leaf based on the standardized leaf sample through a high-resolution scanner, and obtains a binary image with clear vein network through image stitching and binary processing;

[0043] The detection module performs skeletonization processing on the binarized image with a clear network of leaf veins, calculates the fractal dimension of the leaf veins by using box counting method, and obtains a fractal dimension change rate;

[0044] The fusion module continuously measures the advancing angle and the receding angle by using a contact angle measuring instrument in a fixed detection area of the standardized leaf sample, calculates a contact angle hysteresis value, and obtains contact angle hysteresis change data.

[0045] The execution module inputs the fractal dimension change rate and the contact angle hysteresis change data into a random forest model, calculates a drought comprehensive index, obtains a drought grade result, and generates a visual report according to the drought grade result.

[0046] In a third aspect, the present application provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, any step of the monitoring method for identifying crop drought based on leaf morphology according to the first aspect of the present application is implemented.

[0047] In a fourth aspect, the present application provides a computer readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, any step of the monitoring method for identifying crop drought based on leaf morphology according to the first aspect of the present application is implemented.

[0048] The present application has the following beneficial effects: through the innovative analysis of leaf vein fractal dimension and the dynamic measurement of contact angle, the present application constructs a crop drought monitoring system. In terms of leaf vein analysis, skeletonization processing and box counting method are used to accurately quantify the topological structure of the leaf vein network, and early changes in soil water content can be detected. In terms of surface property detection, the dynamic changes of contact angle are captured by high-speed photography, and temperature compensation is combined. The technical closed loop is formed by the steps of standardized sample preparation, high-resolution scanning, random forest model fusion and visual report generation, the multidimensional collaborative monitoring of micro morphology and surface characteristics is realized, and the problem of single monitoring dimension in the prior art is solved. BRIEF DESCRIPTION OF DRAWINGS

[0049] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed in the embodiment description. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0050] Fig. 1 The flowchart of the monitoring method for identifying crop drought based on leaf morphology.

[0051] Fig. 2 The schematic diagram of the monitoring system for identifying crop drought based on leaf morphology.

[0052] Fig. 3 Standardized leaf sample procedure schematic.

[0053] Fig. 4 Visual report generation schematic. DETAILED DESCRIPTION

[0054] In order to make the above objectives, features and advantages of the present application more apparent, specific embodiments of the present application are described in detail below with reference to the accompanying drawings.

[0055] In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present application. However, it will be apparent to one skilled in the art that the present application can be practiced without the specific details and other implementations can be employed. Therefore, the present application is not limited to the details described herein and can be practiced with variations that are within the scope and spirit of the claims.

[0056] Secondly, one embodiment or embodiments as referred to herein covers a particular feature, structure, or characteristic that can be included in at least one implementation of the present application. Occurrences of the expression one embodiment in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments.

[0057] Reference Figs. 1-4 For one embodiment of the present application, the embodiment provides a method for monitoring crop drought based on leaf morphology, comprising the following steps:

[0058] S1, select fully expanded healthy leaves, clean them with ultrapure water and dry them with nitrogen, and then fix them on a standard stage to obtain standardized leaf samples.

[0059] S1.1, use sterile sampling forceps to select fully expanded healthy leaves in the middle of the crop canopy, and place the collected healthy leaves into a pre-cooled constant temperature transport box. A temperature and humidity recorder is placed in the pre-cooled constant temperature transport box to monitor the environment.

[0060] Further, using a sterile sampling forceps, a fully expanded healthy leaf is selected from the third to fifth leaf position in the middle of the crop canopy, avoiding areas of disease, pests and mechanical damage. The collected healthy leaf is immediately placed in a pre-cooling constant temperature transport box with a temperature controlled at 25±1°C and a relative humidity of 60±5%. A temperature and humidity recorder with a precision of ±0.5°C is placed in the pre-cooling constant temperature transport box to monitor the environmental parameters in real time, ensuring that the temperature fluctuation during transportation does not exceed 1°C and the humidity fluctuation does not exceed 3%. The storage time of the healthy leaf in the pre-cooling constant temperature transport box is not more than 2 hours. During transportation, direct sunlight and severe vibration are avoided. The data collected by the temperature and humidity recorder is recorded every minute, including temperature value, relative humidity value and corresponding time stamp. The data is transmitted to a mobile terminal for storage through Bluetooth. After the healthy leaf arrives at the laboratory, the temperature and humidity recorder data is checked to confirm that the environmental parameters during transportation meet the requirements before subsequent processing. The pre-cooling constant temperature transport box is disinfected with 75% ethanol before and after each use. After disinfection, it is left to stand for 30 minutes to ensure complete evaporation of ethanol. The temperature and humidity recorder is calibrated once a week using a standard temperature and humidity source. The calibration record is saved for at least three years. The crop variety, growth stage, collection time and field location information are recorded synchronously when the healthy leaf is collected.

[0061] S1.2, immerse the healthy leaf in ultrapure water and clean it with an ultrasonic cleaner to remove surface contaminants. Dry the leaf surface with high-purity nitrogen gas to obtain a dried healthy leaf.

[0062] Further, immerse the healthy leaf in ultrapure water with a resistivity of ≥18.2 MΩ·cm. Use an ultrasonic cleaner with a frequency of 40 kHz and a power of 100 W to clean for 30 seconds. Maintain the water temperature at 25±1°C during the ultrasonic process. Place the cleaned healthy leaf on a sterile operation table and use high-purity nitrogen gas with a purity of 99.999% to blow at a 45-degree angle from 10 cm away from the leaf surface at a pressure of 0.2 MPa. Control the nitrogen gas flow rate at 5 L / min during the blowing process. Continue blowing until there is no obvious water film reflection on the leaf surface. Place the blown healthy leaf in a constant temperature and humidity environment (25±1°C, 60±5% RH) for 5 minutes. Confirm that there is no visible water stain on the surface to obtain a dried healthy leaf. Rinse the ultrasonic cleaner with ultrapure water three times before and after each use. Pass the high-purity nitrogen gas through a 0.22 μm filter to sterilize before blowing. The dried healthy leaf needs to be placed in the subsequent fixation step within 30 minutes to avoid secondary pollution caused by long-term exposure.

[0063] S1.3, place the dried healthy leaf on a sample stage and fix the edges with a stress-free transparent fixation ring to obtain a standardized leaf sample with a clean surface, stable shape and no stress deformation.

[0064] Further, the dried healthy leaf is placed flat on the central region of an aluminum alloy carrier platform with a surface flatness of ≤0.01 mm, and a stress-free transparent fixing ring with a diameter of 50 mm and a thickness of 2 mm is used to lightly press the edge of the leaf. The contact surface between the fixing ring and the leaf is pre-coated with medical-grade vaseline (e.g., 0.1 g of vaseline) to reduce friction. During the fixing process, the leaf is kept in a natural relaxed state to avoid artificial stretching or folding. After fixing, the distance between the edge of the leaf and the fixing ring is measured using a vernier caliper to ensure that the deviation in each direction is ≤1 mm. After checking that the fixed healthy leaf is free of wrinkles and warping under standard lighting conditions (e.g., 500 lux white light LED light source), a standardized leaf sample with a clean surface, stable shape, and no stress deformation is obtained.

[0065] S2. Based on the standardized leaf sample, the front and back surface gray scale images of the leaf are obtained by a high-resolution scanner. After image stitching and binary processing, a binary image with clear vein network is obtained.

[0066] S2.1. The standardized leaf sample is placed on the glass platform of the high-resolution scanner, and the parameters of the high-resolution scanner are adjusted. The front and back surface gray scale images of the leaf are obtained, and the front and back surface gray scale images of the leaf are registered.

[0067] Further, the standardized leaf sample is placed flat on the central region of the tempered glass platform of the high-resolution scanner. The surface cleanliness of the glass platform reaches the ISO Class 5 cleanliness standard. The high-resolution scanner parameters are set as follows: 2400 dpi optical resolution, 48-bit color depth, no compression TIFF format output, and all automatic enhancement functions are turned off. The front and back surfaces of the standardized leaf sample are scanned respectively. Before each scan, a standard gray scale card is used for white balance calibration. The obtained front and back surface gray scale images of the leaf are registered by SIFT feature point matching algorithm. The number of matching points is not less than 50 pairs, and the registration error is controlled within 5 μm. The registered front and back surface gray scale images of the leaf are stored as 16-bit unsigned integer arrays, with pixel value range of 0-65535. After scanning 50 samples, the high-resolution scanner needs to be verified for optical performance using a special calibration plate. Before each batch of scanning, the glass platform is wiped with dust-free cloth dipped in analytical pure ethanol. During the registration process, feature point pairs with a matching error greater than 2 pixels are removed, and the remaining matching point pairs are evenly distributed in the main vein and lateral vein regions.

[0068] S2.2. The registered front and back surface gray scale images of the leaf are fused into a complete leaf panoramic image by Laplacian pyramid algorithm. The registered front and back surface gray scale images of the leaf are converted into binary images by dynamic Otsu algorithm. A 3x3 circular structure element is used for morphological operation to remove noise points, and a binary image with clear vein network is obtained.

[0069] Further, the registered blade front and back gray scale images are input into the Laplacian pyramid algorithm for multi-scale fusion, the pyramid layer number is set to 5 layers, the Gaussian kernel size is 5x5 pixels, the fusion weight coefficient is 0.5, the complete blade panoramic image after fusion is threshold segmented by the dynamic Otsu algorithm, the algorithm window size is set to 50x50 pixels, the threshold search step is 5 gray levels, the pixel value of the vein in the segmented binary image is 1, and the mesophyll pixel value is 0. The binary image is subjected to morphological operation using a 3x3 circular structure element, first subjected to 3x3 circular structure element erosion operation to remove isolated noise points, and then subjected to 3x3 circular structure element expansion operation to restore the vein morphology, and finally a binary image with clear vein network is obtained.

[0070] S3, skeletonize the binary image with clear vein network, calculate the fractal dimension of the vein by box counting method, and obtain the fractal dimension change rate.

[0071] S3.1, check the connectivity of the binary image with clear vein network, and use Zhang-Suen thinning algorithm to generate a vein skeleton image.

[0072] Further, 8-connected region analysis is performed on the binary image with clear vein network to detect the connectivity state of each vein branch and mark the position of the broken region; after the connectivity check is completed, the binary image with clear vein network is input into the Zhang-Suen thinning algorithm, and two-step iteration is alternately performed: the first step iteration deletes the boundary pixels that satisfy 2≤neighborhood pixel number≤6, connectivity number is 1 and at least one east / south / west / north neighborhood is background, and the second step iteration adjusts the deletion condition to maintain symmetry, until the image no longer changes to generate a vein skeleton image, the skeleton topological integrity is verified after each iteration, the main vein and the connection relationship of the third-order and above lateral veins are preserved, and finally the output vein skeleton image maintains a width of 1 pixel, the branch break error is controlled within 3 pixels, and during the processing of the Zhang-Suen thinning algorithm, the calculation is terminated when the number of pixels deleted by continuous 5 iterations is less than 0.1% of the total number of pixels, the vein skeleton image is stored in 1-bit BMP format, and the skeleton node coordinates and branch level information are attached.

[0073] S3.2, analyze the scale range of the vein skeleton image using box counting method, count the minimum grid number covering the vein skeleton, and calculate the fractal dimension of the vein using the fitted linear regression curve.

[0074] Specifically, the expression is,

[0075] ;

[0076] wherein, is a scale is the minimum grid number required to cover the vein skeleton of a leaf, is the fractal dimension of the vein, is the intercept of the linear regression, is a scale index.

[0077] S3.3, by collecting the same variety of crop health leaf vein fractal dimension under standard irrigation conditions, statistical analysis, get the same variety of healthy sample baseline value.

[0078] Further, under standard irrigation conditions (soil moisture content is maintained at 70-80% of field water capacity), the same variety of crop health leaf sample is not less than 30 pieces in the same growth period, each healthy leaf sample is prepared according to the unified standard after the standardization of leaf sample, the positive and negative gray scale images of the leaf are obtained by high resolution scanner, and the image registration, binary processing and skeletonization processing are carried out, and the box counting method is used to obtain the leaf vein fractal dimension of each healthy leaf sample in the range of 0.1-5mm scale, and the arithmetic mean value of the leaf vein fractal dimension of the remaining healthy leaf sample is obtained as the baseline value of the same variety of healthy sample, and the standard deviation is recorded for subsequent significance test, the crop variety, growth period, sampling time and field location information are recorded synchronously when the healthy leaf sample is collected, all the leaf vein fractal dimension calculation results and metadata are stored, the baseline value of the same variety of healthy sample is updated once a quarter, and the sampling conditions, measurement method and data processing process must be consistent with the previous period.

[0079] S3.4, the leaf vein fractal dimension is compared with the baseline value of the same variety of healthy sample, and the fractal dimension change rate is calculated.

[0080] Specifically, the expression is,

[0081] ;

[0082] is the fractal dimension change rate, is the baseline value of the same variety of healthy sample.

[0083] S4, in the fixed detection area of the standardized leaf sample, the advancing angle and the receding angle are measured continuously by the contact angle measuring instrument, the contact angle hysteresis value is calculated, and the contact angle hysteresis change data is obtained.

[0084] S4.1, mark the detection point at one-third of the distance from the leaf tip of the standardized leaf sample with a laser positioner, place the marked leaf in a constant temperature and humidity chamber for equilibration, calibrate the contact angle measuring instrument with a standard PDMS sheet, inject ultrapure water at the marked detection point, record the profile at the moment of maximum expansion of the droplet, obtain the advancing angle, capture the profile at the moment when the droplet starts to shrink, obtain the receding angle, and calculate the contact angle hysteresis value.

[0085] Specifically, the expression is,

[0086] ;

[0087] wherein, the contact angle hysteresis value, the advancing angle, the receding angle, the temperature compensation coefficient, the temperature deviation.

[0088] It should be noted that the detection point is marked 2 mm away from the main vein at one-third of the distance from the leaf tip of the standardized leaf sample using a laser positioner with a wavelength of 635 nm and a power of <5 mW, and the marking accuracy is ±0.1 mm; the marked standardized leaf sample is placed in a constant temperature and humidity chamber with a temperature of 25±0.5℃ and a relative humidity of 60±2% for equilibration for 30 minutes; the contact angle measuring instrument is calibrated using a standard PDMS sheet with a surface energy of 20.1±0.5 mJ / m², and the calibration error is controlled within ±0.5°; 2 μL of ultrapure water with a resistivity of ≥18.2 MΩ·cm is injected at a rate of 1 μL / s at the marked detection point, a 1000 fps high-speed camera is used to record the droplet dynamics, and the profile image is intercepted when the droplet contact line reaches the maximum expansion radius, and the advancing angle is obtained by fitting through the Young-Laplace equation; continue to suck the droplet back at a rate of 0.5 μL / s until the contact line starts to shrink, and the profile image is intercepted, and the receding angle is obtained in the same way, and finally the contact angle hysteresis value is calculated as the advancing angle minus the receding angle, each detection point is measured 3 times to take the average value, the positioning accuracy of the laser positioner is verified with a standard grid plate before each use, the environmental parameters of the constant temperature and humidity chamber are recorded every minute, the contact angle measuring instrument is flushed with ultrapure water 3 times before and after calibration, and the environmental temperature fluctuation is controlled within ±0.3℃ during the measurement of the advancing angle and the receding angle.

[0089] S4.2, by measuring and statistically analyzing the same growth period healthy leaves after uniform standardization treatment under standard environmental control conditions, the reference value of the same variety healthy sample is obtained.

[0090] Further, under the standard environmental control condition of temperature 25±1℃ and relative humidity 60±5%, at least 30 healthy leaf samples of the same crop species at the same growth stage are collected, the healthy leaf samples are selected from the third to fifth leaf positions of the fully expanded leaves in the middle of the canopy, and all the healthy leaf samples are processed according to the unified standard process, including collection with sterile sampling forceps, pre-cooling in a constant-temperature transport box, ultrasonic cleaning with ultrapure water, high-purity nitrogen blowing and stress-free fixation; the processed healthy leaf samples are scanned by a high-resolution scanner to obtain the front and back gray-scale images of the leaves, after image registration and binarization, the fractal dimension of the leaf veins is calculated by box counting method, the advancing angle and receding angle are measured by a contact angle measuring instrument, and the contact angle hysteresis value is calculated, and the arithmetic mean of the fractal dimension of the leaf veins and the contact angle hysteresis value of the remaining healthy leaf samples is obtained, which is used as the reference value of the fractal dimension of the leaf veins and the reference value of the contact angle hysteresis value of the healthy samples of the same species, respectively, and the information of the crop species, growth stage, sampling time and field location is recorded synchronously when the healthy leaf samples are collected, and all the reference value calculation results are stored in association with the metadata, and the standard environmental control condition is monitored in real time by a calibrated temperature and humidity sensor.

[0091] S4.3, compare the contact angle hysteresis value with the reference value of the healthy samples of the same species to obtain the contact angle hysteresis change data.

[0092] Specifically, the expression is,

[0093] ;

[0094] wherein, is the contact angle hysteresis change data, is the reference value of the healthy samples of the same species.

[0095] S5, input the fractal dimension change rate and the contact angle hysteresis change data into the random forest model to calculate the drought comprehensive index and obtain the drought grade result.

[0096] S5.1, combine the fractal dimension change rate and the contact angle hysteresis change data into a two-dimensional feature vector, and use the RandomForestClassifier of Scikit-learn to perform ten-fold cross-validation training to obtain the trained random forest model.

[0097] Further, the fractal dimension change rate and the contact angle hysteresis change data are combined as a two-dimensional feature vector with a weight of 0.6:0.4, and a RandomForestClassifier of Scikit-learn version 1.2.2 is used, with the number of decision trees set to 500, the maximum depth set to 5, and the minimum leaf sample number set to 3, to perform ten-fold cross-validation training on a feature vector matrix containing at least 300 samples. During the training process, a grid search method is used to optimize the parameters, and the validation set accounts for 20%. Finally, the model with the highest average accuracy is selected as the trained random forest model.

[0098] S5.2, input the two-dimensional feature vector into the pre-trained random forest model to calculate the drought comprehensive index.

[0099] Specifically, the expression is,

[0100] ;

[0101] wherein, is the drought comprehensive index, is the normalized fractal dimension change rate, is the normalized contact angle hysteresis change rate.

[0102] S5.3, based on the ROC curve analysis of the crop physiological response inflection point, an optimal split point is formed by the Youden index to obtain the drought threshold.

[0103] Further, based on the drought comprehensive index data containing at least 500 field measurement samples and the simultaneously obtained crop physiological parameters (such as stomatal conductance, photosynthetic rate), a receiver operating characteristic curve (ROC curve) is drawn, with the horizontal axis representing 1-specificity and the vertical axis representing sensitivity. The crop physiological response inflection point is calculated, the Youden index (Youden index = sensitivity + specificity - 1) corresponding to each candidate drought threshold is calculated, and the drought comprehensive index value corresponding to the maximum Youden index is selected as the optimal split point to obtain the drought threshold.

[0104] S5.4, when the drought comprehensive index is less than the drought threshold, the crop water status is determined to be in the normal range.

[0105] Furthermore, when the drought composite index is less than the drought threshold (e.g., DSI≤0.7), the current crop moisture status is determined to be within the normal range. At this time, the deviation of leaf physiological indicators (e.g., stomatal conductance, photosynthetic rate) from the baseline value of healthy samples of the same variety does not exceed 15%, the leaf vein fractal dimension change rate is ≤5%, and the contact angle hysteresis change rate ΔCAH is ≤8%. The original agronomic management measures are maintained unchanged, and the next test is carried out according to the standard monitoring cycle. During the determination process, the consistency of the fractal dimension change rate and the contact angle hysteresis change rate ΔCAH is checked simultaneously. The two change trends are required to be consistent. Otherwise, the data verification process is triggered. The determination result of the normal state is cross-validated with the field measured soil moisture content data to ensure the reliability of the determination.

[0106] S5.5 When the drought composite index equals the drought threshold, the crop moisture status is determined to be mild drought.

[0107] Furthermore, when the drought composite index calculation result equals the drought threshold (e.g., DSI=0.7), the current crop moisture status is determined to be mild drought. At this time, leaf physiological indicators show that stomatal conductance decreases by 20-30%, photosynthetic rate decreases by 15-25%, leaf vein fractal dimension change rate ΔDf is in the range of 5-15%, and contact angle hysteresis change rate ΔCAH is 8-15%. The system automatically generates a mild drought warning and suggests retesting within 24 hours. At the same time, the fertilization plan is adjusted (e.g., reducing nitrogen fertilizer application by 10-15%). During the determination process, the synergy between fractal dimension change rate and contact angle hysteresis change rate is verified. If the data is abnormal, the quality control review process is initiated. The mild drought determination result is cross-validated with field transpiration meter data.

[0108] S5.6 When the drought composite index is greater than the drought threshold, the crop moisture status is determined to be severe drought.

[0109] Furthermore, when the calculated drought composite index is greater than the drought threshold determined by ROC curve analysis and Youden index optimization (e.g., DSI>0.85), the current crop water status is determined to be severe drought. At this time, leaf physiological indicators show that stomatal conductance decreases by more than 50%, photosynthetic rate decreases by more than 40%, leaf vein fractal dimension change rate is >15%, and contact angle hysteresis change rate is >20%, immediately triggering a severe drought alarm and automatically generating irrigation instructions.

[0110] S6. Generate a visualization report based on the drought level results.

[0111] S6.1. Based on the fractal dimension change rate, perform JET color level mapping on the binarized image with a clear leaf vein network to obtain a degradation heatmap with a color scale, and use Matplotlib to plot the time series curves of the advance angle and the retreat angle.

[0112] Further, the binary image with clear vein network and the corresponding fractal dimension change rate data are input into the applyColorMap function of the OpenCV library, the JET color scale mode is selected for color mapping, and a pseudo-color heat map is generated, in which blue (RGB 0, 0, 255) represents ΔDf≤5%, yellow (RGB 255, 255, 0) represents 5%<ΔDf≤15%, and red (RGB 255, 0, 0) represents ΔDf>15%; a standard color scale is added to the heat map to mark the corresponding relationship between the ΔDf value and the color, and the pyplot module of the Matplotlib library is used to draw a time series curve with marked points, with the measurement time points of the advancing angle and the receding angle as the horizontal coordinates and the angle values as the vertical coordinates, the advancing angle is represented by a green solid line, and the receding angle is represented by a red dashed line, and the change trend of the contact angle hysteresis CAH is marked above the curve.

[0113] S6.2, based on the drought comprehensive index, the drought threshold and the irrigation event mark, a vector trend chart is generated.

[0114] Further, based on the historical drought comprehensive index time series data, the pre-set drought threshold and the irrigation event log, a vector format trend chart is created using the Matplotlib library, with time as the horizontal axis (for example, date scale) and drought comprehensive index as the vertical axis (range 0-1.0), a line chart is drawn to represent the drought comprehensive index change trend; two horizontal dashed lines are added to the chart to mark the mild drought threshold (for example, orange dashed line y=0.7) and the severe drought threshold (for example, red dashed line y=0.85); the irrigation event occurrence time point and the corresponding water amount are marked with green triangle symbols, the trend chart adds coordinate axis labels (for example, time (days) and drought comprehensive index) and a title (for example, wheat jointing stage drought trend), and is saved as a scalable vector graphics (SVG) format, with metadata records of crop variety, monitoring period and chart generation date, linear interpolation is used for missing data in the drawing process, and outliers (for example, DSI>1.0) are excluded after review, the trend chart color scheme is consistent with the heat map, blue represents normal, yellow represents mild drought, and red represents severe drought.

[0115] S6.3, the degradation heat map with color scale, the time series curve of advancing angle and receding angle and the vector trend chart are inserted into the HTML5 template through Jinja2 to generate a visual report.

[0116] Further, the Jinja2 template engine is used to insert the PNG file of the degradation thermodynamic map with color scale, the PNG file of the time series curve of the advancing angle and the receding angle, and the SVG file of the vector trend map into a predefined HTML5 report template, the template including a title area (displaying the crop variety, the detection date, and the geographical location), a data visualization area (displaying the three graphs side by side), and a conclusion area (summarizing the drought grade and the agronomic suggestion), adding an interactive pop-up box in the HTML5 template for the degradation thermodynamic map to display the accurate value of the change rate of the fractal dimension of the leaf vein at the position of the mouse hovering, adding a legend for the time series curve to explain the corresponding relationship between the advancing angle, the receding angle, and the contact angle hysteresis value CAH, and embedding clickable irrigation event marker points in the vector trend map, which display the specific irrigation time and water volume after being clicked, to generate a visual report.

[0117] The embodiment also provides a monitoring system for identifying crop drought based on leaf morphology, including:

[0118] A standardization module selects a fully expanded healthy leaf, washes the leaf with ultrapure water, dries the leaf with nitrogen, and fixes the leaf on a standard carrier table to obtain a standardized leaf sample.

[0119] A quantification module obtains a grayscale image of the front and back of the leaf based on the standardized leaf sample through a high-resolution scanner, and obtains a binary image with a clear leaf vein network through image stitching and binarization processing.

[0120] A detection module performs skeletonization processing on the binary image with a clear leaf vein network, calculates the fractal dimension of the leaf vein by using box counting, and obtains the change rate of the fractal dimension.

[0121] A fusion module continuously measures the advancing angle and the receding angle through a contact angle measuring instrument in a fixed detection area of the standardized leaf sample, calculates the contact angle hysteresis value, and obtains the change data of the contact angle hysteresis.

[0122] An execution module inputs the change rate of the fractal dimension and the change data of the contact angle hysteresis into a random forest model, calculates a drought comprehensive index, obtains a drought grade result, and generates a visual report according to the drought grade result.

[0123] The embodiment also provides a computer device suitable for the monitoring method for identifying crop drought based on leaf morphology, including a memory and a processor, the memory is used to store computer executable instructions, and the processor is used to execute the computer executable instructions to implement the monitoring method for identifying crop drought based on leaf morphology as proposed in the above embodiment.

[0124] The computer device can be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen and an input device connected by a system bus. The processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for running the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is configured to communicate with external terminals in a wired or wireless manner. The wireless manner can be achieved by WIFI, an operator network, NFC (Near Field Communication) or other technologies. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer overlaid on the display screen, or a key, trackball or touchpad arranged on the shell of the computer device, or an external keyboard, touchpad or mouse, etc.

[0125] The embodiment also provides a storage medium having a computer program stored thereon, the program being executed by a processor to implement the monitoring method for identifying crop drought based on leaf morphology according to the above embodiment. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as a static random access memory (SRAM), an electrically erasable programmable read-only memory (EEPROM), an erasable programmable read-only memory (EPROM), a programmable read-only memory (PROM), a read-only memory (ROM), a magnetic storage, a flash memory, a magnetic disk or an optical disk.

[0126] In summary, the present application constructs a crop drought monitoring system through innovative leaf vein fractal dimension analysis and contact angle dynamic measurement. In terms of leaf vein analysis, skeletonization processing and box counting method are used to accurately quantify the topological structure of the leaf vein network, so that early changes in soil water content can be detected. In terms of surface property detection, the dynamic changes of the contact angle are captured by high-speed photography, and temperature compensation is combined to form a technical closed loop with the steps of standardized sample preparation, high-resolution scanning, random forest model fusion and visual report generation, so that multi-dimensional collaborative monitoring of micro morphology and surface properties is achieved, and the problem of single monitoring dimension in the prior art is solved.

[0127] It should be noted that the above examples are only used to illustrate the technical solutions of the present application but not limit the present application. Although the present application is described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or equivalently replaced, without departing from the spirit and scope of the technical solutions of the present application, which should be covered in the scope of the claims of the present application.

Claims

1. A method for monitoring crop drought based on leaf morphology identification, characterized in that: include, Fully unfolded healthy leaves were selected, washed with ultrapure water and dried with nitrogen, and then fixed on a standard stage to obtain standardized leaf samples. Based on standardized leaf samples, grayscale images of the front and back of the leaves were obtained using a high-resolution scanner. After image stitching and binarization, a binarized image with a clear leaf vein network was obtained. Skeletonization processing was performed on the binarized image with a clear leaf vein network, and the fractal dimension of the leaf vein was calculated using the box counting method to obtain the rate of change of fractal dimension; In a fixed testing area of ​​a standardized blade sample, the advance angle and retreat angle are continuously measured using a contact angle measuring instrument, and the contact angle hysteresis value is calculated to obtain contact angle hysteresis change data. Input the fractal dimension change rate and contact angle hysteresis change data into the random forest model to calculate the drought comprehensive index, obtain the drought level results, and generate a visualization report based on the drought level results. The binarized image with a clear leaf vein network is skeletonized, and the fractal dimension of the leaf veins is calculated using box counting to obtain the rate of change of fractal dimension. Includes the following steps, The connectivity of the binarized image with a clear leaf vein network is checked. The binarized image with the clear leaf vein network is then processed iteratively using the Zhang-Suen thinning algorithm to generate a leaf vein skeleton image. The scale range of the leaf vein skeleton image was analyzed using box counting, the minimum number of grids covering the leaf vein skeleton was counted, and the fractal dimension of the leaf vein was calculated using the fitted linear regression curve. By collecting leaf vein fractal dimension data from healthy leaves of the same crop variety under standard irrigation conditions and performing statistical analysis, a baseline value for healthy samples of the same variety was obtained. The fractal dimension of leaf veins was compared with the baseline value of healthy samples of the same variety, and the rate of change of fractal dimension was calculated. In a fixed testing area of ​​a standardized blade sample, the advance angle and retreat angle are continuously measured using a contact angle measuring instrument. The contact angle hysteresis value is calculated to obtain contact angle hysteresis change data, including the following steps. At one-third of the distance from the leaf tip, a laser locator was used to mark the detection point on the standardized leaf sample. The marked leaf was placed in a constant temperature and humidity chamber for equilibration. The contact angle measuring instrument was calibrated using a standard PDMS sheet. Ultrapure water was injected at the marked detection point, and the profile of the droplet at the moment of maximum expansion was recorded to obtain the advance angle. The profile of the droplet at the moment of beginning to contract was captured to obtain the retreat angle, and the contact angle hysteresis value was calculated. By measuring and statistically analyzing healthy leaves of the same growth period under standard environmental control conditions, a baseline value for healthy samples of the same variety was obtained. The contact angle hysteresis value was compared with the baseline value of healthy samples of the same variety to obtain the contact angle hysteresis change data.

2. The method for monitoring crop drought based on leaf morphology identification according to claim 1, characterized in that: Fully expanded, healthy leaves were selected, washed with ultrapure water, dried with nitrogen, and then fixed on a standard stage to obtain standardized leaf samples. This process included the following steps: Using sterile sampling forceps, select fully expanded healthy leaves from the middle of the crop canopy. Place the collected healthy leaves into a pre-cooled constant temperature transport box, and place a temperature and humidity recorder inside the pre-cooled constant temperature transport box to monitor the environment. Healthy leaves are immersed in ultrapure water, cleaned with an ultrasonic cleaner to remove surface contaminants, and then dried with high-purity nitrogen to obtain dried healthy leaves. The dried healthy leaf blades are laid on a stage, and the edges are fixed with a stress-free transparent fixing ring to obtain standardized leaf samples with clean surface, stable morphology and no stress deformation.

3. The method for monitoring crop drought based on leaf morphology identification according to claim 2, characterized in that: Based on the standardized leaf samples, the gray scale images of the front and back surfaces of the leaves are obtained by a high-resolution scanner, and through image stitching and binary processing, a binary image with clear vein network is obtained, including the following steps, Place the standardized leaf sample on the glass platform of the high-resolution scanner, and adjust the parameters of the high-resolution scanner to obtain the gray scale images of the front and back surfaces of the leaf, and register the gray scale images of the front and back surfaces of the leaf; Through the Laplacian pyramid algorithm, the registered gray scale images of the front and back surfaces of the leaf are fused into a complete leaf panoramic image, the dynamic Otsu algorithm is used to convert the registered gray scale images of the front and back surfaces of the leaf into binary images, and the 3*3 circular structure element is used for morphological operation to remove noise points, and a binary image with clear vein network is obtained.

4. The method for monitoring crop drought based on leaf morphology identification as claimed in claim 1, wherein: The fractal dimension change rate and the contact angle hysteresis change data are input into the random forest model to calculate the drought comprehensive index, and the drought grade result is obtained, including the following steps, The fractal dimension change rate and the contact angle hysteresis change data are combined into a two-dimensional feature vector, and the RandomForestClassifier of Scikit-learn is used for ten-fold cross-validation training to obtain the trained random forest model; The two-dimensional feature vector is input into the pre-trained random forest model to calculate the drought comprehensive index; Based on the ROC curve analysis of the measured data in the field, the inflection point of the crop physiological response is combined to form an optimized segmentation point through the Youden index to obtain a drought threshold value; When the drought comprehensive index is less than the drought threshold value, it is determined that the crop water state is in the normal range; When the drought comprehensive index is equal to the drought threshold value, it is determined that the crop water state is in the mild drought range; When the drought comprehensive index is greater than the drought threshold value, it is determined that the crop water state is in the severe drought range.

5. The method for monitoring crop drought based on leaf morphology identification according to claim 4, characterized in that: According to the drought grade result, a visual report is generated, including the following steps, Based on the fractal dimension change rate, the binary image with clear vein network is subjected to JET color scale mapping to obtain a degradation thermodynamic map with a color scale, and the time series curves of the advancing angle and the receding angle are drawn using Matplotlib; Based on the drought comprehensive index, the drought threshold value and the irrigation event mark, a vector trend chart is generated; The degradation thermodynamic map with color scale, the time series curves of the advancing angle and the receding angle, and the vector trend chart are inserted into an HTML5 template through Jinja2 to generate a visual report.

6. A monitoring system for identifying crop drought based on leaf morphology, based on the monitoring method for identifying crop drought based on leaf morphology according to any one of claims 1 to 5, characterized in that: including, The standardized module selects fully expanded healthy leaf blades, which are cleaned with ultrapure water and dried with nitrogen gas, and then fixed on a standard stage to obtain standardized leaf samples; The quantification module obtains the gray scale images of the front and back surfaces of the leaves based on the standardized leaf samples through a high-resolution scanner, and through image stitching and binary processing, a binary image with clear vein network is obtained; The detection module performs skeletonization processing on the binary image with clear vein network, calculates the vein fractal dimension using box counting method, and obtains the fractal dimension change rate; The fusion module measures the advancing angle and receding angle by the contact angle measuring instrument in the fixed detection area of the standardized leaf sample, calculates the contact angle hysteresis value, and obtains the contact angle hysteresis change data; The execution module inputs the fractal dimension change rate and the contact angle hysteresis change data into a random forest model, calculates a drought comprehensive index, obtains a drought grade result, and generates a visual report according to the drought grade result. 7.A computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the computer device is characterized in that: The processor executes the computer program to implement the steps of the monitoring method for identifying crop drought based on leaf morphology according to any one of claims 1-5.

8. A computer readable storage medium having stored thereon a computer program, characterized in that: The computer program is executed by the processor to implement the steps of the monitoring method for identifying crop drought based on leaf morphology according to any one of claims 1-5.

Images