Method for decision of water and fertilizer management during flowering and fruit setting period of citrus

Abstract

The application discloses a method for decision-making of water and fertilizer management in the flowering and fruit setting period of citrus, and is characterized in that the method comprises the following steps: collecting crown layer images of the citrus trees in the daytime, air temperature and soil moisture content data, obtaining spring shoot growth states, high temperature indexes and soil moisture abundance and deficiency indexes; comprehensively evaluating daily environmental stress conditions based on the three indexes, and making decisions on supplementing appropriate irrigation management; comprehensively evaluating the high temperature stress degree in the flowering period, and making decisions on normal application of flower fertilizer in the fruit setting period, increase of foliar spraying fertilizer or increase of foliar spraying fertilizer and hormone management measures. The application can effectively solve the problems that the decision-making of water and fertilizer management in the flowering and fruit setting period of conventional citrus trees relies on artificial experience and lacks automatic decision-making, and improves the intelligent level of the cultivation and management of the citrus trees.

Description

Technical Field

[0001] This invention relates to the field of intelligent fruit tree cultivation and management, specifically to a method for monitoring growth stress and environmental conditions to make decisions on water and fertilizer management during the flowering and fruit setting period of citrus trees, belonging to the field of smart agricultural management technology. Background Technology

[0002] Citrus is a fruit tree with a large flowering volume but a low natural fruit set rate. The first physiological fruit drop occurs 1-2 weeks after flowering, accounting for approximately 60% of the total fruit drop. During the flowering to fruit set period, factors such as soil drought or excessive moisture, high temperatures or freezing, inadequate tree nutrition, imbalances in endogenous plant hormones, and unscientific use of exogenous hormones exacerbate physiological fruit drop, hindering later management and yield formation. In traditional cultivation management, decisions regarding the application of post-flowering fertilizer, foliar fertilizer, irrigation, and drainage are made manually based on tree vigor, flowering volume, and fruit set. Due to differences in professional knowledge and planting management experience among different managers, many unreasonable and inappropriate management decisions and operations occur. Utilizing IoT devices to monitor growth status and environmental conditions, combined with appropriate algorithmic decision-making methods, offers better rationality and accuracy in standardized planting management. This addresses the reliance on experience in traditional cultivation management and represents a key technology for the development of smart agriculture.

[0003] Currently, automated decision-making technologies for crop water and fertilizer management mainly include: 1) making precise variable fertilization decisions for citrus trees based on rapid detection results of major soil nutrient content; 2) using fuzzy control algorithms to make decisions based on growth and development processes and soil moisture conditions, and executing different water and fertilizer programs in an integrated water and fertilizer machine; 3) using machine vision to identify fruit tree phenology, analyze and diagnose nutrient deficiencies, and make decisions on water and fertilizer management; 4) establishing expert systems based on knowledge and databases in fruit tree cultivation, and making decisions on water and fertilizer application based on similarity of rules, conditions, and fertilization strategies. Existing management decisions are applicable to scenarios involving the management of major nutrients such as nitrogen, phosphorus, and potassium in fruit trees, but specific implementation depends on different observation data; however, they lack precise water and fertilizer management decisions for situations where high temperatures may cause stress during the flowering and fruit-setting period, making it difficult to achieve the goal of protecting flowers and fruits in citrus. To achieve more precise automated decision-making for water and fertilizer management during the flowering and fruit-setting period, it is necessary to formulate monitoring indicators, monitoring periods, and decision-making criteria that affect water and fertilizer decisions during this period. Summary of the Invention

[0004] The purpose of this invention is to provide a method for decision-making regarding water and fertilizer management during the flowering and fruit-setting period of citrus trees. This method utilizes IoT monitoring equipment to collect temporal images of the tree canopy, air temperature, and soil moisture data. It identifies growth stress conditions, determines potential temperature and soil moisture indicators that may cause stress, and comprehensively assesses the daily high-temperature stress during the flowering-fruit-setting period. In response to high-temperature stress, it makes decisions to supplement irrigation management accordingly. The method also statistically analyzes the degree of high-temperature stress encountered during the flowering period and, based on the total stress level, decides whether to increase the application of foliar fertilizer and exogenous hormones during the fruit-setting period. Through multiple decisions regarding irrigation, fertilizer, and pesticide application, this method achieves a scientific and rational response to high-temperature stress during the flowering and fruit-setting period, improves plant resistance, and ultimately helps to preserve flowers and fruits in citrus trees.

[0005] To achieve the above objectives, the present invention adopts the following technical solution:

[0006] A method for decision-making regarding water and fertilizer management during the flowering and fruit-setting period of citrus includes the following steps:

[0007] Collect daytime canopy images, air temperature, and soil moisture content data of citrus trees to obtain spring shoot growth status, high temperature index, and soil moisture abundance / deficiency index;

[0008] The daily environmental stress situation is assessed by comprehensively evaluating three indicators, and decisions are made to supplement appropriate irrigation management.

[0009] A comprehensive assessment of the degree of high-temperature stress during the flowering period is used to determine whether to apply fertilizer normally during the fruit setting period, increase foliar fertilizer application, or increase foliar fertilizer application and hormone management measures.

[0010] A further step is:

[0011] The spring shoot growth status includes detecting spring shoot targets using daytime canopy images and deep learning methods. The growth status of spring shoots in the canopy image is identified based on the target window. The growth status includes two states: normal and wilted. The sequence of spring shoot growth status at N time points each day is as follows: The two states are represented by numerical values, with the value recorded as 0 for the normal state and 1 for the wilted state.

[0012] A further step is:

[0013] When more than 50% of the leaves in the canopy image have a leaf tilt angle ≥90°, they are considered wilted; the rest are considered normal.

[0014] A further step is:

[0015] The aforementioned high-temperature index is obtained by comparing the air temperature at each moment with a set high-temperature threshold to obtain a high-temperature index sequence corresponding to the growth status at N moments each day. Its state includes two states: present and absent. The two states are represented by numerical values, where the value is 0 when the state is absent and 1 when the state is present.

[0016] That is, when the temperature is below the high temperature threshold at a certain moment, the state is none; when the temperature is not below the high temperature threshold at a certain moment, the state is present.

[0017] A further step is:

[0018] The set high temperature threshold is 32℃.

[0019] A further step is:

[0020] The aforementioned soil moisture abundance / deficit index is obtained by comparing the acquired soil moisture content with a set soil moisture abundance / deficit threshold to obtain a soil moisture abundance / deficit index sequence for N times each day. The status includes three states: abundant water, normal, and scarce water. Each state of the three indicators is represented by a numerical value: -1 for abundant water, 0 for normal water, and 1 for scarce water.

[0021] A further step is:

[0022] A comprehensive assessment of daily environmental stress using three indicators is used to determine appropriate supplementary irrigation management, including:

[0023] Based on indicators at each time point , , The summation of the state values ​​is as follows:

[0024] ,

[0025] Assess whether the new buds, leaves, and flowers of fruit trees are suffering from high-temperature stress under this environment; when t i When the sum of the state values ​​at time t is greater than the threshold, it means that t i When a stress occurs, the stress duration is recorded as 1. When the cumulative daily stress duration exceeds the threshold, the decision is to add quantitative sprinkler irrigation based on the stress level.

[0026] A further step is:

[0027] when Increase irrigation by 15% when Increase irrigation by 30% at that time.

[0028] A further step is:

[0029] The comprehensive assessment of the degree of high-temperature stress during the flowering period includes using temporal canopy images and deep learning methods to detect flower targets and determine the flowering period; when flower buds are detected, the initial date is determined. The end of the flowering period is determined when a significant reduction or disappearance of blooming flowers is detected. ,statistics to Degree of coercion at various moments during the period Obtain the total degree of coercion The specific formula is as follows:

[0030] ,

[0031] Where temp is the average temperature at all times, ( ) + This means that positive values ​​are retained, and values ​​less than 0 are returned as 0.

[0032] ;

[0033]

[0034] in, The high temperature threshold; The attenuation coefficient is determined by using historical air temperature as input and field observation stress characterization data as a benchmark. The nonlinear least squares fitting algorithm is used for optimization to finally determine the λ value with the highest matching degree between the calculated value and the actual observed value, which is usually 0.5-2.

[0035] Based on the total degree of stress ,when When <1, the assessment is normal; when 1≤ <15:00, assessed as mild stress, when When the value is greater than 15, it is assessed as severe stress. The corresponding decisions are: normal application of fertilizer after flowering, increased application of foliar fertilizer (0.2% potassium dihydrogen phosphate + 0.1% boron-zinc micronutrient fertilizer), or increased application of foliar fertilizer (0.2% potassium dihydrogen phosphate + 0.1% boron-zinc micronutrient fertilizer) and regulation with exogenous hormones (5-10 ppm brassinolide or 20 ppm gibberellin).

[0036] The beneficial effects that the method for decision-making on water and fertilizer management during the flowering and fruit-setting period of citrus disclosed in this application may bring include, but are not limited to:

[0037] This invention monitors images of the citrus tree canopy and the surrounding environment, including temperature and soil moisture. Deep learning and machine vision technologies are used to perceive the state of newly sprouted buds and leaves on spring shoots, which are sensitive to stress. A threshold method is used to quickly determine the state of the main environmental factors causing stress. By integrating the stress state of buds and leaves perceived by machine vision, along with high temperature and soil moisture abundance / deficiency, the daily stress situation is assessed, providing an accurate basis for management decisions. Compared to assessment and decision-making based on transpiration and evapotranspiration models, this method, which integrates machine vision and perception of key environmental factors, avoids the need to input meteorological and radiation flux observation data and set parameters for different crops and stages during the assessment and decision-making process. It enables intuitive and accurate assessment of stress occurrence and rapid decision-making in response to short-term high-temperature stress, thereby reducing the impact of high-temperature stress during the flowering and fruit-setting period.

[0038] Compared with existing water and fertilizer decision-making methods, the method of this invention comprehensively utilizes machine vision and deep learning to obtain high-temperature sensitive flowering period information. It addresses the problems of poor flower quality, incomplete flower development, and poor fertilization caused by high temperatures, which increase flower and fruit drop. It more accurately assesses the overall degree of high-temperature stress during this period. It not only makes decisions based on phenological stages but also adjusts fertilizer and pesticide management based on the degree of stress during that period, achieving more reasonable water, fertilizer, and pesticide management decisions for flower and fruit preservation, and providing a foundation for high-quality and high-yield management in the later stages. Attached Figure Description

[0039] Figure 1 This is a flowchart illustrating the decision-making process for water and fertilizer management during the flowering and fruit-setting period of citrus trees, according to an embodiment of the present invention.

[0040] Figure 2 This is a schematic diagram illustrating the decision-making principle of water and fertilizer management during the flowering and fruit-setting period of citrus trees according to an embodiment of the present invention.

[0041] Figure 3 The monitoring effect of indicators for water and fertilizer management decision-making during the flowering and fruit setting period of citrus trees according to an embodiment of the present invention;

[0042] Figure 4 This invention illustrates the effect of irrigation decision-making on citrus trees under high temperature stress during the flowering and fruit setting period according to one embodiment of the present invention.

[0043] Figure 5 This invention provides an evaluation of the degree of high-temperature stress during the flowering and fruit-setting period of citrus trees according to one embodiment of the present invention. Detailed Implementation

[0044] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0045] Conversely, this application covers any substitutions, modifications, equivalent methods, and schemes made within the spirit and scope of this application as defined in the claims. Furthermore, to provide the public with a better understanding of this application, certain specific details are described in detail below. However, this application can be fully understood by those skilled in the art even without these detailed descriptions.

[0046] The following will provide a detailed description of a method for decision-making regarding water and fertilizer management during the flowering and fruit setting period of citrus, as described in the embodiments of this application.

[0047] Example 1

[0048] like Figure 1 As shown, a method for decision-making regarding water and fertilizer management during the flowering-fruit setting period of citrus includes the following steps:

[0049] Step 1: Collect images of the citrus tree canopy, temperature, and soil moisture content data to monitor the daytime growth status of spring shoots, high temperature indicators, and soil moisture abundance / deficiency indicators. Using IoT cameras, canopy images are collected at various times during the day. Based on the color and shape characteristics of newly sprouted leaves and buds of citrus spring shoots, a deep learning model is used to train and fine-tune parameters to identify and detect spring shoot targets. The location and distribution of the spring shoot target window are obtained from the first canopy image of each day. This detection window is used as the region of interest to identify the daytime growth status of spring shoots within the region. Straight leaves indicate a normal state, while drooping leaves indicate a wilted state, resulting in the spring shoot growth status at N times each day. Using the temperature at each moment and the high temperature threshold The comparison shows that temperatures below a threshold indicate no high-temperature state, while temperatures equal to or above the threshold indicate a high-temperature state, thus obtaining the time sequence of whether or not high temperatures exist at each time point of the day. ; Utilizing soil moisture content at each moment and abundance / deficiency thresholds ( , , By comparing these data, we can obtain a sequence of soil water-abundant, normal, or water-deficient states at each time point. The status of the three indicators is expressed numerically.

[0050] Step 2: Assess the daily environmental stress based on three indicators: spring shoot growth status, high temperature index, and soil moisture abundance / deficiency index, and make irrigation management decisions accordingly. (Based on the indicators at each time point...) , , The sum of the state values ​​is used to assess whether the new buds, leaves and flowers of fruit trees are under high temperature stress under the given temperature and soil moisture conditions. When the state value is greater than the threshold, stress occurs and the stress duration is recorded as 1. When the cumulative stress duration is greater than the threshold, the decision is made to increase irrigation to reduce the subsequent possible high temperature stress.

[0051] Step 3: Comprehensively assess the degree of stress during the flowering period to determine the application and management of foliar fertilizers and exogenous hormones during the fruit-setting period. Daily canopy images and deep learning methods are used to determine the start and end of the flowering period. White flower buds and blooming flowers are identified based on flower color and trait characteristics. The initial date is determined when flower buds appear. The end of the flowering period is determined when the number of blooming flowers decreases significantly or disappears. ;statistics to Degree of coercion at various moments during the period Obtain the total degree of coercion The specific formula is as follows:

[0052] ,

[0053] Where temp is the average temperature at all times, ( ) + This means that positive values ​​are retained, and values ​​less than 0 are returned as 0.

[0054] ;

[0055]

[0056] Based on the overall stress level, assess the stress levels as normal, mild, and severe, and decide on appropriate measures such as applying normal fertilizer after flowering, increasing foliar fertilizer application, and hormone management.

[0057] Example 2

[0058] The present invention will be further described in detail below with reference to a more specific embodiment.

[0059] In a standardized citrus orchard equipped with integrated water and fertilizer management systems, an IoT sensor system is used to monitor the citrus trees and their environment, collecting temporal images of the canopy, air temperature, and soil moisture. This embodiment uses this data to identify and assess daily environmental stress during the flowering-fruit setting period, making decisions on irrigation management after encountering high-temperature stress, sensing the degree of high-temperature stress during flowering, and determining whether to apply foliar fertilizers and exogenous hormones during the fruit setting period based on the total stress level. Specifically, according to… Figure 2 The principle of water and fertilizer management during the flowering and fruit setting period of citrus is to use computer vision technology and deep learning methods to process time-series image data, obtain the stress status of spring shoots and flowering time, perceive stress indicators based on ambient temperature and soil moisture, and comprehensively judge the degree of short-term stress and total stress by combining visual perception and environmental indicators, and make decisions on sprinkler irrigation for cooling, foliar fertilizer application and exogenous hormone management.

[0060] Step 1: Monitor the daytime growth stress status, high temperature index, and soil moisture abundance / deficiency index of spring shoots. Using canopy images taken at 10 times between 8:00 AM and 5:00 PM, a deep learning model is used to train and fine-tune parameters based on the color and shape characteristics of newly emerging leaves of citrus spring shoots to identify and detect spring shoot targets. The location and distribution of the spring shoot target window are obtained from the canopy images taken at 8:00 AM each day. Using this detection window as the region of interest, the daytime growth status of spring shoots within the region is identified, with 0 and 1 representing normal and wilted states, respectively, thus obtaining the spring shoot growth status at 10 times each day. Using the temperature at each moment and the high temperature threshold ( By comparing the results, and using 0 and 1 to represent the absence and presence of the high-temperature state respectively, a sequence of the presence and absence of the high-temperature state at each time point during the day is obtained. The soil moisture content obtained by the sensor at each time point was compared with the abundance / deficiency threshold, and -1, 0, and 1 were used to represent abundant soil moisture, respectively. ),normal( ) or lack of water ( By comparing the states, the time series of soil moisture abundance and scarcity states can be obtained. The monitoring results of the three indicators are shown in the figure. Figure 3 .

[0061] Step 2: Assess the daily environmental stress based on three indicators: spring shoot growth status, high temperature index, and soil moisture abundance / deficiency index, and make irrigation management decisions accordingly. (Based on the indicators at each time point...) , , The sum of state values ​​is used to assess whether the new buds, leaves, and flowers of fruit trees are under high-temperature stress under the given temperature and soil moisture conditions. A state value greater than 2 indicates stress, meaning that sensitive buds and leaves show signs of stress and that one or more environmental indicators triggering stress are present, or two such indicators are present simultaneously. The stress duration is recorded as 1. When the cumulative stress duration exceeds 3, a decision is made to increase irrigation to mitigate potential subsequent high-temperature stress. Daily high-temperature stress assessments and decision-making effectiveness are detailed in [link to relevant documentation]. Figure 4 .

[0062] Step 3: Using daily canopy images and deep learning methods, determine the start and end of the flowering period. Identify white flower buds and blooming flowers based on flower color and trait characteristics. Determine the initial date when flower buds appear. The flowering period ends on April 10th, when the blooming flowers disappear. April 19th, according to to During this period, there were three days of high temperatures. Based on the high-temperature stress response curve of citrus fruits, and by fitting local historical stress data, the attenuation coefficient was determined to be 1.0, and the total stress level was calculated. =10.95; Based on 0 being no stress, 1-15 being mild stress, and greater than 15 being moderate to severe stress, the degree of high temperature stress during this period was assessed as mild, and the decision was to apply normal flowering fertilizer and increase foliar fertilizer application; the effect of stress degree assessment is shown in 5.

[0063] This embodiment can make good use of the Internet of Things and automatic algorithms to make scientific decisions on water and fertilizer management during the flowering and fruit setting period of citrus trees under stress, which has great practical value.

[0064] Although the present invention has been described herein with reference to illustrative embodiments, the above embodiments are merely preferred embodiments of the present invention, and the implementation of the present invention is not limited to the above embodiments. It should be understood that those skilled in the art can devise many other modifications and implementations, which will fall within the scope and spirit of the principles disclosed in this application.

Claims

1. A method for decision-making on water and fertilizer management during the flowering-fruit setting period of citrus, characterized in that... Includes the following steps: Collect daytime canopy images, air temperature, and soil moisture content data of citrus trees to obtain spring shoot growth status, high temperature index, and soil moisture abundance / deficiency index; The daily environmental stress situation is assessed by comprehensively evaluating three indicators, and decisions are made to supplement appropriate irrigation management. A comprehensive assessment of the degree of high-temperature stress during the flowering period is used to determine whether to apply fertilizer normally during the fruit setting period, increase foliar fertilizer application, or increase foliar fertilizer application and hormone management measures.

2. The method for decision-making on water and fertilizer management during the flowering-fruit setting period of citrus according to claim 1, characterized in that: The spring shoot growth status includes detecting spring shoot targets using daytime canopy images and deep learning methods, identifying the spring shoot growth status in the canopy image based on the target window, and the growth status includes two states: normal and wilted. The spring shoot growth status sequence for N time points each day is as follows: The two states are represented by numerical values, with the value recorded as 0 for the normal state and 1 for the wilted state.

3. The method for decision-making on water and fertilizer management during the flowering-fruit setting period of citrus according to claim 2, characterized in that: When more than 50% of the leaves in the canopy image have a leaf tilt angle ≥90°, they are considered wilted; the rest are considered normal.

4. The method for decision-making on water and fertilizer management during the flowering-fruit setting period of citrus according to claim 2, characterized in that: The aforementioned high-temperature index is obtained by comparing the air temperature at each moment with a set high-temperature threshold to obtain a high-temperature index sequence corresponding to the growth status at N moments each day. Its state includes two states: present and absent; That is, when the temperature is below the high temperature threshold at a certain moment, the state is none; when the temperature is not below the high temperature threshold at a certain moment, the state is present. The two states are represented by numerical values, where the value is 0 when the state is none and 1 when the state is some.

5. The method for decision-making on water and fertilizer management during the flowering-fruit setting period of citrus according to claim 4, characterized in that: The set high temperature threshold is 32℃.

6. The method for decision-making on water and fertilizer management during the flowering-fruit setting period of citrus according to claim 4, characterized in that: The aforementioned soil moisture abundance / deficit index is obtained by comparing the acquired soil moisture content with a set soil moisture abundance / deficit threshold to obtain a soil moisture abundance / deficit index sequence for N times each day. The status includes three states: abundant water, normal, and scarce water. Each state of the three indicators is represented by a numerical value: when the state is abundant water, the value is recorded as -1; when the state is normal, the value is recorded as 0; and when the state is scarce, the value is recorded as 1.

7. The method for decision-making on water and fertilizer management during the flowering-fruit setting period of citrus according to claim 6, characterized in that: A comprehensive assessment of daily environmental stress using three indicators is used to determine appropriate supplementary irrigation management, including: Based on indicators at each time point , , The summation of the state values ​​is as follows: ， Assess whether the new buds, leaves, and flowers of fruit trees are suffering from high-temperature stress under this environment; when t i When the sum of the state values ​​at time t is greater than the threshold, it means that t i When a stress occurs, the stress duration is recorded as 1. When the cumulative daily stress duration exceeds the threshold, the decision is to add quantitative sprinkler irrigation based on the stress level.

8. The method for decision-making on water and fertilizer management during the flowering-fruit setting period of citrus according to claim 7, characterized in that: when Increase irrigation by 15% when Increase irrigation by 30% at that time.

9. The method for decision-making on water and fertilizer management during the flowering-fruit setting period of citrus according to claim 1, characterized in that: The comprehensive assessment of the degree of high-temperature stress during the flowering period includes using temporal canopy images and deep learning methods to detect flower targets and determine the flowering period; when flower buds are detected, the initial date is determined. The end of the flowering period is determined when a significant reduction or disappearance of blooming flowers is detected. ,statistics to Degree of coercion at various moments during the period Obtain the total degree of coercion The specific formula is as follows: ， in ; , Attenuation coefficient ， in, The high temperature threshold; Using historical air temperature as input and field observation stress characterization data as benchmark, the attenuation coefficient is optimized using a nonlinear least squares fitting algorithm. Finally, the calculated value with the highest matching degree with the actual observed value is determined to be 0.5-2.

10. The method for decision-making on water and fertilizer management during the flowering-fruit setting period of citrus according to claim 9, characterized in that: Based on the total degree of stress ,when When <1, the assessment is normal; when 1≤ <15:00, assessed as mild stress, when When the stress level is greater than 15, it is assessed as severe stress, and the corresponding decisions are to apply fertilizer normally after flowering, increase foliar fertilizer application, or increase foliar fertilizer application and exogenous hormone regulation.

Images