A method and system for dynamically collecting maize field phenotype group data

Abstract

The application provides a corn field phenotype group data dynamic acquisition method and system, and relates to the corn planting data acquisition technical field.The method comprises the following steps: identifying macroscopic phenotype changes indicated in macroscopic phenotype data;querying a preset corn phenotype characteristic map according to the identified macroscopic phenotype changes;determining specific data acquisition parameters or parameter combinations for obtaining diagnostic microscopic characteristics according to the query result;adjusting sensor configuration and acquisition platform behavior of an automatic data acquisition device by driving the automatic data acquisition device according to the specific data acquisition parameters or parameter combinations to obtain diagnostic microscopic characteristic data;analyzing the diagnostic microscopic characteristic data, and classifying the macroscopic phenotype changes according to a preset diagnostic characteristic mode in the corn phenotype characteristic map.The method of the application overcomes the inefficiency of the existing general fine acquisition mode, and significantly improves the effectiveness and diagnostic accuracy of field phenotype group data acquisition.

Description

Technical Field

[0001] This invention relates to the field of maize planting data acquisition technology, and more specifically, to a method and system for dynamic acquisition of maize field phenotypic data. Background Technology

[0002] Automated data acquisition equipment in corn fields, typically equipped with high-resolution image sensors such as unmanned aerial vehicles (UAVs) or ground inspection robots, periodically collects large-scale phenotypic data from corn plants according to preset paths and schedules. These devices utilize sensors such as visible light cameras and multispectral cameras to acquire macroscopic phenotypic data such as plant height, leaf area, canopy coverage, and leaf color. The collected data is transmitted to a backend processing system for image analysis and pattern recognition to monitor the growth, development, and health status of the corn. The standard data acquisition strategy is preset, such as the equipment taking images at fixed spatial or time intervals to ensure data uniformity and comparability.

[0003] However, in actual field operations, while the system can identify macroscopic changes in the phenotypic traits of maize plants through preliminary analysis of continuously acquired images—such as abnormal leaf color, leaf curling, or changes in canopy morphology in a certain area—this preliminary identification remains at the superficial level and fails to distinguish the specific type and underlying cause. For example, slight yellowing of maize leaves could be an early symptom of nitrogen deficiency, water stress, or disease; the system can only identify "changes in leaf color" but cannot determine the specific type. This ambiguity in detecting macroscopic changes is a problem currently faced by the technology.

[0004] When a system detects such ambiguous phenotypic changes, existing adaptive strategies typically employ a pre-defined, generalized, refined acquisition mode. For example, regardless of the cause of yellowing, the system might instruct the device to hover over the area and capture high-resolution images from multiple angles, or switch to a multispectral camera for supplementary acquisition. This generalized, refined acquisition mode lacks efficiency and specificity. For nutrient deficiency problems requiring specific spectral information for diagnosis, visible light multi-angle images have limited value; for disease problems requiring observation of microscopic features of lesions, multispectral images lack sufficient spatial resolution. This leads to unnecessary resource consumption and fails to efficiently acquire the crucial information needed to diagnose specific problems.

[0005] The underlying microscopic features or specific geometric morphologies behind different types of phenotypic changes differ, and these differences are difficult to capture in conventional macroscopic images. For example, yellowing caused by nitrogen deficiency starts from the leaf tip and spreads along the veins, presenting as a uniform pale green; yellowing caused by fungal diseases may appear as irregular spots. Slight leaf curling may be due to water stress or insect infestation. These subtle features are key to accurately distinguishing different phenotypic changes. Under conventional image acquisition modes with fixed perspectives and parameters, these key microscopic features or specific geometric morphological information are often ignored or cannot be effectively captured. For example, capturing the texture of subtle leaf lesions requires macro photography; analyzing the three-dimensional morphology of leaf curling requires multi-view stereo imaging or lidar data; distinguishing specific nutrient deficiencies requires specific band spectral data. The lack of key discriminative features hinders the system from accurately classifying initially identified, ambiguous changes.

[0006] Therefore, after initially identifying macroscopic phenotypic changes in maize plants, automated phenotypic data acquisition equipment faces the challenge of overcoming the inefficiency of existing generalized, refined acquisition methods, avoiding unnecessary resource consumption, and ensuring efficient acquisition of key information needed to diagnose specific problems, thereby achieving accurate classification and diagnosis of phenotypic traits. This requires the system to adjust its acquisition behavior after initially detecting macroscopic changes to obtain specific microscopic features or geometric morphological data that reveal the nature of the changes. Summary of the Invention

[0007] The purpose of this invention is to provide a method and system for dynamic acquisition of maize field phenotypic data, which overcomes the inefficiency of existing general refined acquisition modes and significantly improves the effectiveness and diagnostic accuracy of field phenotypic data acquisition.

[0008] In a first aspect, the present invention provides a method for dynamically collecting maize field phenotypic data, comprising the following steps:

[0009] S1. Obtain macroscopic phenotypic data of maize plants and identify macroscopic phenotypic changes indicated in the macroscopic phenotypic data;

[0010] S2. Based on the identified macroscopic phenotypic changes, query the preset maize phenotypic feature map; the maize phenotypic feature map includes diagnostic microscopic features associated with macroscopic phenotypic changes, as well as specific data acquisition parameters required to obtain diagnostic microscopic features;

[0011] S3. Based on the query results, determine the specific data acquisition parameters or parameter combinations used to obtain diagnostic microscopic features; the parameter combination includes multiple specific data acquisition parameters;

[0012] S4. Based on specific data acquisition parameters or combinations of parameters, diagnostic microscopic feature data is obtained by driving automated data acquisition equipment to adjust its sensor configuration and acquisition platform behavior.

[0013] S5. Analyze the diagnostic microscopic feature data, and classify the macroscopic phenotypic changes according to the analysis results and the pre-set diagnostic feature patterns in the maize phenotypic feature map.

[0014] The dynamic data acquisition method for maize field phenotypic sets provided by this invention overcomes the inefficiency of existing general refined acquisition modes by establishing an intelligent mechanism that can determine and execute targeted data acquisition behaviors based on the type of macroscopic changes. This effectively acquires key microscopic features or specific geometric morphological information corresponding to different types of phenotypic changes, thereby supporting accurate classification of phenotypic traits, avoiding redundant data acquisition, and improving the effectiveness and diagnostic accuracy of data acquisition in resource-constrained field operation environments.

[0015] Secondly, the present invention provides a dynamic acquisition system for maize field phenotypic data, comprising:

[0016] The acquisition module is used to acquire macroscopic phenotypic data of maize plants and identify macroscopic phenotypic changes indicated in the macroscopic phenotypic data.

[0017] The query module is used to query a preset maize phenotypic feature map based on the identified macroscopic phenotypic changes. The maize phenotypic feature map includes diagnostic microscopic features associated with macroscopic phenotypic changes, as well as specific data acquisition parameters required to obtain the diagnostic microscopic features.

[0018] The determination module is used to determine, based on the query results, specific data acquisition parameters or combinations of parameters for obtaining diagnostic microscopic features; the parameter combination includes multiple specific data acquisition parameters.

[0019] The control acquisition module is used to obtain diagnostic microscopic feature data by driving automated data acquisition equipment to adjust its sensor configuration and acquisition platform behavior according to specific data acquisition parameters or parameter combinations.

[0020] The analysis module is used to analyze diagnostic microscopic feature data and classify macroscopic phenotypic changes based on the analysis results and the preset diagnostic feature patterns in the maize phenotypic feature map.

[0021] As can be seen from the above, the dynamic acquisition method for maize field phenotypic data provided by this invention, by constructing and applying maize phenotypic feature maps, enables automated acquisition equipment to intelligently determine and execute targeted data acquisition actions after initial identification of macroscopic phenotypic changes. This effectively overcomes the inefficiency of existing general-purpose refined acquisition modes, avoiding redundant data acquisition and unnecessary resource waste. By efficiently acquiring key microscopic features or specific geometric morphological information corresponding to different types of phenotypic changes, and supporting precise type classification of maize phenotypic traits, it significantly improves the effectiveness and diagnostic accuracy of field phenotypic data acquisition, providing more reliable data support for precision agriculture management.

[0022] Other features and advantages of the invention will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing embodiments of the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description and the accompanying drawings. Attached Figure Description

[0023] Figure 1 This is a flowchart of a method for dynamically collecting maize field phenotypic data according to an embodiment of the present invention.

[0024] Figure 2 This is a schematic diagram of a dynamic acquisition system for maize field phenotypic group data provided in an embodiment of the present invention.

[0025] Label Explanation:

[0026] 100. Acquisition Module; 200. Query Module; 300. Confirmation Module; 400. Control Acquisition Module; 500. Analysis Module. Detailed Implementation

[0027] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0028] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this invention, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0029] Reference Appendix Figure 1 This invention provides a method for dynamically collecting maize field phenotypic data, comprising the following steps:

[0030] S1. Obtain macroscopic phenotypic data of maize plants and identify macroscopic phenotypic changes indicated in the macroscopic phenotypic data;

[0031] S2. Based on the identified macroscopic phenotypic changes, query the preset maize phenotypic feature map; the maize phenotypic feature map includes diagnostic microscopic features associated with macroscopic phenotypic changes, as well as specific data acquisition parameters required to obtain diagnostic microscopic features;

[0032] S3. Based on the query results, determine the specific data acquisition parameters or parameter combinations used to obtain diagnostic microscopic features; the parameter combination includes multiple specific data acquisition parameters;

[0033] S4. Based on specific data acquisition parameters or combinations of parameters, diagnostic microscopic feature data is obtained by driving automated data acquisition equipment to adjust its sensor configuration and acquisition platform behavior.

[0034] S5. Analyze the diagnostic microscopic feature data, and classify the macroscopic phenotypic changes according to the analysis results and the pre-set diagnostic feature patterns in the maize phenotypic feature map.

[0035] A maize phenotypic feature map refers to a pre-built knowledge base or database, which can be implemented using structured databases, knowledge graphs, or machine learning models, such as relational databases or graph databases. Its main purpose is to store and associate the mapping relationships between macroscopic phenotypic changes in maize and diagnostic microscopic features and their acquisition parameters, guiding subsequent refined data acquisition. Specific data acquisition parameters or parameter combinations refer to sensor configurations and acquisition platform behavior parameters specifically set to obtain specific diagnostic microscopic features. These can be implemented using parameters such as focal length, aperture, exposure time, spectral band selection, sensor type switching, acquisition altitude, flight speed, hovering position, and shooting angle, or combinations thereof. For example, adjusting the camera focal length for macro photography or switching to a specific spectral camera to obtain nutrient information. Their main purpose is to ensure that automated data acquisition equipment can specifically capture the key microscopic information needed behind macroscopic changes. Automated data acquisition equipment refers to devices capable of autonomously or semi-autonomously performing data acquisition tasks in the field. These can take the form of unmanned aerial vehicles (UAVs), ground inspection robots, or fixed sensor arrays. Examples include UAVs equipped with multispectral cameras and ground robots equipped with lidar. Their primary purpose is to efficiently and flexibly acquire phenotypic data of corn plants in the field environment. Sensor configuration and acquisition platform behavior refer to the hardware settings used for data acquisition on automated data acquisition equipment and their spatial movement. This can be achieved by adjusting sensor type, resolution, focal length, aperture, exposure parameters, spectral band, and changing the acquisition platform's altitude, speed, attitude, path, hovering position, and shooting angle. For example, a UAV might lower its flight altitude and adjust the camera focal length to capture leaf details. This is primarily to optimize the data acquisition process to obtain high-quality diagnostic microscopic feature data according to specific data acquisition parameter requirements. Diagnostic microscopic feature data refers to data that reveals the essential causes of macroscopic phenotypic changes at the microscopic level or in specific spectral bands. This data can be obtained using high-resolution images, multispectral images, hyperspectral images, 3D point cloud data, or thermal infrared images. Examples include texture images of leaf lesions and chlorophyll absorption spectrum data in specific bands. Its primary purpose is to provide specific pathological, physiological, or environmental stress information behind macroscopic phenotypic changes, thereby supporting accurate type classification. Diagnostic feature patterns refer to pre-defined, identifiable patterns or feature sets in diagnostic microscopic feature data that are directly associated with specific types of macroscopic phenotypic changes. These can be obtained using image texture features, spectral curve features, geometric morphological features, or combinations thereof. Examples include the shape, color, and edge features of lesions in specific diseases. This data primarily serves as a basis for classification, matching the collected microscopic feature data with known pathophysiological patterns to accurately classify macroscopic phenotypic changes.

[0036] The working principle of this invention is as follows: By establishing a predefined "maize phenotypic feature atlas," macroscopic phenotypic changes in maize plants (such as leaf yellowing) are correlated with microscopic features that can accurately diagnose their underlying causes (such as specific lesion textures, leaf vein discoloration patterns, and specific spectral absorption peaks). This is further mapped to specific data acquisition parameters (such as sensor type, focal length, shooting angle, and platform posture) required to acquire these microscopic features. When automated data acquisition equipment conducts routine field inspections, if macroscopic phenotypic changes in maize plants are initially identified, the system no longer performs blind, general, and refined data acquisition. Instead, it intelligently queries the atlas. Based on the type of macroscopic change, the atlas provides a series of possible diagnostic microscopic features and their corresponding precise acquisition instructions. The system then generates targeted, refined acquisition instructions to guide the acquisition equipment in adjusting its sensor configuration and platform behavior (such as lowering the height, switching sensors, and changing the shooting angle) to purposefully capture the key microscopic feature data hidden behind the macroscopic changes. Subsequently, the system analyzes these finely collected microscopic feature data and compares them with preset diagnostic feature patterns in the atlas, thereby achieving accurate classification and diagnosis of maize phenotypic changes, such as accurately distinguishing between nitrogen deficiency, water stress, or specific diseases. The entire process forms a closed loop, ensuring the relevance and effectiveness of the collected data and avoiding resource waste.

[0037] Based on the synergistic effect of the above-mentioned technical features, the core innovation of this application lies in combining the preliminary identification of macroscopic phenotypic changes with a preset maize phenotypic feature map, and dynamically adjusting the sensor configuration and acquisition platform behavior of the automated data acquisition equipment based on this map, thereby achieving on-demand and efficient acquisition of diagnostic microscopic feature data, and overcoming the ambiguity of macroscopic phenotypic change detection, the inefficiency of general fine acquisition mode, and the lack of key discriminative feature acquisition.

[0038] Specifically, this application provides a dynamic and intelligent maize field phenotypic data acquisition strategy. First, by acquiring macroscopic phenotypic data of maize plants and identifying indicated macroscopic phenotypic changes, such as abnormal leaf color or morphological changes, preliminary judgment criteria and triggering conditions are provided for subsequent dynamic data acquisition. This step is the starting point of the entire dynamic acquisition process, avoiding indiscriminate and blind acquisition and laying the foundation for targeted data acquisition. Next, based on the identified macroscopic phenotypic changes, the system can intelligently query a pre-set maize phenotypic feature map. This map pre-stores diagnostic microscopic features associated with specific macroscopic phenotypic changes, as well as specific data acquisition parameters required to acquire these microscopic features. This map-based query mechanism allows the system to specifically understand the microscopic features that may correspond to a certain macroscopic change, and how to obtain this key information by adjusting the acquisition parameters. This effectively solves the inefficiency problem of existing general refined acquisition modes, avoids unnecessary resource consumption, and provides guidance for efficiently acquiring the key information needed to diagnose specific problems. Subsequently, based on the query results, the system can accurately determine the specific data acquisition parameters or parameter combinations used to acquire diagnostic microscopic features. This determination process, based on a preliminary understanding of macroscopic phenotypic changes and pre-defined knowledge of diagnostic microscopic features, ensures the targeted nature and effectiveness of subsequent data acquisition. By determining specific parameters or combinations of parameters, the automated data acquisition equipment can effectively capture key discriminant features that are difficult to identify at the macroscopic level, thus overcoming the problem of missing key discriminant feature acquisition. Then, according to the determined specific data acquisition parameters or combinations of parameters, the system drives the automated data acquisition equipment to adjust its sensor configuration and acquisition platform behavior to acquire diagnostic microscopic feature data. This adaptive adjustment capability allows the equipment to break away from preset fixed acquisition modes and instead efficiently and accurately acquire diagnostic microscopic feature data based on actual macroscopic phenotypic change requirements, such as the texture of leaf lesions, the three-dimensional morphology of leaf curling, or spectral information of specific nutrient deficiencies. This directly solves the inefficiency problem of general refined acquisition modes and the problem of missing key discriminant feature acquisition, ensuring the diagnostic value of the acquired data. Finally, the acquired diagnostic microscopic feature data is analyzed, and based on the analysis results and the pre-defined diagnostic feature patterns in the maize phenotypic feature map, the macroscopic phenotypic changes are classified. For example, "leaf yellowing" can be precisely classified as "nitrogen deficiency" or "a certain disease." This classification ability based on microscopic features overcomes the ambiguity of macroscopic change detection, enabling the system to make accurate judgments on the health status and growth and development of corn plants, providing a reliable basis for subsequent precision agricultural management.

[0039] As a preferred embodiment, the solution of this application is implemented as follows: In a cornfield, an unmanned aerial vehicle (UAV) equipped with a visible light camera and a multispectral camera patrols along a preset path, periodically acquiring visible light images of corn plants. A background image processing system analyzes these images and identifies macroscopic yellowing of corn leaves in a certain area. Based on the identified macroscopic phenotypic change of "leaf yellowing," the system queries a preset corn phenotypic feature map. This map stores diagnostic microscopic features that may correspond to "leaf yellowing," such as "uniform yellowing starting at the leaf tip" or "irregular spot-like yellowing," as well as specific data acquisition parameters required to acquire these microscopic features. For example, for nitrogen deficiency, "near-infrared spectral data" may be needed, and for fungal diseases, "high-resolution macro images" may be needed. Based on the query results, the system determines that "near-infrared spectral data" and "high-resolution macro images" are needed as diagnostic microscopic features. To this end, the system determined a specific combination of data acquisition parameters, including adjusting the UAV to a lower flight altitude, switching to a multispectral camera and selecting the near-infrared band, and simultaneously adjusting the focal length of the visible light camera to achieve macro photography. Subsequently, the system drove the UAV to adjust its flight altitude and attitude, switched the multispectral camera to the near-infrared band for data acquisition, and simultaneously adjusted the focal length and exposure parameters of the visible light camera to acquire high-resolution macro images of the leaves in the yellowed area. Through these adjustments, the UAV acquired near-infrared spectral data and high-resolution lesion images of the maize leaves in this area. Finally, the background system analyzed the acquired near-infrared spectral data, extracted chlorophyll content-related indicators, and performed texture and morphological analysis on the high-resolution images to identify lesion characteristics. The system compared these analysis results with the preset "nitrogen deficiency spectral pattern" and "fungal disease lesion pattern" in the maize phenotypic feature map. If the spectral pattern matched nitrogen deficiency, the macroscopic yellowing was classified as "nitrogen deficiency"; if the lesion pattern matched fungal disease, it was classified as "fungal disease." In this way, accurate classification of yellowing phenomena in corn leaves was achieved.

[0040] Through the above-described solution, this application effectively addresses the ambiguity problem in the detection of macroscopic phenotypic changes in existing technologies, achieving accurate classification of maize phenotypic traits. Simultaneously, by dynamically adjusting the sensor configuration of the data acquisition equipment and the behavior of the acquisition platform, the inefficiency of generalized refined acquisition modes is avoided, significantly improving the targeting and efficiency of data acquisition and reducing unnecessary resource consumption. Furthermore, this application ensures the efficient acquisition of key microscopic feature information required for diagnosing specific problems, overcoming the challenge of missing key discriminative features, thereby providing reliable data support for precision agriculture management.

[0041] In some embodiments, the specific steps in step S2 include:

[0042] S21. Obtain auxiliary information related to the macroscopic phenotypic changes identified; auxiliary information includes the growth stage of maize plants, environmental data of the region, or soil test data;

[0043] S22. Based on the identified macrophenotypic changes and auxiliary information, make a preliminary judgment on the macrophenotypic changes to obtain a judgment result indicating the potential type of macrophenotypic changes;

[0044] S23. Based on the judgment result, select diagnostic micro-features that match the judgment result from the preset maize phenotypic feature map, and obtain the specific data acquisition parameters required for the diagnostic micro-features.

[0045] Ancillary information refers to data that provides additional contextual clues to the nature of macrophenotypic changes beyond the macrophenotypic data itself. It can take various forms, such as real-time environmental data directly acquired from sensors, historical meteorological data retrieved from databases, or soil nutrient data obtained through soil sampling analysis. This information is crucial for understanding the underlying causes of macrophenotypic changes. Preliminary judgment refers to a preliminary, non-final assessment of the potential type of macrophenotypic changes based on the identified changes and the acquired ancillary information. It can be implemented using various techniques, such as logical reasoning through a pre-defined expert rule system, classification prediction using machine learning models (e.g., decision trees, support vector machines, or neural networks), or probabilistic inference using fuzzy logic or Bayesian networks. The judgment result refers to the conclusion output by the preliminary judgment process, indicating the potential type of macrophenotypic change. It can be presented in various forms, such as one or more labels indicating the highest probability of the potential type (e.g., "suspected nitrogen deficiency," "suspected water stress"), a probability distribution, or a confidence score. Selecting diagnostic micro-features consistent with the judgment results refers to the purposeful selection of micro-features directly related to the potential type obtained from the preliminary judgment, from a pre-defined maize phenotypic feature map, that can be used for further precise diagnosis. This can be achieved through database queries, index matching, or semantic association-based retrieval.

[0046] This application's solution refines and optimizes the query process for macrophenotypic changes by introducing auxiliary information and preliminary judgment. It aims to improve the accuracy and efficiency of diagnostic micro-feature screening, effectively focusing on the most likely diagnostic path and providing more targeted parameters for subsequent refined data collection. Specifically, after identifying a macrophenotypic change, the system acquires auxiliary information associated with that change. This auxiliary information, such as the growth stage of the maize plant, environmental data of the region, or soil testing data, provides crucial context for understanding the background of the macrophenotypic change. This information forms the basis for subsequent judgments, as the same macrophenotypic change may indicate different potential problems under different conditions. Next, the system makes a preliminary judgment on the macrophenotypic change based on the identified macrophenotypic change and the acquired auxiliary information. This judgment process is no longer a simple direct mapping but comprehensively considers the macrophenotypic change itself as well as the environmental and physiological background in which it occurs. Through this comprehensive analysis, the system can refine vague macroscopic phenotypic changes (such as "leaf yellowing") into more specific potential types (such as "suspected nitrogen deficiency" or "suspected early stage of disease"), thereby obtaining a judgment result indicating the potential type of macroscopic phenotypic changes. This preliminary judgment significantly narrows the scope of subsequent feature screening and improves the initial accuracy of diagnosis. Finally, based on this preliminary judgment, the system filters diagnostic microscopic features that match the judgment result from a pre-set maize phenotypic feature map and obtains the specific data acquisition parameters required for these microscopic features. This screening mechanism replaces the method of directly querying the entire map, enabling the system to more accurately locate diagnostic microscopic features related to specific potential types. This means that the system no longer needs to traverse the entire feature map, but purposefully selects the microscopic features most likely to reveal the essence of the problem and obtains their corresponding specific data acquisition parameters. By introducing auxiliary information and preliminary judgment, this solution refines the original direct query step into a targeted screening process. This improvement enables the system to overcome the challenges posed by the ambiguity of macroscopic phenotypic changes and avoids the problem of overly broad query results. By predicting potential types, the system can effectively focus on the most likely diagnostic path, thus ensuring the efficiency and relevance of subsequent refined data collection. This purposeful screening avoids unnecessary general data collection, ensuring the efficiency and effectiveness of subsequent data collection, and enabling the collected data to be used more directly for accurate diagnosis, thereby significantly improving the intelligence level and resource utilization efficiency of the entire data collection method.

[0047] By introducing auxiliary information and making preliminary judgments on macroscopic phenotypic changes, this method can effectively overcome the ambiguity in identifying macroscopic phenotypic changes. This allows the system to perform targeted filtering based on more specific potential type judgments when querying preset maize phenotypic feature maps, rather than performing broad direct matching. This filtering mechanism significantly narrows the query scope, enabling the system to more accurately focus on the most likely diagnostic path, thereby greatly improving the efficiency and targeting of subsequent refined data collection, avoiding unnecessary resource consumption, and ensuring that the acquired data can be used more directly for accurate diagnosis.

[0048] In some embodiments, the specific steps in step S22 include:

[0049] A1. Determine the time for acquiring auxiliary information;

[0050] A2. Based on the time interval between the acquisition time of auxiliary information and the recognition time of macroscopic phenotypic changes, determine whether the auxiliary information needs to be updated according to the preset auxiliary information validity rules;

[0051] A3. If the auxiliary information needs to be updated, obtain the latest data of the auxiliary information and use the latest data as the auxiliary information for preliminary judgment;

[0052] A4. If the auxiliary information does not need to be updated, the already acquired auxiliary information shall be used as the auxiliary information for preliminary judgment;

[0053] A5. Based on the identified macroscopic phenotypic changes and the auxiliary information used for preliminary judgment, a preliminary judgment is made on the macroscopic phenotypic changes to obtain a judgment result indicating the potential type of macroscopic phenotypic changes.

[0054] The validity rules for auxiliary information refer to the set of standards used to assess whether auxiliary information is still applicable to the current judgment. These rules can take various forms. For example, setting a time threshold to stipulate that auxiliary information is considered invalid if it exceeds a certain period after collection; or establishing a dynamic model based on historical data analysis to predict the validity period of auxiliary information according to the rate of environmental change; or setting different validity standards based on the type of auxiliary information (e.g., soil nutrient data may be updated less frequently, while temperature data may require real-time updates). Obtaining the latest auxiliary information refers to acquiring the auxiliary information corresponding to the current or most recent time point through methods such as re-collection, querying from a real-time database, or synchronizing from external data sources. This can be achieved through methods such as real-time sensor acquisition, subscription from an IoT platform, or obtaining the latest reports from weather stations or soil monitoring stations.

[0055] The preliminary judgment method in this application optimizes the preliminary judgment process by introducing a mechanism for judging and updating the timeliness of auxiliary information. The method first determines the specific acquisition time of the acquired auxiliary information. This timestamp provides a foundation for subsequent evaluation of the timeliness of the auxiliary information. Next, the system compares the acquisition time of the auxiliary information with the identification time of the macroscopic phenotypic change, and intelligently judges whether the current auxiliary information is still valid or needs to be updated based on preset auxiliary information validity rules. This judgment mechanism is the core of this solution; it enables the system to identify potential lags in auxiliary information, thereby avoiding the use of outdated data for judgment. If the judgment result indicates that the auxiliary information needs updating, the system will proactively trigger the process of acquiring the latest auxiliary data, ensuring that the information used for preliminary judgment best reflects the current situation. Conversely, if the auxiliary information is judged to still be valid, the system will continue to use the existing auxiliary information, thereby avoiding unnecessary duplicate collection and improving data processing efficiency. Ultimately, both the updated latest auxiliary information and the original auxiliary information verified to still be valid will be used together with the identified macroscopic phenotypic change for preliminary judgment, thereby obtaining a judgment result indicating the potential type of macroscopic phenotypic change. Through this series of steps, this approach ensures that the auxiliary information relied upon for initial assessment remains timely and accurate. This allows the initial assessment results to more accurately indicate the potential types of macroscopic phenotypic changes, thus providing a solid foundation for subsequently screening more precise diagnostic microscopic features from the maize phenotypic feature map. This dynamic management of the timeliness of auxiliary information significantly improves the targeting and effectiveness of the entire dynamic acquisition method, enabling the system to promptly acquire and utilize the latest auxiliary information for assessment, thereby avoiding missing the optimal intervention opportunity or causing subsequent acquisition strategies to fail.

[0056] This solution addresses the problem of insufficient timeliness in preliminary assessments caused by the continuous changes in maize plant growth and environmental conditions across vast maize fields. By introducing a mechanism for judging and updating the timeliness of auxiliary information, the system ensures that the auxiliary information used for preliminary assessments is always up-to-date and valid. This avoids using outdated or invalid information, thereby improving the accuracy and reliability of preliminary assessments. It also helps prevent missing the optimal intervention window for observing macroscopic phenotypic changes in maize plants and ensures the effectiveness of subsequent data collection strategies.

[0057] In some embodiments, the specific steps in step S22 include:

[0058] B1. Obtain the spatial location and identification time of the macroscopic phenotypic data, as well as the spatial location and acquisition time of the auxiliary information;

[0059] B2. Based on the spatial location and identification time of the macroscopic phenotypic data, as well as the spatial location and acquisition time of the auxiliary information, determine the spatial and temporal correlations between the macroscopic phenotypic data and each piece of auxiliary information;

[0060] B3. Based on spatial and temporal correlations, by performing spatial interpolation or temporal synchronization processing on each auxiliary information, a set of auxiliary information aligned with the macroscopic phenotypic data in space and time is generated;

[0061] B4. Determine the weight of each piece of auxiliary information in the preliminary judgment based on the data source type or data quality of each piece of auxiliary information;

[0062] B5. The identified macroscopic phenotypic changes are fused with the aligned set of auxiliary information, and the fused information is comprehensively evaluated according to the weight of each auxiliary information in the preliminary judgment to obtain the judgment result indicating the potential type of macroscopic phenotypic changes.

[0063] Spatial correlation refers to the relative positions of different data points in geographic space, which can be achieved using geographic coordinate matching, regional overlap analysis, or distance threshold judgment. Temporal correlation refers to the sequential or synchronous relationship of different data points on a time axis, which can be achieved using timestamp matching, time window analysis, or event sequence correspondence. Spatial interpolation is a method of inferring the value of data points at unknown spatial locations based on data points at known spatial locations, which can be achieved using algorithms such as Kriging interpolation, inverse distance weighted interpolation, or spline interpolation. Time synchronization processing refers to the method of adjusting data collected at different time points to the same time reference, which can be achieved using linear interpolation, nearest neighbor interpolation, or time series alignment algorithms. Auxiliary information set refers to a multi-source heterogeneous data set that maintains spatiotemporal consistency with macroscopic phenotypic data after spatiotemporal alignment processing, which may include environmental sensor data, soil testing data, or historical phenotypic data. Data source type refers to the classification of the source or nature of data, which can be distinguished according to sensor type, acquisition method, or data provider. Data quality refers to the degree to which data meets its intended use, which can be evaluated based on dimensions such as data completeness, accuracy, consistency, or timeliness. Weighting refers to the degree of importance assigned to different data items or information sources during data fusion or evaluation. It can be dynamically adjusted based on data source type, data quality, or expert experience. Fusion refers to the process of integrating information from different sources or of different types into a unified representation, which can be achieved using methods such as weighted averaging, decision tree fusion, or neural network fusion. Comprehensive evaluation refers to the process of comprehensively considering and judging a target based on multi-dimensional information, which can be achieved using multi-criteria decision analysis, fuzzy comprehensive evaluation, or machine learning classification models.

[0064] To overcome the potential spatial and temporal mismatches in multi-source heterogeneous data, which could affect the accuracy and reliability of initial judgments, this application proposes a data preprocessing and fusion mechanism. This mechanism first acquires the spatial location and identification time of the identified macroscopic phenotypic data, as well as the spatial location and acquisition time of each auxiliary information, laying the foundation for subsequent data alignment and correlation analysis. This metadata forms the basis for spatiotemporal matching, enabling the system to clearly define the time and location of macroscopic phenotypic changes, as well as the time and location represented by each auxiliary information. Based on this, the system determines the spatial and temporal correlations between the macroscopic phenotypic data and each auxiliary information based on this spatiotemporal information. This step establishes the spatiotemporal correspondence between different data sources, identifying their temporal and spatial differences, and providing a basis for subsequent alignment processing. Furthermore, based on these determined spatial and temporal correlations, the system generates a set of auxiliary information aligned spatially and temporally with the macroscopic phenotypic data by performing spatial interpolation or temporal synchronization processing on each auxiliary information. This step is a crucial step in resolving the spatiotemporal inconsistency problem of auxiliary information. Spatial interpolation converts auxiliary information from different spatial locations into data from the same spatial location as the macrophenotypic data. Temporal synchronization ensures that auxiliary information collected at different time points aligns with the macrophenotypic data identification time, guaranteeing that all auxiliary information used for initial judgment is consistent with macrophenotypic changes in both time and space, avoiding judgment biases caused by temporal and spatial misalignment. Furthermore, considering the potential differences in the reliability and importance of different auxiliary information, the system determines the weight of each auxiliary information in the initial judgment based on its data source type or data quality. By evaluating the data source type or data quality, each piece of auxiliary information is assigned a corresponding weight, allowing the system to prioritize high-quality, reliable auxiliary information in subsequent judgments, further improving accuracy. Finally, the system fuses the identified macrophenotypic changes with the spatiotemporally aligned and weighted auxiliary information set, and comprehensively evaluates the fused information based on the weight of each auxiliary information in the initial judgment, obtaining a judgment result indicating the potential type of macrophenotypic changes. This fusion and weighting process makes the initial judgment robust and precise, accurately indicating the potential type of macrophenotypic changes. Through the synergistic effect of the above steps, this scheme enables the effective integration of multi-source heterogeneous data during the initial judgment, overcoming the problem of data mismatch in space and time, thereby improving the accuracy and reliability of the judgment results and providing a foundation for subsequent diagnostic micro-feature screening and determination of specific data acquisition parameters.

[0065] This application addresses the problem of potential spatial and temporal mismatches in multi-source heterogeneous data from vast cornfields, which can hinder effective data fusion for initial assessments, by preprocessing and fusing auxiliary information. By acquiring and utilizing spatiotemporal metadata of macroscopic phenotypic data and auxiliary information, the system establishes correlations between them. Furthermore, spatial interpolation or temporal synchronization ensures that all auxiliary information remains spatiotemporally consistent with macroscopic phenotypic changes, preventing assessment biases caused by data misalignment. Simultaneously, weighting based on data source type and quality allows the system to prioritize high-quality, reliable data during information fusion, thereby improving the accuracy and reliability of initial assessments. This comprehensive evaluation, achieved through spatiotemporal alignment and weighted fusion, makes the initial assessment robust and precise, accurately indicating the potential types of macroscopic phenotypic changes. This provides guidance for subsequent diagnostic micro-feature screening and determination of specific data collection parameters, avoiding resource waste and insufficient information acquisition due to unclear initial assessments.

[0066] In some embodiments, the specific steps in step S4 include:

[0067] S41. Based on specific data acquisition parameters or combinations of parameters, drive the automated data acquisition device to adjust its sensor configuration and acquisition platform behavior, and perform data acquisition. At the same time, monitor the environmental parameters and the operating status of the automated data acquisition device during the data acquisition process.

[0068] S42. Based on the acquired diagnostic microscopic feature data, environmental parameters, and equipment operating status, assess the data quality of the diagnostic microscopic feature data and determine whether the data quality meets the preset data quality standards.

[0069] S43. If the data quality does not meet the preset data quality standard, then based on the reason for the data quality failure, determine the adaptive adjustment parameters for the sensor configuration or acquisition platform behavior, and drive the automated data acquisition device to readjust its sensor configuration and acquisition platform behavior according to the adaptive adjustment parameters in order to reacquire diagnostic microscopic feature data until the data quality of the diagnostic microscopic feature data meets the preset data quality standard or reaches the preset upper limit of the number of reacquisitions.

[0070] Specific data acquisition parameters or combinations of parameters refer to the specific settings used to guide automated data acquisition equipment in data acquisition. These can be achieved using sensor-related parameters such as sensor type, resolution, exposure time, gain, focal length, aperture, and white balance, as well as acquisition platform behavior-related parameters such as flight altitude, flight speed, hovering attitude, shooting angle, and path planning. Automated data acquisition equipment refers to devices capable of autonomously or semi-autonomously performing data acquisition tasks in field environments. These can take the form of unmanned aerial vehicles (UAVs), ground inspection robots, or fixed intelligent monitoring stations. Sensor configuration refers to the operating status and parameter settings of various sensors mounted on the automated data acquisition equipment. This can be achieved by adjusting the sensor's exposure time, gain, white balance, focal length, aperture, filter selection, or sensor array activation mode. Acquisition platform behavior refers to the motion mode and attitude control of the automated data acquisition equipment when performing data acquisition tasks. This can be achieved by adjusting the equipment's flight altitude, flight speed, hovering position, shooting angle, attitude, or planning specific acquisition paths. Environmental parameters refer to the external natural conditions that affect data quality during data acquisition, and can be represented by data such as light intensity, wind speed, temperature, humidity, rainfall, or air transparency. Equipment operating status refers to the internal working condition of automated data acquisition equipment during data acquisition, and can be represented by data such as battery level, attitude stability, sensor operating temperature, storage space, or internal vibration level. The data quality of diagnostic microscopic feature data refers to the degree to which the acquired diagnostic microscopic feature data meets the needs of subsequent analysis and diagnosis in terms of clarity, completeness, signal-to-noise ratio, color accuracy, or spatial resolution, and can be measured by indicators such as image clarity, signal-to-noise ratio, data completeness, or the identifiability of specific features. Preset data quality standards refer to the benchmarks or thresholds used to determine whether data meets requirements during data quality assessment, and can be defined by pre-set values ​​or sets of rules, such as a clarity threshold, a lower limit for signal-to-noise ratio, or the minimum identifiable size of a specific feature. The reasons for unsatisfactory data quality refer to technical or environmental factors that cause the acquired diagnostic microscopic feature data to fail to meet preset standards. These can be identified by specific issues such as insufficient lighting, overexposure, equipment vibration, sensor malfunction, or target blur. Adaptive parameter adjustment refers to targeted modifications to sensor configuration or acquisition platform behavior to improve data acquisition quality based on the reasons for unsatisfactory data quality. This can be determined by adjusting sensor exposure time, gain, focal length, or by adjusting the acquisition platform's flight altitude, speed, hovering attitude, or replanning local acquisition paths. The maximum number of re-acquisition attempts refers to the maximum number of times an automated data acquisition device is allowed to re-execute data acquisition when data quality is unsatisfactory. This can be limited by a preset integer value to avoid unlimited invalid re-acquisitions.

[0071] This solution incorporates a dynamic adaptive mechanism to ensure the quality and reliability of the acquired diagnostic microscopic feature data. Specifically, while the automated data acquisition equipment adjusts its sensor configuration and platform behavior based on specific data acquisition parameters or combinations, and performs data acquisition, the system simultaneously monitors environmental parameters and the operating status of the automated data acquisition equipment. This synchronous monitoring is crucial, enabling the system to acquire real-time information on external environmental factors affecting data quality and the internal operating status of the equipment, providing comprehensive foundational data for subsequent data quality assessment. Based on these real-time monitored environmental parameters and equipment operating status, as well as the acquired diagnostic microscopic feature data itself, the system comprehensively evaluates the data quality of the diagnostic microscopic feature data and determines whether it meets preset data quality standards. This multi-dimensional and comprehensive evaluation method can more accurately identify quality problems in the data, avoiding potential biases caused by relying solely on visual inspection or a single indicator. If the evaluation results show that the data quality does not meet the preset standards, the system will enter an adaptive adjustment phase. At this point, the system intelligently analyzes the specific reasons for the unsatisfactory data quality. Based on these identified causes, the system determines targeted adaptive adjustment parameters, which can be adjustments to sensor configuration or the behavior of the data acquisition platform. Subsequently, the system drives the automated data acquisition equipment to readjust its sensor configuration and platform behavior according to these adaptive adjustment parameters and performs data acquisition again. This process continues until the quality of the diagnostic microscopic feature data meets preset standards or reaches a preset limit on the number of re-acquisitions, avoiding an infinite number of invalid attempts. Through this iterative monitoring, evaluation, and adaptive adjustment mechanism, this solution effectively addresses the complex and ever-changing environmental conditions in the field, ensuring that the final acquired diagnostic microscopic feature data is high-quality and reliable. This is closely aligned with the overall method of this application, which aims to determine specific data acquisition parameters based on macroscopic phenotypic changes and ultimately classify these changes. High-quality diagnostic microscopic feature data is the foundation for accurate analysis and classification. By ensuring data quality, this solution directly improves the accuracy and reliability of subsequent phenotypic classification, making the diagnosis of problems such as pests, diseases, or nutrient deficiencies in maize plants more precise, thus providing solid data support for precision agriculture management.

[0072] This solution effectively addresses the challenge of automated data acquisition equipment continuously acquiring high-quality diagnostic micro-feature data in dynamically changing field environments by introducing synchronous monitoring of environmental parameters and equipment operating status during data acquisition, diagnostic micro-feature data quality assessment based on multi-dimensional information, and a mechanism for adaptive adjustment and re-acquisition based on the quality assessment results. This ensures the validity and reliability of the acquired data, avoids data quality degradation due to environmental interference or poor equipment condition, and thus provides a solid data foundation for the subsequent accurate classification and diagnosis of macro-phenotypic changes in maize, significantly improving diagnostic accuracy. Simultaneously, this adaptive mechanism reduces the need for manual intervention, improves the automation level and overall efficiency of data acquisition, and avoids resource waste caused by ineffective acquisition.

[0073] In some embodiments, step S42, which involves evaluating the data quality of the diagnostic microscopic feature data based on the acquired diagnostic microscopic feature data, environmental parameters, and device operating status, includes:

[0074] S421. Obtain information on the type of the diagnostic microscopic features to be evaluated;

[0075] S422. Based on the type information of diagnostic micro-features, determine a set of diagnostic specificity assessment indicators for evaluating diagnostic micro-feature data from a pre-defined diagnostic specificity data quality assessment rule base; the set of assessment indicators includes quality assessment dimensions directly related to the diagnostic purpose of the diagnostic micro-features.

[0076] S423. Based on environmental parameters and equipment operating status, adjust the weights or thresholds of each evaluation indicator in the diagnostic specificity evaluation indicator set to obtain the corrected evaluation indicator set.

[0077] S424. Evaluate the data quality of the diagnostic microscopic features based on the acquired diagnostic microscopic features data and the revised set of evaluation indicators.

[0078] The type information of diagnostic micro-features refers to the specific category or attribute identifier of the diagnostic micro-feature to be evaluated. This can be implemented using predefined codes, text labels, or enumerated values, such as "leaf lesion texture," "leaf three-dimensional morphology," or "specific spectral absorption peak." The diagnostic-specific data quality assessment rule base refers to a collection of data quality assessment standards and methods customized for different types of diagnostic micro-features. This can be stored in the form of a database, configuration file, or knowledge graph, containing assessment indicators, weights, and thresholds corresponding to each feature type. The diagnostic-specific assessment indicator set refers to a set of dimensions directly related to the diagnostic purpose of a specific diagnostic micro-feature, used to measure data quality. These can include image spatial resolution, texture sharpness, contrast, spectral signal-to-noise ratio, band accuracy, three-dimensional point cloud density, or morphological feature matching degree. Weight adjustment or threshold correction refers to changing the importance of evaluation indicators or adjusting the threshold value for judging whether data quality is qualified based on real-time environmental conditions or equipment performance status. It can be achieved by rule-based condition judgment, machine learning model output or preset lookup table. For example, increase the tolerance of image brightness index under low light conditions or reduce the weight of image sharpness index when the equipment is shaken.

[0079] Based on the above characteristics, the overall working principle of this scheme is as follows: when evaluating the data quality of acquired diagnostic microscopic feature data, the system first obtains the type information of the diagnostic microscopic feature to be evaluated. This is because different types of microscopic features, such as the texture of leaf lesions or the three-dimensional morphology of leaf curling, have different requirements for data quality. Based on this type information, the system determines a set of evaluation indicators directly related to the diagnostic purpose of the microscopic feature from a pre-set diagnostic-specific data quality evaluation rule base. This customized indicator selection ensures the specificity of the evaluation and avoids the limitations of general indicators. Furthermore, considering the real-time changes in environmental parameters (e.g., light intensity, wind speed) and equipment operating status (e.g., sensor stability, acquisition platform vibration) during data acquisition, the system adjusts the weights or corrects the thresholds of each indicator in the above evaluation indicator set. For example, when there is insufficient light, the brightness or contrast threshold of the image can be relaxed; when the equipment vibrates slightly, the image sharpness weight can be reduced. This dynamic correction allows the data quality evaluation to adapt to actual acquisition conditions, avoiding misjudgments caused by external factors, thereby making the evaluation results more accurate and robust. Finally, the system assesses data quality based on the acquired diagnostic microscopic feature data and a modified set of evaluation indicators. Through the synergistic effect of the above steps, this scheme achieves accurate and adaptive evaluation of the quality of diagnostic microscopic feature data. Compared with the basic scheme that only performs general data quality evaluation, this approach can more accurately reflect the actual diagnostic value and information completeness of diagnostic microscopic features. For example, for images that need to identify tiny lesions, the system focuses on spatial resolution and texture sharpness; for spectral data used to determine nutrient deficiency, it focuses on signal-to-noise ratio and absorption peak characteristics. This targeted evaluation ensures that even if the data is slightly deficient in some general indicators, it can still be deemed acceptable as long as the diagnostic information is sufficient, and vice versa. This effectively solves the problem that general evaluation methods cannot accurately reflect diagnostic value, ensures the accuracy of subsequent diagnoses, and avoids unnecessary duplicate data collection, thereby improving the overall efficiency of data acquisition and the reliability of diagnosis.

[0080] Through the above implementation methods, this solution achieves the following technical effects. By incorporating considerations of diagnostic microscopic feature types, environmental parameters, and equipment operating status, this solution enables accurate and adaptive evaluation of diagnostic microscopic feature data quality. This solves the problem that existing general data quality assessment methods cannot accurately reflect the diagnostic value or information completeness of different diagnostic microscopic features. By customizing evaluation indicators according to the type of feature to be evaluated and dynamically adjusting the weights or thresholds of these indicators in conjunction with real-time environmental and equipment status, this solution ensures that the data quality assessment results are highly matched with actual diagnostic needs. This avoids situations where data is qualified on general indicators but cannot support accurate diagnosis, or is misjudged as unqualified when it is slightly insufficient on general indicators but the diagnostic information is sufficient. Therefore, this solution ensures that the acquired data truly has diagnostic value, reduces unnecessary duplicate collection, and thus improves the efficiency of data collection and the accuracy of subsequent diagnosis.

[0081] In some embodiments, the specific steps in step S5 include:

[0082] S51. By analyzing diagnostic microscopic feature data, obtain the diagnostic microscopic feature patterns indicated in the diagnostic microscopic feature data;

[0083] S52. Obtain contextual information related to changes in macroscopic phenotypic characteristics; contextual information includes historical phenotypic data, phenotypic data of neighboring plants, or regional environmental data;

[0084] S53. Based on the correlation between diagnostic micro-feature patterns and contextual information, determine the auxiliary weight of contextual information for the classification of macro-phenotypic change types;

[0085] S54. Integrate diagnostic microscopic feature patterns with contextual information, and perform weighted processing on the fused information according to auxiliary weights to obtain comprehensive diagnostic information;

[0086] S55. Based on the comprehensive diagnostic information and the pre-set diagnostic feature patterns in the maize phenotypic feature map, classify the macroscopic phenotypic changes by type; the classification process includes identifying the similarity or overlap between the diagnostic microscopic feature patterns and the pre-set diagnostic feature patterns in the maize phenotypic feature map, and distinguishing them based on the comprehensive diagnostic information.

[0087] Diagnostic microscopic feature patterns refer to a refined set of features extracted from diagnostic microscopic feature data that characterizes specific macroscopic phenotypic changes. These patterns can be achieved using techniques such as image texture analysis, shape recognition, spectral feature extraction, or cell morphology analysis. Contextual information refers to auxiliary data related to macroscopic phenotypic changes in maize that goes beyond the microscopic features of a single plant. This can include historical phenotypic data, phenotypic data of neighboring plants, or regional environmental data, which can be obtained from agricultural IoT platforms, weather stations, or historical databases. The correlation between diagnostic microscopic feature patterns and contextual information refers to the logical or statistical relationship between the diagnostic microscopic feature patterns and different types of contextual information. This relationship can reveal the auxiliary value of contextual information in diagnosing specific microscopic patterns and can be established and evaluated using machine learning models or expert rule systems. Auxiliary weights refer to the degree of importance contributed by different contextual information to the auxiliary judgment of diagnostic microscopic feature patterns during the classification of macroscopic phenotypic change types. These weights can be dynamically determined based on preset rules, historical data analysis, or machine learning algorithms. The fusion of diagnostic micro-feature patterns and contextual information refers to the process of integrating feature patterns extracted from micro-data with multi-dimensional contextual information. This can be achieved using feature vector concatenation, multimodal data fusion algorithms, or deep learning networks. Weighted processing refers to the process of applying different influences to different information components in the fused information according to pre-determined auxiliary weights. This can be achieved using algorithms such as linear weighting, non-linear weighting, or attention mechanisms. Comprehensive diagnostic information refers to the comprehensive set of information, after weighted fusion processing, that simultaneously includes diagnostic micro-feature patterns and contextual information, used for final classification. The pre-defined diagnostic feature patterns in the maize phenotypic feature map refer to the standard micro-feature patterns pre-stored in the maize phenotypic feature map that correspond to various known macro-phenotypic change types. The classification process involves identifying the similarity or overlap between diagnostic micro-feature patterns and preset diagnostic feature patterns in the maize phenotypic feature map, and distinguishing them based on comprehensive diagnostic information. This means that when classifying macro-phenotypic changes, it is not only necessary to compare the similarity between the currently acquired diagnostic micro-feature patterns and preset patterns in the map, but more importantly, to use comprehensive diagnostic information to resolve confusion between similar patterns, thereby achieving more accurate type attribution.

[0088] The proposed solution, based on the dynamic collection of maize field phenotypic data and the acquisition of diagnostic microscopic feature data, further optimizes the classification process for macroscopic phenotypic changes. Specifically, after acquiring diagnostic microscopic feature data from automated data acquisition equipment, the system first extracts indicative diagnostic microscopic feature patterns through in-depth analysis of this data. This step forms the basis for subsequent classification, ensuring that key, diagnostically applicable microscopic information is obtained from the refined data collection. Building upon this, the system further acquires contextual information associated with macroscopic phenotypic changes. This contextual information, such as historical phenotypic data, neighboring plant phenotypic data, or regional environmental data, provides broader background information beyond the current microscopic features of a single plant. Historical data can reveal trends and stages of change, helping to determine the persistence and severity of current changes; neighboring plant data can help determine whether it is a local or regional problem, thus distinguishing between individual differences and generalized stress; while regional environmental data can provide important clues such as environmental stress or nutrient status, providing environmental context for understanding the potential causes of phenotypic changes. By acquiring this multi-dimensional and multi-source contextual information, a more comprehensive basis is provided for subsequent integrated judgment, compensating for the shortcomings of single micro-feature information. To intelligently utilize this contextual information, the system dynamically determines the auxiliary weights of contextual information for classifying macroscopic phenotypic change types based on the correlation between diagnostic micro-feature patterns and contextual information. This mechanism recognizes that different contextual information may have different diagnostic values ​​for specific micro-feature patterns; for example, in a certain micro-pattern, historical data may be more important than environmental data. Dynamically determining auxiliary weights ensures that contextual information plays its optimal auxiliary role in the classification process, avoiding the limitations of simple averaging or fixed weights, enabling the system to utilize auxiliary information more flexibly and accurately. Subsequently, the diagnostic micro-feature patterns are fused with the acquired contextual information, and the fused information is weighted according to the determined auxiliary weights to obtain comprehensive diagnostic information. This weighted fusion mechanism allows the system to fully utilize the direct diagnostic capabilities of micro-features and the auxiliary judgment capabilities of contextual information, forming a more comprehensive and convincing diagnostic basis. The comprehensive diagnostic information not only includes the most direct microscopic evidence but also incorporates multi-dimensional background knowledge such as time, space, and environment, thus providing a more solid foundation for the final classification. Finally, based on the obtained comprehensive diagnostic information and the pre-defined diagnostic feature patterns in the maize phenotypic feature map, macroscopic phenotypic changes are classified. The classification process particularly emphasizes identifying the similarity or overlap between the diagnostic microscopic feature patterns and the pre-defined diagnostic feature patterns in the maize phenotypic feature map, and distinguishing them based on the comprehensive diagnostic information. This means that classification no longer relies solely on a simple match between microscopic feature patterns and the map, but rather on refining and modifying this match through comprehensive diagnostic information.For example, when two different diseases exhibit similar patterns at the microscopic level, comprehensive diagnostic information (such as historical disease occurrence and environmental humidity contextual information) can help the system make more accurate distinctions, thereby avoiding misjudgments and achieving precise classification of macroscopic phenotypic changes. Overall, this solution, after acquiring highly targeted diagnostic microscopic feature data, effectively solves the problem of difficulty in accurately distinguishing single microscopic feature patterns in complex field environments by introducing and intelligently integrating multi-dimensional contextual information. This comprehensive diagnostic mechanism, combining microscopic details and macroscopic background, makes the classification of macroscopic phenotypic changes in maize more accurate and robust, thus providing a more reliable basis for subsequent precision agriculture management decisions. This approach, combined with the aforementioned method for acquiring diagnostic microscopic feature data, forms a complete closed loop from data collection to intelligent diagnosis, significantly improving the practicality and effectiveness of the entire dynamic phenotypic data collection method.

[0089] This solution effectively addresses the challenge of accurately distinguishing single microscopic feature patterns in complex field environments by introducing and intelligently integrating contextual information with diagnostic microscopic feature patterns. Specifically, when microscopic feature patterns of different types of phenotypic changes exhibit similarity or overlap—for example, when microscopic spots of early fungal diseases differ slightly in morphology from certain physiological spots—this solution can utilize contextual information such as historical phenotypic data, phenotypic data of adjacent plants, or regional environmental data to assist in judging and weighting the distinction of these similar patterns. This enables the system to avoid misjudgments or diagnostic ambiguity, thereby achieving accurate classification of macroscopic phenotypic changes and providing a more reliable and refined basis for subsequent agricultural management decisions.

[0090] Reference Appendix Figure 2 This invention provides a dynamic data acquisition system for maize field phenotypic sets, comprising:

[0091] The acquisition module 100 is used to acquire macroscopic phenotypic data of maize plants and identify macroscopic phenotypic changes indicated in the macroscopic phenotypic data.

[0092] The query module 200 is used to query a preset maize phenotypic feature map based on the identified macroscopic phenotypic changes. The maize phenotypic feature map includes diagnostic microscopic features associated with macroscopic phenotypic changes, as well as specific data acquisition parameters required to obtain the diagnostic microscopic features.

[0093] The determination module 300 is used to determine, based on the query results, specific data acquisition parameters or parameter combinations for obtaining diagnostic microscopic features; the parameter combination includes multiple specific data acquisition parameters.

[0094] The control acquisition module 400 is used to acquire diagnostic microscopic feature data by driving the automated data acquisition equipment to adjust its sensor configuration and acquisition platform behavior according to specific data acquisition parameters or parameter combinations.

[0095] The analysis module 500 is used to analyze diagnostic microscopic feature data and classify macroscopic phenotypic changes according to the analysis results and the preset diagnostic feature patterns in the maize phenotypic feature map.

[0096] In this document, relational terms such as first and second are used only to distinguish one entity or operation from another entity or operation, without necessarily requiring or implying any such actual relationship or order between these entities or operations.

[0097] The above description is merely an embodiment of the present invention and is not intended to limit the scope of protection of the present invention. For those skilled in the art, the present invention can have various modifications and variations. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for dynamically collecting maize field phenotypic data, characterized in that, Includes the following steps: S1. Obtain macroscopic phenotypic data of maize plants and identify macroscopic phenotypic changes indicated in the macroscopic phenotypic data; S2. Based on the identified macroscopic phenotypic changes, query the preset maize phenotypic feature map; the maize phenotypic feature map includes diagnostic microscopic features associated with macroscopic phenotypic changes, as well as specific data acquisition parameters required to obtain diagnostic microscopic features; S3. Based on the query results, determine the specific data acquisition parameters or parameter combinations used to obtain diagnostic microscopic features; the parameter combination includes multiple specific data acquisition parameters; S4. Based on specific data acquisition parameters or combinations thereof, diagnostic microscopic feature data is obtained by adjusting the sensor configuration and acquisition platform behavior of the automated data acquisition equipment; sensor configuration and acquisition platform behavior refer to the hardware settings used for data acquisition on the automated data acquisition equipment and its movement in space. S5. Analyze the diagnostic microscopic feature data, and classify the macroscopic phenotypic changes according to the analysis results and the pre-set diagnostic feature patterns in the maize phenotypic feature map.

2. The method for dynamic acquisition of maize field phenotypic data according to claim 1, characterized in that... The specific steps in step S2 include: S21. Obtain auxiliary information related to the identified macroscopic phenotypic changes; S22. Based on the identified macrophenotypic changes and auxiliary information, make a preliminary judgment on the macrophenotypic changes to obtain a judgment result indicating the potential type of macrophenotypic changes; S23. Based on the judgment result, select diagnostic micro-features that match the judgment result from the preset maize phenotypic feature map, and obtain the specific data acquisition parameters required for the diagnostic micro-features.

3. The method for dynamic acquisition of maize field phenotypic data according to claim 2, characterized in that, Supporting information includes the growth stage of the corn plants, environmental data of the area, or soil test data.

4. The method for dynamic acquisition of maize field phenotypic data according to claim 2, characterized in that, The specific steps in step S22 include: A1. Determine the time for acquiring auxiliary information; A2. Based on the time interval between the acquisition time of auxiliary information and the recognition time of macroscopic phenotypic changes, determine whether the auxiliary information needs to be updated according to the preset auxiliary information validity rules; A3. If the auxiliary information needs to be updated, obtain the latest data of the auxiliary information and use the latest data as the auxiliary information for preliminary judgment; A4. If the auxiliary information does not need to be updated, the already acquired auxiliary information shall be used as the auxiliary information for preliminary judgment; A5. Based on the identified macroscopic phenotypic changes and the auxiliary information used for preliminary judgment, a preliminary judgment is made on the macroscopic phenotypic changes to obtain a judgment result indicating the potential type of macroscopic phenotypic changes.

5. The method for dynamic acquisition of maize field phenotypic data according to claim 2, characterized in that, The specific steps in step S22 include: B1. Obtain the spatial location and identification time of the macroscopic phenotypic data, as well as the spatial location and acquisition time of the auxiliary information; B2. Based on the spatial location and identification time of the macroscopic phenotypic data, as well as the spatial location and acquisition time of the auxiliary information, determine the spatial and temporal correlations between the macroscopic phenotypic data and each piece of auxiliary information; B3. Based on spatial and temporal correlations, by performing spatial interpolation or temporal synchronization processing on each auxiliary information, a set of auxiliary information aligned with the macroscopic phenotypic data in space and time is generated; B4. Determine the weight of each piece of auxiliary information in the preliminary judgment based on the data source type or data quality of each piece of auxiliary information; B5. The identified macroscopic phenotypic changes are fused with the aligned set of auxiliary information, and the fused information is comprehensively evaluated according to the weight of each auxiliary information in the preliminary judgment to obtain the judgment result indicating the potential type of macroscopic phenotypic changes.

6. The method for dynamic acquisition of maize field phenotypic data according to claim 1, characterized in that, The specific steps in step S4 include: S41. Based on specific data acquisition parameters or combinations of parameters, drive the automated data acquisition device to adjust its sensor configuration and acquisition platform behavior, and perform data acquisition. At the same time, monitor the environmental parameters and the operating status of the automated data acquisition device during the data acquisition process. S42. Based on the acquired diagnostic microscopic feature data, environmental parameters, and equipment operating status, assess the data quality of the diagnostic microscopic feature data and determine whether the data quality meets the preset data quality standards. S43. If the data quality does not meet the preset data quality standard, then based on the reason for the data quality failure, determine the adaptive adjustment parameters for the sensor configuration or acquisition platform behavior, and drive the automated data acquisition device to readjust its sensor configuration and acquisition platform behavior according to the adaptive adjustment parameters in order to reacquire diagnostic microscopic feature data until the data quality of the diagnostic microscopic feature data meets the preset data quality standard or reaches the preset upper limit of the number of reacquisitions.

7. The method for dynamic acquisition of maize field phenotypic data according to claim 6, characterized in that, In step S42, the step of evaluating the data quality of the diagnostic microscopic feature data based on the acquired diagnostic microscopic feature data, environmental parameters, and equipment operating status includes: S421. Obtain information on the type of the diagnostic microscopic features to be evaluated; S422. Based on the type information of diagnostic micro-features, determine the set of diagnostic specificity assessment indicators for evaluating diagnostic micro-feature data from the preset diagnostic specificity data quality assessment rule base; S423. Based on environmental parameters and equipment operating status, adjust the weights or thresholds of each evaluation indicator in the diagnostic specificity evaluation indicator set to obtain the corrected evaluation indicator set. S424. Evaluate the data quality of the diagnostic microscopic features based on the acquired diagnostic microscopic features data and the revised set of evaluation indicators.

8. The method for dynamic acquisition of maize field phenotypic data according to claim 1, characterized in that, The specific steps in step S5 include: S51. By analyzing diagnostic microscopic feature data, obtain the diagnostic microscopic feature patterns indicated in the diagnostic microscopic feature data; S52. Obtain contextual information related to changes in macroscopic phenotypic features; S53. Based on the correlation between diagnostic micro-feature patterns and contextual information, determine the auxiliary weight of contextual information for the classification of macro-phenotypic change types; S54. Integrate diagnostic microscopic feature patterns with contextual information, and perform weighted processing on the fused information according to auxiliary weights to obtain comprehensive diagnostic information; S55. Based on the comprehensive diagnostic information and the pre-set diagnostic feature patterns in the maize phenotypic feature map, classify the macroscopic phenotypic changes by type.

9. The method for dynamic acquisition of maize field phenotypic group data according to claim 8, characterized in that, The specific steps in step S55 include: Identify the similarity or overlap between diagnostic micro-feature patterns and pre-defined diagnostic feature patterns in maize phenotypic feature maps, and differentiate them based on comprehensive diagnostic information.

10. A dynamic data acquisition system for maize field phenotypic sets, characterized in that, include: The acquisition module is used to acquire macroscopic phenotypic data of maize plants and identify macroscopic phenotypic changes indicated in the macroscopic phenotypic data. The query module is used to query a preset maize phenotypic feature map based on the identified macroscopic phenotypic changes. The maize phenotypic feature map includes diagnostic microscopic features associated with macroscopic phenotypic changes, as well as specific data acquisition parameters required to obtain the diagnostic microscopic features. The determination module is used to determine, based on the query results, specific data acquisition parameters or combinations of parameters for obtaining diagnostic microscopic features; the parameter combination includes multiple specific data acquisition parameters. The control acquisition module is used to obtain diagnostic microscopic feature data by driving automated data acquisition equipment to adjust its sensor configuration and acquisition platform behavior according to specific data acquisition parameters or parameter combinations. Sensor configuration and acquisition platform behavior refers to the hardware setup used for data acquisition on automated data acquisition equipment and its movement in space; The analysis module is used to analyze diagnostic microscopic feature data and classify macroscopic phenotypic changes based on the analysis results and the preset diagnostic feature patterns in the maize phenotypic feature map.

Images