Agricultural assistance strategy matching method and system based on growth stage assessment

Abstract

This application discloses an agricultural auxiliary strategy matching method and system based on growth stage assessment, belonging to the field of data processing technology. By fusing real-time crop images and environmental parameters, and comparing crop morphological features, crop texture features, crop spectral features with standard templates, the current growth stage of the crop is accurately determined and an ideal growth curve is generated. Crop growth parameters are extracted from multi-channel feature maps to construct the actual growth curve. The weighted deviation between the actual growth curve and the ideal growth curve is calculated. When the deviation exceeds the limit, the causes and weights of abnormal nutrient content deviations are comprehensively diagnosed based on the actual nutrient absorption curve and the standard crop absorption curves for each major nutrient. Based on this, the direction of increasing or decreasing the application of target nutrients is determined. The application rate is corrected by calculating adjustment coefficients based on the deviation degree, and the fertilization ratio is dynamically adjusted according to the standard demand ratio and the deviation weight. Finally, instructions are generated and sent to the integrated water and fertilizer equipment to achieve precise closed-loop control.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a method and system for matching agricultural auxiliary strategies based on growth stage assessment. Background Technology

[0002] Precision agriculture is the core pathway to improving crop yield and resource utilization efficiency, and growth stage assessment is a prerequisite for achieving precision agricultural management. Crops have significantly different nutrient and water requirements at different growth stages. Traditional agriculture, relying on experience-based judgment and a fixed calendar, employs extensive management methods that struggle to dynamically match the actual growth status and needs of crops. This leads to deviations in fertilization and irrigation timing, nutrient imbalances, resource waste, and environmental pollution. With the development of sensor, computer vision, and IoT technologies, real-time acquisition of crop images and environmental parameters has become possible, providing a technological foundation for data-driven agricultural decision support.

[0003] However, in the implementation of existing agricultural support strategies, the determination of crop growth stages usually relies on manual observation or calculation based on a single time dimension, lacking the fusion analysis of real-time crop image morphological, textural, and spectral characteristics, making it difficult to accurately match the actual growth process. Growth status monitoring often uses manual sampling to measure parameters such as plant height and leaf area index, resulting in discrete and lagging data that cannot form continuous growth curves. Fertilization decisions are mainly based on pre-set fixed agronomic schemes or soil sampling laboratory test results, which suffer from poor timeliness and inability to dynamically respond to the real-time nutrient needs of crops. When crop growth abnormalities occur, there is a lack of quantitative assessment of the degree of deviation in the growth curve, as well as the ability to diagnose the causes of deviations by combining data on actual crop nutrient absorption efficiency with soil nitrogen, phosphorus, and potassium content.

[0004] Therefore, existing technologies struggle to achieve agricultural support strategies that integrate multi-source information, quantitatively assess deviations, and dynamically match precise fertilization adjustments. Summary of the Invention

[0005] This invention provides a method and system for matching agricultural auxiliary strategies based on growth stage assessment. It can dynamically match fertilization adjustment strategies according to real-time crop images and environmental parameters at different crop growth stages, automatically generate adjustment instructions including target nutrient application rates and fertilizer ratios, and drive integrated water and fertilizer equipment to perform precision fertilization operations, thus achieving an intelligent closed-loop agricultural auxiliary management system. The technical solution provided by this application is as follows: According to a first aspect of this application, an agricultural auxiliary strategy matching method based on growth stage assessment is provided. The method includes: acquiring real-time crop images and environmental parameters of a target crop; determining the current growth stage of the crop based on the similarity between the real-time crop images and a preset standard growth stage morphology template, and acquiring the ideal growth curve of the crop at that growth stage; stitching the real-time crop images and environmental parameters to obtain a crop growth feature vector, and extracting crop growth parameters based on the crop growth feature vector to generate a crop growth curve corresponding to the determined growth stage; calculating the deviation between the crop growth curve corresponding to the determined growth stage and the ideal crop growth curve, and evaluating the degree of deviation through curve parameters; if the degree of deviation exceeds a preset deviation threshold, extracting the crop absorption curves of each major nutrient at the current growth stage, comprehensively diagnosing the cause of the deviation based on the crop absorption curves of each major nutrient and the crop growth curve, dynamically generating fertilization adjustment instructions based on the cause and degree of deviation, and sending them to the integrated water and fertilizer control equipment for execution; if the degree of deviation does not exceed the preset deviation threshold, maintaining the preset standard water and fertilizer management plan, and continuing to acquire real-time crop images and environmental parameters for the next monitoring cycle, continuously monitoring the deviation of the crop growth curve.

[0006] According to another aspect of this application, an agricultural auxiliary strategy matching system based on growth stage assessment is provided. This system is applied to an agricultural auxiliary strategy matching method based on growth stage assessment. The system includes: a data acquisition module for acquiring real-time crop images and environmental parameters of the target crop; a growth stage determination module for determining the current growth stage of the crop based on the similarity between the real-time crop image and a preset standard growth stage morphology template, and obtaining the ideal growth curve of the crop at that growth stage; a growth curve generation module for stitching the real-time crop image and environmental parameters to obtain a crop growth feature vector, extracting crop growth parameters based on the crop growth feature vector, and generating the crop growth curve corresponding to the determined growth stage; and a deviation assessment module. The system comprises four modules: a module for calculating the deviation between the crop growth curve corresponding to the determined growth stage and the ideal crop growth curve, and an evaluation of the deviation degree through curve parameters; a diagnostic decision module for extracting the crop absorption curves of each major nutrient at the current growth stage when the deviation degree exceeds a preset deviation degree threshold, comprehensively diagnosing the cause of the deviation based on the crop absorption curves of each major nutrient and the crop growth curve, and dynamically generating fertilization adjustment instructions based on the cause and degree of deviation; and an execution control module for issuing fertilization adjustment instructions to the integrated water and fertilizer control equipment for execution when the deviation degree does not exceed the preset deviation degree threshold.

[0007] According to another aspect of this application, an agricultural assistance strategy matching device based on growth stage assessment is provided, comprising: one or more processors and one or more memories, wherein the one or more memories store at least one computer program, and the at least one computer program is loaded and executed by the one or more processors to implement the agricultural assistance strategy matching method based on growth stage assessment.

[0008] The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following: 1. By integrating crop morphology, texture, and spectral features with standard growth templates for similarity matching, multi-level verification is performed to determine the real-time growth stage of the crop and match the ideal growth curve for the corresponding stage. This solves the problems of low accuracy of single feature recognition and ambiguous growth stage determination, thereby achieving accurate staged crop growth assessment and benchmark standardization.

[0009] 2. By integrating real-time crop images with multi-source environmental data such as temperature, humidity, light, and soil parameters, and through channel stitching, deep learning network feature extraction, and parameter regression fitting, a refined real-time crop growth curve is generated. This achieves deep integration and utilization of image information and environmental information, thereby completing the quantitative, dynamic, and comprehensive characterization of the actual growth status of crops.

[0010] 3. By extracting multiple core feature parameters of the curve, such as growth rate, curve inflection point, and cumulative growth area, and using a normalized weighted algorithm, the deviation between the actual and ideal growth curves of crops is quantified. This avoids the one-sidedness and randomness of a single evaluation indicator, thereby achieving a scientific, comprehensive, and precise quantitative evaluation of crop growth deviation.

[0011] 4. By dividing crops into refined continuous time units and combining nutrient absorption efficiency calculations with cross-comparison analysis of soil nutrient measurement data, the abnormal nitrogen, phosphorus, and potassium nutrient causes of crop growth deviations can be traced from multiple dimensions. This allows for precise differentiation of the root causes of growth problems in different nutrients and different continuous time units, thereby achieving refined and targeted diagnosis of the causes of growth deviations.

[0012] 5. By associating deviation diagnosis results, deviation degree, nutrient deviation weight and other multi-dimensional factors, the nutrient application rate and fertilizer ratio are dynamically and adaptively adjusted and water and fertilizer control instructions are generated. This breaks through the limitation of poor adaptability of traditional fixed water and fertilizer management schemes, and realizes dynamic matching, precise control and intelligent implementation of water and fertilizer auxiliary strategies under different abnormal growth scenarios.

[0013] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of this disclosure, nor is it intended to limit the scope of this disclosure. Other features of this disclosure will become readily apparent from the following description. Attached Figure Description

[0014] The accompanying drawings are provided for a better understanding of this solution and do not constitute a limitation of this application. Wherein: Figure 1 This is a flowchart of an agricultural auxiliary strategy matching method based on growth stage assessment provided by an embodiment of the present invention; Figure 2 This is a schematic diagram of an agricultural auxiliary strategy matching system based on growth stage assessment provided in an embodiment of the present invention. Detailed Implementation

[0015] The following description, in conjunction with the accompanying drawings, illustrates exemplary embodiments of this application, including various details to aid understanding. These should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this application. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0016] This invention provides a method for matching agricultural assistance strategies based on growth stage assessment. This method can be implemented using an agricultural assistance strategy matching device based on growth stage assessment, which can be a terminal or a server. Figure 1 The flowchart shown is for an agricultural auxiliary strategy matching method based on growth stage assessment. This method is applicable to the monitoring and water and fertilizer regulation of the entire growth period of field food crops and cash crops. It is executed by a processing device deployed in field edge computing equipment or a remote server, and includes the following steps: acquiring real-time crop images and environmental parameters of the target crop; determining the current growth stage of the crop based on the similarity between the real-time crop images and a preset standard growth stage morphological template, and obtaining the ideal growth curve of the crop at the growth stage; stitching the real-time crop images and environmental parameters to obtain a crop growth feature vector, and extracting crop growth parameters based on the crop growth feature vector to generate the crop corresponding to the determined growth stage. The crop growth curve is calculated to determine the deviation between the crop growth curve corresponding to the current growth stage and the ideal crop growth curve. The degree of deviation is evaluated through curve parameters. If the degree of deviation exceeds the preset deviation threshold, the crop absorption curves of each major nutrient at the current growth stage are extracted. The cause of the deviation is diagnosed based on the crop absorption curves of each major nutrient and the crop growth curve. Based on the cause and degree of deviation, a fertilization adjustment instruction is dynamically generated and sent to the integrated water and fertilizer control equipment for execution. If the degree of deviation does not exceed the preset deviation threshold, the preset standard water and fertilizer management plan is maintained, and real-time crop images and environmental parameters for the next monitoring cycle are continuously acquired to continuously monitor the deviation of the crop growth curve.

[0017] This invention establishes a closed-loop monitoring and water and fertilizer regulation system for the entire crop growth cycle, adapting to standardized planting scenarios for various cyclical crops such as field grains and cash crops, achieving intelligent agricultural auxiliary regulation without human intervention. The scenarios are set as conventional agricultural planting scenarios such as open fields and greenhouses, covering routine monitoring and dynamic water and fertilizer regulation throughout the entire crop growth cycle. The data structure is defined as comprising four core data systems: a real-time monitoring dataset (real-time crop images, multi-dimensional environmental parameters), a growth curve dataset (actual growth curve, ideal growth curve), a deviation threshold dataset, and a water and fertilizer regulation instruction dataset. Parameter values ​​are preset with a general threshold for crop growth deviation, which can be calibrated differently based on crop variety, planting area, and planting mode. First, image and environmental data are integrated to determine the crop growth stage and match the corresponding ideal growth curve. Then, growth parameters are extracted through feature splicing to construct the actual growth curve, quantifying the degree of hyperbola deviation. Finally, the branch execution regulation logic is determined based on the threshold. The output results are divided into two categories: when the deviation exceeds the threshold, a precise dynamic fertilization regulation instruction is output and the water and fertilizer equipment is driven to execute; when the deviation does not exceed the threshold, the standard water and fertilizer scheme is maintained and the next cycle of continuous monitoring begins. The core technological advantage of this solution is that it breaks away from the drawbacks of traditional agricultural water and fertilizer management being fixed and experience-based, and achieves closed-loop management with real-time perception of crop growth status, automatic diagnosis of deviations, and dynamic adaptation of strategies. This significantly improves the targeting and accuracy of water and fertilizer management, and avoids problems such as poor crop growth and resource waste caused by excessive or insufficient fertilization from the source.

[0018] Specifically, based on the similarity between real-time crop images and preset standard growth stage morphological templates, the current growth stage of the crop is determined, and the ideal growth curve of the crop at that stage is obtained. This includes: extracting crop morphological features, crop texture features, and crop spectral features from real-time crop images, and calculating the similarity with each preset standard growth stage morphological template using cosine similarity measurement, obtaining crop similarity calculation results, including crop morphological similarity, crop texture similarity, and crop spectral similarity; comparing the crop similarity calculation results with preset crop morphological similarity thresholds, crop texture similarity thresholds, and... The crop spectral similarity threshold is compared. If the similarity calculation result of any crop does not reach the corresponding similarity threshold, it is determined that the current growth stage does not match the morphological template of the standard growth stage, and the comparison with the next standard growth stage morphological template continues. If the crop similarity calculation results all reach the corresponding similarity threshold, the growth stage corresponding to the standard growth stage morphological template is determined as the current growth stage of the crop. Based on the determined growth stage, the ideal growth curve of the crop corresponding to the growth stage is retrieved from the pre-constructed crop growth model database. The ideal growth curve of the crop represents the trajectory of the standard growth index over time in the current growth stage.

[0019] This invention employs a multi-dimensional crop feature fusion and comparison mechanism to accurately identify crop growth stages in complex field environments, addressing the problem of misjudgment caused by interference from lighting, weeds, and leaf occlusion in single visual features. The scenario is set as accurate stage identification for various crops at different growth stages, adapting to complex field shooting environments and actual planting scenarios with uneven crop growth. The data structure includes a multi-feature dataset of crops (morphological, texture, and spectral features), a standard growth stage morphological template library for the entire growth stage, a dataset of specific similarity thresholds for the three types of features, and a database of ideal growth curves for each stage. Parameter values ​​are set with independent matching thresholds for morphological, texture, and spectral features, with differentiated threshold parameters for different crop varieties to ensure adaptability for different crop identification. First, three core growth features are extracted from real-time images, and their similarity is calculated by comparing them one by one with preset standard templates. A fully matching standard template is selected through multi-threshold joint verification, ultimately pinpointing the current precise growth stage of the crop and retrieving the corresponding standard ideal growth curve. The output is a precise label for the current crop growth stage and the matching standard ideal growth trajectory curve for that stage. The core technical advantage of this solution is that it uses multi-feature fusion verification instead of single-feature judgment, effectively avoiding the growth stage identification error caused by field environmental interference, greatly improving the accuracy of growth stage determination, providing a precise benchmark for subsequent growth deviation calculation and water and fertilizer strategy matching, and ensuring the effectiveness of the overall control plan.

[0020] Specifically, real-time crop images are stitched together with environmental parameters to obtain a crop growth feature vector. Crop growth parameters are then extracted from this feature vector to generate crop growth curves corresponding to the determined growth stages. This includes: stitching real-time crop images with environmental parameters through multiple channels to obtain a multi-channel feature map. Environmental parameters include air temperature, air humidity, light intensity, soil temperature, and soil moisture. Specifically, the environmental parameter values ​​are spatially interpolated to generate a two-dimensional matrix of environmental parameters with the same spatial resolution as the real-time crop images, with each environmental parameter corresponding to one matrix channel; and then stitching the real-time crop images... Each image channel and the corresponding matrix channel of each environmental parameter are concatenated along the channel dimension to obtain a multi-channel feature map. The number of channels in the multi-channel feature map is the sum of the number of channels in the real-time crop image and the number of environmental parameters. The multi-channel feature map is then input into a pre-trained crop growth feature extraction network to obtain a crop growth feature vector. The crop growth feature extraction network is a pre-trained convolutional neural network, which includes multiple convolutional layers and pooling layers for multi-level feature extraction from the multi-channel feature map. The last convolutional layer of the convolutional neural network outputs a feature map, which is then compressed in space by a global average pooling operation. The crop growth feature vector is mapped to a fixed length through fully connected layers. Based on this feature vector, crop growth parameters are predicted using a fully connected regression network. This network consists of multiple stacked fully connected layers, each followed by a nonlinear activation function. The last layer has the same number of output nodes as the number of crop growth parameters to be predicted. The fully connected regression network performs a layer-by-layer nonlinear transformation on the crop growth feature vector, regressing and mapping continuous predicted values ​​of each crop growth parameter from it. These continuous predicted values ​​include plant height, leaf area index, and aboveground growth rate. The training process of the fully connected regression network for predicted crop yield, predicted normalized vegetation index (NZD), and predicted crop cover is as follows: Crop growth feature vector samples labeled with actual crop growth parameters are used as training data. The mean squared error between the predicted and actual values ​​is used as the loss function. The weight parameters of the fully connected regression network are iteratively updated through gradient backpropagation until the loss function converges. The crop growth parameters are arranged according to a time series, and a pre-defined fitting function is used to generate crop growth curves corresponding to the determined growth stages. The fitting function is a polynomial fitting function with the form y(t) = a0 + a1t + a2t. 2 +...+a n t n Where t is the normalized time variable, and the coefficients are a0~a n The crop growth parameters arranged in time series are determined by fitting the data using the least squares method, where n represents the time number.

[0021] This invention utilizes image and environmental data fusion modeling and deep learning feature regression mechanisms to adapt to the quantitative modeling of crop growth status in complex field environments, solving the problems of low efficiency in traditional manual measurement and one-sided data from single image modeling. The scenario is set as an automated collection, quantification, and modeling of crop growth parameters in the field throughout all time periods, adapting to planting environments with dynamic changes in temperature, humidity, light, and soil conditions. The data structure is defined as including multi-channel feature map data, convolutional feature vector data, a fully connected regression network weight parameter set, and a time-series crop growth parameter dataset. The parameters are fixed, with environmental monitoring parameters including air temperature and humidity, light intensity, and soil temperature and humidity. First, spatial interpolation unifies the spatial resolution of environmental parameters and images; channel stitching generates a multi-dimensional fused feature map; then, a pre-trained convolutional network extracts fixed-length growth feature vectors; relying on a fully connected regression network, it accurately predicts five core growth parameters such as plant height and leaf area index; finally, a continuous, staged growth curve is constructed using a time-series fitting function; the output is a time-continuous, parameter-quantified actual growth curve of the crop at the current stage, fully restoring the real-time growth dynamics of the crop. The core technical effect of this solution is to integrate visual morphology data and field environment data, and use deep learning models to achieve automated and accurate prediction of growth parameters, replacing the traditional manual measurement mode. The data dimensions are more comprehensive and the accuracy is higher, which can truly reflect the actual growth status of crops in complex environments and provide real and reliable data support for deviation analysis.

[0022] Specifically, the deviation between the crop growth curve corresponding to the determined growth stage and the ideal crop growth curve is calculated, and the degree of deviation is evaluated through curve parameters. This includes: extracting curve feature parameters of the crop growth curve and the ideal crop growth curve corresponding to the determined growth stage, including the growth rate slope, curve inflection point time, maximum growth rate, and cumulative growth integral area; calculating the deviation between the curve feature parameters of the crop growth curve and the ideal crop growth curve, performing normalization processing, and then performing a weighted sum based on the weights pre-assigned to each curve feature parameter to obtain a deviation degree index. The deviation degree index is used to comprehensively reflect the deviation of the actual crop growth state from the ideal growth state.

[0023] This invention employs a multi-dimensional curve feature weighted evaluation mechanism to accurately quantify crop growth state deviations, addressing the limitations of traditional single-point data comparison evaluations that are one-sided and fail to reflect overall growth deviations. The scenario is defined as a precise evaluation of the differences between crop stage growth states and the standard ideal state, adapting to the dynamic changes in crop growth rate and the differentiation of growth nodes. The data structure includes a set of core curve feature parameters, a parameter deviation dataset, a feature weight configuration set, and a comprehensive deviation index dataset. Fixed parameter values ​​are extracted from four core curve parameters: growth rate slope, curve inflection point time, maximum growth rate, and cumulative growth integral area. Each parameter is assigned a unique weight adapted to the crop's growth pattern. The four types of feature parameters from the actual and ideal growth curves are extracted separately, and the relative deviations of each parameter are calculated and normalized. Combined with preset weights, a weighted sum is obtained to finally obtain a unique comprehensive deviation index. The output is a quantified comprehensive deviation value for crop growth, intuitively reflecting the deviation of the actual crop growth from the standard state. The core technical effect of this solution is to quantify deviations from multiple dimensions such as growth dynamics, growth nodes, and total growth, abandoning the rough evaluation method of traditional single data comparison. The evaluation results are more comprehensive, objective, and accurate, and can accurately distinguish between slight deviations, moderate deviations, and severe deviations, providing a grading basis for subsequent differentiated water and fertilizer regulation.

[0024] Specifically, the comprehensive diagnosis of discrepancies between crop absorption curves and crop growth curves for each major nutrient includes: retrieving standard crop absorption curves for each major nutrient (nitrogen, phosphorus, and potassium) based on the determined current growth stage; inverting the actual absorption sequence of each major nutrient by the crop based on the crop growth curve corresponding to the determined growth stage, combined with the relationship model between crop biomass growth and major nutrient content; specifically, extracting the aboveground biomass sequence as a function of time from the crop growth curve corresponding to the determined growth stage; and inputting the aboveground biomass sequence into a pre-constructed relationship model between crop biomass growth and major nutrient content, where the model is a biomass-nutrient content model calibrated based on crop variety characteristics. The mapping function is used to calculate the cumulative amount of major nutrients absorbed by crops per unit area based on the aboveground biomass per unit area. Through a relational model, the aboveground biomass sequence values ​​of each location are mapped one by one to the cumulative amount of major nutrients sequence values, and the cumulative amount of major nutrients sequence is subjected to time difference to obtain the net absorption of major nutrients per unit time. The net absorption of major nutrients per unit time at each moment is arranged in chronological order to form the actual absorption sequence of each major nutrient by the crop. The current growth stage of the crop is divided into multiple continuous time units according to time. Taking the jointing stage of maize as an example, this growth stage lasts for about 20 days. With 4 days as a continuous time unit, the jointing stage is divided into 5 continuous time units T1-T5, and the absorption efficiency of nitrogen, phosphorus, and potassium in each unit is calculated. Based on the actual absorption sequence and the standard crop absorption curve, the absorption efficiency of each major nutrient in different continuous time units is calculated. For each major nutrient, the actual absorption sequence is integrated within each continuous time unit to obtain the actual stage absorption total for that continuous time unit. The standard crop absorption curve is then integrated within the same continuous time unit to obtain the standard stage absorption total for that continuous time unit. The actual stage absorption total is divided by the standard stage absorption total to obtain the absorption efficiency of that major nutrient in that continuous time unit. If the absorption efficiency of any major nutrient in any continuous time unit is lower than the corresponding absorption efficiency threshold, a marker indicating insufficient absorption efficiency for that major nutrient in that continuous time unit is generated. If the absorption efficiency of any major nutrient in any continuous time unit is not lower than the corresponding absorption efficiency threshold, a marker indicating normal absorption efficiency for that major nutrient in that continuous time unit is generated. All generated markers indicating insufficient absorption efficiency are collected and combined with the current soil major nutrient data for comprehensive diagnosis to determine the cause of the deviation in the growth curve. The causes of the deviation include abnormal nitrogen content, abnormal phosphorus content, and abnormal potassium content.

[0025] This invention utilizes a nutrient absorption characteristic analysis mechanism across continuous time units to precisely trace the root causes of crop growth deviations, addressing the problem that traditional overall assessments cannot pinpoint stage-specific nutrient absorption anomalies. The scenario is set as the detection and anomaly localization of nitrogen, phosphorus, and potassium nutrient absorption status at various subdivided growth stages of crops, adapting to the significant differences in nutrient requirements across different continuous time units. The data structure includes a library of standard nutrient absorption curves for each stage, a biomass-nutrient content mapping model, a stage-specific nutrient absorption efficiency dataset, and a nutrient absorption status marker dataset. Parameter values ​​are fixed, with nitrogen, phosphorus, and potassium as the core detected nutrients. Based on the current growth stage, the corresponding standard nutrient absorption curve is retrieved, and the actual nutrient absorption is inverted from biomass time-series data. The ratio of actual nutrient absorption efficiency to standard absorption efficiency is calculated for each continuous time unit, and a threshold is used to generate markers indicating insufficient or normal absorption efficiency. The output is anomaly markers for the absorption status of each core nutrient in each continuous time unit, accurately locating the stage and type of nutrient absorption anomaly. The core technical advantage of this solution is that it aligns with the nutrient requirements of plants at different stages, conducts nutrient absorption diagnosis in detail at each growth stage, and breaks through the limitations of traditional overall tracing. It can accurately pinpoint that crop growth deviations are caused by abnormal absorption of specific nutrients at a specific stage, providing a targeted basis for subsequent precise water and fertilizer regulation.

[0026] Specifically, all generated insufficient absorption efficiency markers are collected, and combined with current soil nutrient data, a comprehensive diagnosis is performed to determine the cause of the growth curve deviation. This includes: soil nutrient data such as nitrogen, phosphorus, and potassium content; if the nutrient content of any major nutrient in any continuous time unit is lower than the minimum value of the corresponding suitable range, an insufficient nutrient marker is generated for that continuous time unit; if the nutrient content of any major nutrient in any continuous time unit is higher than the maximum value of the corresponding suitable range, an excessive nutrient marker is generated for that continuous time unit; if the nutrient content of any major nutrient in any continuous time unit is within the corresponding suitable range, a normal nutrient marker is generated for that continuous time unit; all generated insufficient markers and content... Excessive and normal content markers are cross-statistically analyzed with insufficient absorption efficiency markers. The number of first joint markers (insufficient absorption efficiency and insufficient content), second joint markers (insufficient absorption efficiency and excessive content), and third joint markers (insufficient absorption efficiency and normal content) are counted for each major nutrient in each continuous time unit. If any major nutrient has an insufficient content or excessive content marker in any continuous time unit, and the number of continuous time units with insufficient absorption efficiency markers exceeds a preset number of stages, then the abnormal content of the corresponding major nutrient is taken as the cause of deviation. The ratio of the sum of the first and second joint markers to the total number of markers is used as the deviation weight for that nutrient type. The deviation weights for each target nutrient type are then statistically generated.

[0027] This invention utilizes a cross-validation mechanism between crop absorption characteristics and soil baseline data to accurately determine the core causes of growth deviations, addressing the problem that single crop absorption data cannot distinguish between nutrient deficiency and nutrient excess. The scenario is set as a precise source tracing scenario where uneven soil nutrient baselines and water-fertilizer imbalances lead to abnormal crop growth. The data structure includes a real-time soil nutrient dataset, a set of suitable nutrient content intervals for each stage, a soil-absorption joint marker statistical data set, and a nutrient deviation weight dataset. Parameter values ​​are preset to the upper and lower limits of suitable soil nitrogen, phosphorus, and potassium content for each continuous time unit, and statistical thresholds for abnormal stages. First, three categories of markers—nutrient deficiency, excess, and normal—are generated based on measured soil nutrient data. Then, cross-statistical analysis is performed with crop nutrient absorption efficiency markers to quantify the number of markers for different abnormal combinations. Finally, the deviation weight of each nutrient type is calculated. The output is a clear core cause of growth deviation (abnormal nitrogen / phosphorus / potassium content) and the corresponding nutrient deviation weight, quantifying the impact of different nutrient abnormalities on growth deviations. The core technical effect of this solution is to link two-way data from the crop absorption end and the soil supply end, so as to achieve accurate source tracing and weight quantification of the causes of deviations. It can effectively distinguish between two types of abnormal problems: nutrient deficiency and nutrient enrichment, avoid misjudgment caused by single-dimensional diagnosis, and make the regulation strategy more in line with the actual soil conditions in the field.

[0028] Specifically, fertilization adjustment instructions are dynamically generated based on the cause and degree of deviation. This includes: querying the preset mapping relationship between the cause of deviation and the target nutrient type to determine the target nutrient type that needs adjustment. The preset mapping relationship includes: abnormal nitrogen content corresponds to nitrogen as the target nutrient type, abnormal phosphorus content corresponds to phosphorus as the target nutrient type, and abnormal potassium content corresponds to potassium as the target nutrient type; determining the adjustment direction for each target nutrient type based on the insufficient and excessive content markers generated during the comprehensive diagnosis process. If a target nutrient type has an insufficient content marker, the adjustment direction for that target nutrient type is determined to be positive increase; if a target nutrient type has an excessive content marker, the adjustment direction for that target nutrient type is determined to be negative decrease; and matching the deviation degree index with the preset... The adjustment coefficient calculation function is matched to obtain the adjustment coefficient of the current total fertilizer application. The adjustment coefficient calculation function is a piecewise linear function with the deviation degree index as the independent variable. According to the adjustment direction and the adjustment coefficient, the current planned application amount of the target nutrient type is adjusted to obtain the adjusted target nutrient application amount. Specifically, if the adjustment direction is positive (increasing application), the current planned application amount is multiplied by the adjustment coefficient to obtain the adjusted target nutrient application amount. If the adjustment direction is negative (reducing application), the current planned application amount is divided by the adjustment coefficient to obtain the adjusted target nutrient application amount. According to the deviation degree weight and deviation degree index of each target nutrient type, the adjusted fertilizer ratio is obtained. Based on the adjusted target nutrient application amount and fertilizer ratio, a fertilizer adjustment instruction is generated and sent to the water and fertilizer integrated control equipment for execution.

[0029] This invention employs a dynamic fertilization adjustment mechanism linked to deviation levels to adapt to differentiated water and fertilizer precision control scenarios, solving the problem that traditional fixed fertilization rates cannot adapt to crops with varying degrees of deviation. The scenario is set as a personalized fertilization adjustment scenario for crops with different growth deviation levels and different nutrient anomaly types. The data structure is defined as including a nutrient anomaly-control type mapping table, a fertilization adjustment direction dataset, a deviation-adjustment coefficient piecewise function set, and a nutrient application rate correction dataset. Parameter values ​​use piecewise linear adjustment coefficient functions related to deviation levels, distinguishing between two differentiated calculation logics: positive increase and negative decrease. Based on the diagnosed deviation causes, the target nutrient for adjustment is identified, and the direction of increase or decrease in application is determined in conjunction with soil nutrient status. By matching the corresponding adjustment coefficient with the deviation level, the current planned fertilization rate is quantitatively corrected. The output is the precise application rate of each target nutrient after correction, forming quantitative basic parameters for fertilization adjustment. The core technical effect of this solution is to achieve intelligent control by determining the adjustment range based on the degree of deviation and the adjustment direction based on the type of anomaly. It abandons the one-size-fits-all fixed fertilization scheme. The greater the deviation, the greater the adjustment range. It accurately matches the crop's growth and repair needs, which not only solves the problem of weak growth caused by insufficient nutrients, but also avoids the problems of excessive growth and soil compaction caused by excessive nutrients, thereby improving water and fertilizer utilization efficiency.

[0030] Specifically, based on the deviation weights and deviation indices of each target nutrient type, the adjusted fertilization ratio is obtained, including: retrieving the standard nutrient requirement ratio corresponding to the current growth stage of the crop from a pre-built crop growth model database, where the standard nutrient requirement ratio is the recommended application ratio of nitrogen, phosphorus, and potassium; inputting the deviation index into a preset ratio adjustment step size mapping function to obtain the ratio adjustment step size, which is a monotonically increasing function with the deviation index as the independent variable; using the standard nutrient requirement ratio as the baseline ratio, and combining the deviation weights of each target nutrient type in the deviation reasons, the baseline ratio is corrected. If the deviation weight of any target nutrient type is higher than that of other nutrient types, the ratio adjustment step size is added to the ratio of that target nutrient type based on the baseline ratio, and the ratios of other nutrient types are adjusted downwards accordingly, with the total downward adjustment equal to the ratio adjustment step size, to obtain the adjusted fertilization ratio.

[0031] This invention addresses the problem that fixed nutrient ratios cannot meet the needs of crop growth abnormality repair by adapting to nutrient ratio optimization scenarios at different crop growth stages through a standard nutrient ratio benchmark and a dynamic deviation correction mechanism. The scenario is defined as a precise nutrient ratio control scenario combining normal crop growth nutrient supply with abnormal growth repair. The data structure includes a database of standard nitrogen, phosphorus, and potassium ratios for each growth stage, a deviation-ratio adjustment step size mapping function, and a nutrient ratio correction dataset. The parameter values ​​are set such that the ratio adjustment step size monotonically increases with the degree of deviation, employing a ratio correction rule that superimposes the step size for advantageous nutrients and halves the step size for other nutrients. The standard nutrient requirement ratio for the current growth stage is retrieved, and the corresponding adjustment step size is obtained based on the comprehensive deviation degree. The benchmark ratio is then adjusted by combining the deviation weights of each nutrient to optimize the relative ratios of nitrogen, phosphorus, and potassium. The output is a dynamically optimal fertilization ratio adapted to the current crop growth deviation. The core technical effect of this solution is to take into account both the inherent nutrient requirements of crops and the need to repair abnormal growth in real time, and to dynamically optimize the nitrogen, phosphorus and potassium fertilizer ratio. Compared with fixed fertilizer ratio, it can specifically repair nutrient imbalances, quickly correct crop growth deviations, and ensure the normal growth and development of crops in the future, thus greatly improving the scientificity and effectiveness of precision agriculture water and fertilizer regulation.

[0032] This application provides an agricultural auxiliary strategy matching system based on growth stage assessment, such as... Figure 2 A schematic diagram of an agricultural auxiliary strategy matching system based on growth stage assessment is shown, comprising: a data acquisition module for acquiring real-time crop images and environmental parameters of the target crop; a growth stage determination module for determining the current growth stage of the crop based on the similarity between the real-time crop image and a preset standard growth stage morphology template, and obtaining the ideal growth curve of the crop at the growth stage; a growth curve generation module for stitching the real-time crop image and environmental parameters to obtain a crop growth feature vector, extracting crop growth parameters based on the crop growth feature vector, and generating the crop growth curve corresponding to the determined growth stage; and a deviation assessment module for calculating the deviation corresponding to the determined growth stage. The deviation between the crop growth curve and the ideal crop growth curve is evaluated using curve parameters to determine the degree of deviation. The diagnostic decision module extracts the crop absorption curves for each major nutrient at the current growth stage when the deviation exceeds a preset threshold. Based on the crop absorption curves and the crop growth curve, it comprehensively diagnoses the cause of the deviation and dynamically generates fertilization adjustment instructions based on the cause and degree of deviation. When the deviation does not exceed the preset threshold, it maintains the preset standard water and fertilizer management plan and continues to acquire real-time crop images and environmental parameters for the next monitoring cycle, continuously monitoring the deviation of the crop growth curve. The execution control module sends fertilization adjustment instructions to the integrated water and fertilizer control equipment for execution.

[0033] This application also provides an agricultural assistance strategy matching device based on growth stage assessment, including one or more processors and one or more memories. The one or more memories store at least one computer program, which is loaded and executed by the one or more processors to implement the agricultural assistance strategy matching method based on growth stage assessment.

[0034] Those skilled in the art will understand that all or part of the steps of the above embodiments can be implemented by hardware or by a program instructing related hardware. The program can be stored in a computer-readable storage medium, such as a read-only memory, a disk, or an optical disk.

[0035] The specific embodiments described above do not constitute a limitation on the scope of protection of this application. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for matching agricultural support strategies based on growth stage assessment, characterized in that, The method includes: The system acquires real-time crop images and environmental parameters of the target crop. Based on the similarity between the real-time crop images and a preset standard growth stage morphology template, it determines the current growth stage of the crop and obtains the ideal growth curve of the crop at the growth stage. The real-time crop image is stitched together with environmental parameters to obtain a crop growth feature vector, and crop growth parameters are extracted based on the crop growth feature vector to generate the crop growth curve corresponding to the determined growth stage. The deviation between the crop growth curve corresponding to the determined growth stage and the ideal crop growth curve is calculated, and the degree of deviation is evaluated through curve parameters. If the deviation exceeds a preset deviation threshold, the crop absorption curves of each major nutrient at the current growth stage are extracted. The cause of the deviation is diagnosed based on the crop absorption curves of each major nutrient and the crop growth curves. Based on the cause of the deviation and the deviation level, a fertilization adjustment instruction is dynamically generated and sent to the integrated water and fertilizer control equipment for execution. If the deviation does not exceed the preset deviation threshold, the preset standard water and fertilizer management plan is maintained, and real-time crop images and environmental parameters for the next monitoring cycle are continuously acquired to continuously monitor the deviation of the crop growth curve.

2. The agricultural auxiliary strategy matching method based on growth stage assessment as described in claim 1, characterized in that, The step of determining the current growth stage of the crop based on the similarity between the real-time crop image and a preset standard growth stage morphology template, and obtaining the ideal growth curve of the crop at the growth stage, includes: Based on the real-time crop images, crop morphological features, crop texture features, and crop spectral features are extracted, and similarity calculations are performed with each preset standard growth stage morphological template to obtain crop similarity calculation results. The crop similarity calculation results include crop morphological similarity, crop texture similarity, and crop spectral similarity. The crop similarity calculation results are compared with preset crop morphological similarity thresholds, crop texture similarity thresholds, and crop spectral similarity thresholds. If any crop similarity calculation result fails to reach the corresponding similarity threshold, it is determined that the current growth stage does not match the morphological template of that standard growth stage, and comparison with the next standard growth stage morphological template continues. If all crop similarity calculation results reach the corresponding similarity thresholds, the growth stage corresponding to the standard growth stage morphological template is determined as the current growth stage of the crop. Based on the determined growth stage, the ideal growth curve of the crop corresponding to that growth stage is retrieved from a pre-built crop growth model database. The ideal growth curve of the crop represents the trajectory of the standard growth index over time in the current growth stage.

3. The agricultural support strategy matching method based on growth stage assessment as described in claim 1, characterized in that, The step of stitching the real-time crop image with environmental parameters to obtain a crop growth feature vector, and extracting crop growth parameters based on the crop growth feature vector to generate the crop growth curve corresponding to the determined growth stage includes: The real-time crop images are stitched together with environmental parameters to obtain a multi-channel feature map. The environmental parameters include air temperature, air humidity, light intensity, soil temperature, and soil moisture. The multi-channel feature map is input into a pre-trained crop growth feature extraction network to obtain a crop growth feature vector; Based on the crop growth feature vector, crop growth parameters are predicted by a fully connected regression network. The crop growth parameters include predicted plant height, predicted leaf area index, predicted aboveground biomass, predicted normalized vegetation index, and predicted crop cover. The crop growth parameters are arranged in a time series, and a preset fitting function is used to generate the crop growth curve corresponding to the determined growth stage.

4. The agricultural support strategy matching method based on growth stage assessment as described in claim 1, characterized in that, The deviation between the crop growth curve corresponding to the determined growth stage and the ideal crop growth curve is calculated, and the degree of deviation is evaluated through curve parameters, including: The curve feature parameters of the crop growth curve and the ideal crop growth curve corresponding to the determined growth stage are extracted respectively. The curve feature parameters include the growth rate slope, the curve inflection point time, the maximum growth rate, and the cumulative growth integral area. The deviation between the characteristic parameters of the crop growth curve and the ideal crop growth curve is calculated, normalized, and then weighted and summed based on the weights pre-assigned to each characteristic parameter to obtain a deviation index. The deviation index is used to comprehensively reflect the deviation of the actual crop growth state from the ideal growth state.

5. The agricultural support strategy matching method based on growth stage assessment as described in claim 1, characterized in that, The reasons for the discrepancy between the crop absorption curve based on the main nutrients and the crop growth curve include: Based on the current growth stage determined by the assessment, the standard crop absorption curves for each major nutrient are retrieved, including nitrogen, phosphorus, and potassium. Based on the crop growth curves corresponding to the growth stages obtained from the determination, and combined with the pre-constructed relationship model between crop biomass growth and main nutrient content, the actual absorption sequence of each main nutrient by the crop is obtained by inversion. The current growth stage of the crop is divided into multiple continuous time units according to time. Based on the actual absorption sequence and the standard crop absorption curve, the absorption efficiency of each major nutrient in different continuous time units is calculated. If the absorption efficiency of any major nutrient in any consecutive time unit is lower than the corresponding absorption efficiency threshold, an insufficient absorption efficiency mark is generated for that major nutrient in that consecutive time unit; if the absorption efficiency of any major nutrient in any consecutive time unit is not lower than the corresponding absorption efficiency threshold, a normal absorption efficiency mark is generated for that major nutrient in that consecutive time unit. All generated insufficient absorption efficiency markers are collected and combined with the current soil nutrient data for comprehensive diagnosis to determine the causes of the growth curve deviation. The causes of the deviation include abnormal nitrogen content, abnormal phosphorus content, and abnormal potassium content.

6. The agricultural support strategy matching method based on growth stage assessment as described in claim 5, characterized in that, The process involves collecting all generated insufficient absorption efficiency markers, combining them with current soil nutrient data, and conducting a comprehensive diagnosis to determine the causes of growth curve deviations, including: The main soil nutrient data include soil nitrogen content, soil phosphorus content, and soil potassium content; If the nutrient content of any major nutrient in any consecutive time unit is lower than the minimum value of the corresponding suitable content range, a "insufficient" label for that major nutrient in that consecutive time unit is generated; if the nutrient content of any major nutrient in any consecutive time unit is higher than the maximum value of the corresponding suitable content range, a "excessive" label for that major nutrient in that consecutive time unit is generated; if the nutrient content of any major nutrient in any consecutive time unit is within the corresponding suitable content range, a "normal" label for that major nutrient in that consecutive time unit is generated. All generated insufficient content markers, excessive content markers, and normal content markers are collected and cross-statistically analyzed with the insufficient absorption efficiency markers. The number of first joint markers, the number of second joint markers, and the number of third joint markers that simultaneously exist for each major nutrient in each continuous time unit are counted. If any major nutrient has a deficiency or excess content marker in any continuous time unit, and there is a deficiency absorption efficiency marker in more than a preset number of continuous time units, then the abnormal content of the corresponding major nutrient is taken as the cause of the deviation, and the ratio of the sum of the first and second joint markers to the total number of markers is taken as the deviation weight of that nutrient type; the deviation weight of each target nutrient type is statistically generated.

7. The agricultural support strategy matching method based on growth stage assessment as described in claim 6, characterized in that, The dynamic generation of fertilization adjustment instructions based on the cause and degree of deviation includes: Based on the reasons for the deviation, query the preset mapping relationship between the reasons for the deviation and the target nutrient type to determine the target nutrient type that needs to be adjusted. The preset mapping relationship includes: abnormal nitrogen content corresponds to the target nutrient type of nitrogen, abnormal phosphorus content corresponds to the target nutrient type of phosphorus, and abnormal potassium content corresponds to the target nutrient type of potassium. Based on the insufficient and excessive content markers generated during the comprehensive diagnosis process, the adjustment direction for each target nutrient type is determined. If a target nutrient type has an insufficient content marker, the adjustment direction for that target nutrient type is determined to be positive increase application; if a target nutrient type has an excessive content marker, the adjustment direction for that target nutrient type is determined to be negative decrease application. The deviation degree index is matched with the preset adjustment coefficient calculation function to obtain the adjustment coefficient of the current total fertilizer application. The adjustment coefficient calculation function is a piecewise linear function with the deviation degree index as the independent variable. Based on the adjustment direction and the adjustment coefficient, the current planned application amount of the target nutrient type is adjusted to obtain the adjusted target nutrient application amount. Based on the deviation weight and deviation index of each target nutrient type, the adjusted fertilization ratio is obtained. Based on the adjusted target nutrient application amount and fertilization ratio, a fertilization adjustment instruction is generated and sent to the integrated water and fertilizer control equipment for execution.

8. The agricultural auxiliary strategy matching method based on growth stage assessment according to claim 7, characterized in that, The process of obtaining the adjusted fertilization ratio based on the deviation weights and deviation indexes for each target nutrient type includes: Based on the current growth stage of the crop as determined, the standard nutrient requirement ratio corresponding to that growth stage is retrieved from the pre-constructed crop growth model database. The standard nutrient requirement ratio is the recommended application ratio of nitrogen, phosphorus, and potassium. The deviation degree index is input into a preset proportional adjustment step size mapping function to obtain the proportional adjustment step size. The proportional adjustment step size mapping function is a monotonically increasing function with the deviation degree index as the independent variable. Using the standard nutrient requirement ratio as the baseline ratio, and combining the deviation weights of each target nutrient type in the reasons for deviation, the baseline ratio is corrected. If the deviation weight of any target nutrient type is higher than that of other nutrient types, the ratio adjustment step size is added to the ratio of that target nutrient type based on the baseline ratio, and the ratios of other nutrient types are adjusted downwards accordingly to obtain the adjusted fertilization ratio.

9. An agricultural assistance strategy matching system based on growth stage assessment, applied in the agricultural assistance strategy matching method based on growth stage assessment as described in any one of claims 1 to 8, characterized in that, include: The data acquisition module is used to acquire real-time crop images and environmental parameters of the target crop; The growth stage determination module is used to determine the current growth stage of the crop based on the similarity between the real-time crop image and the preset standard growth stage morphology template, and to obtain the ideal growth curve of the crop at the growth stage. The growth curve generation module is used to stitch the real-time crop image with environmental parameters to obtain a crop growth feature vector, and extract crop growth parameters based on the crop growth feature vector to generate the crop growth curve corresponding to the determined growth stage. The deviation assessment module is used to calculate the deviation between the crop growth curve corresponding to the determined growth stage and the ideal crop growth curve, and to assess the degree of deviation through curve parameters. The diagnostic decision module extracts the crop absorption curves of each major nutrient at the current growth stage of the crop when the deviation exceeds a preset deviation threshold. Based on the crop absorption curves of each major nutrient and the crop growth curve, it comprehensively diagnoses the cause of the deviation and dynamically generates fertilization adjustment instructions based on the cause of the deviation and the degree of deviation. When the deviation does not exceed the preset deviation threshold, it maintains the preset standard water and fertilizer management plan and continues to acquire real-time crop images and environmental parameters for the next monitoring cycle to continuously monitor the deviation of the crop growth curve. The execution control module is used to send fertilization adjustment commands to the integrated water and fertilizer control equipment for execution.

10. An agricultural auxiliary strategy matching device based on growth stage assessment, characterized in that, The device includes one or more processors and one or more memories, wherein at least one computer program is stored in the one or more memories, and the at least one computer program is loaded and executed by the one or more processors to implement the agricultural assistance strategy matching method based on growth stage assessment as described in any one of claims 1 to 8.

Images