High-throughput plant phenotype dynamic monitoring system and method based on SPAC factor

Abstract

The invention discloses a high-throughput plant phenotype dynamic monitoring system and method based on SPAC factors, and relates to the technical field of plant phenotype dynamic monitoring. Comprising a data acquisition module used for acquiring image data of a plant and acquiring SPAC factor data of a plant growth environment; the data processing module is used for processing the collected data; and the feature extraction module is used for extracting features from the processed data. According to the method, a correlation analysis and information entropy-based weight determination method is adopted, a correlation model of the SPAC factor and the plant phenotype is established, the influence degree of each environmental factor on the plant phenotype can be accurately reflected, and a basis is provided for prediction and scene analysis of the plant phenotype; on the basis of the established relevancy model, the system can accurately deduce the future change trend of the plant phenotype in combination with SPAC factor data collected in real time, and introduces a confidence interval to evaluate the reliability of a prediction result; and the influence of different environmental factors on plant growth can be known.

Description

Technical Field

[0001] This invention relates to the field of plant phenotypic dynamic monitoring technology, and in particular to a high-throughput plant phenotypic dynamic monitoring system and method based on SPAC factors. Background Technology

[0002] In modern agricultural and plant science research, monitoring and analyzing plant phenotypes is crucial for understanding plant growth patterns, optimizing crop management, and improving yield and quality. However, traditional methods of monitoring plant phenotypes often rely on manual observation and measurement, which is not only time-consuming and labor-intensive but also difficult to implement continuously, failing to meet the needs of modern research.

[0003] In recent years, with the rapid development of information technology and sensing technology, research on plant phenotypic monitoring using technologies such as image recognition and sensor networks has gradually increased. These technologies enable rapid and non-destructive monitoring of plant growth status.

[0004] A search revealed Chinese patent application CN202211244907.3, which discloses a high-throughput plant phenotyping platform and method for field applications. The transportation system includes a transport track and an electric intelligent transport vehicle to transport potted plants from the field to an imaging chamber for image acquisition. After imaging, the potted plants are transported back to the field. The imaging system includes an imaging chamber equipped with a light source, a trigger camera, and multiple image acquisition devices to acquire plant phenotypic information and match this information with experimental treatment information. The plant phenotyping platform in this patent has the following shortcomings: it only performs image acquisition and analysis of plant phenotypes. While it possesses some analytical capabilities, it cannot analyze or predict plant phenotypic changes based on the SPAC factor. Summary of the Invention

[0005] The purpose of this invention is to address the shortcomings of existing technologies by proposing a high-throughput plant phenotypic dynamic monitoring system and method based on SPAC factors.

[0006] To achieve the above objectives, the present invention adopts the following technical solution:

[0007] A high-throughput plant phenotypic dynamic monitoring system based on SPAC factors includes:

[0008] The data acquisition module is used to collect image data of plants and SPAC factor data of plant growth environment;

[0009] The data processing module processes the collected data;

[0010] The feature extraction module extracts features from the processed data;

[0011] The model management module initially establishes a correlation model between SPAC factors and plant phenotypes based on the extracted features. This model should reflect the influence of various environmental factors on various plant characteristics, and a correlation weight should be set for each factor.

[0012] The extrapolation module extrapolates future trends in plant phenotypic changes based on the established correlation model and real-time collected SPAC factor data.

[0013] The data evaluation module assesses the accuracy of the simulation module and provides feedback to the model management module. The model management module then dynamically adjusts the correlation weights of each factor in the correlation model based on the accuracy.

[0014] Preferably, the data acquisition module includes:

[0015] The image data acquisition submodule uses a multispectral imaging device to acquire data at a preset acquisition frequency, capturing phenotypic changes in plants at different growth stages.

[0016] The SPAC factor data acquisition submodule uses soil sensors, plant physiological monitoring instruments, and meteorological sensors to measure SPAC factor data in real time during plant growth. Specifically, the soil sensors are placed at the plant roots to monitor soil environmental parameters in real time; the plant physiological monitoring instruments acquire photosynthetic characteristic parameters of the plants in real time; and the meteorological sensors monitor atmospheric environmental parameters.

[0017] Preferably, the data processing module includes:

[0018] The image data processing submodule preprocesses the acquired raw images to improve their quality and clarity; and uses an image segmentation algorithm to separate the plants from the background.

[0019] The SPAC factor data processing submodule removes outliers and erroneous data from the collected SPAC factor data according to a set range.

[0020] Preferably, the SPAC factor data processing submodule arranges the collected data according to the time axis, and denotes them sequentially as C1, C2, ..., C... n The entire data exhibits a linear variation; if at a certain moment C... i The value compared to C i-1 C i+1 There is a significant change, namely C i ≥[(C i-1 +C i+1 ) / 2]×(1+X)%, or C i <[(C i-1 +C i+1 If ) / 2]×(1-X)%, where X is a set value, then it is considered an abnormal value, and the Ci Delete, or remove C i The value is replaced with (C) i-1 +C i+1 ) / 2.

[0021] Preferably, the feature extraction module includes:

[0022] The conventional phenotypic feature extraction submodule is used to extract conventional phenotypic features, including plant size features and leaf density features;

[0023] The abnormal phenotypic feature extraction submodule is used to extract abnormal phenotypic features, including localized yellowing of leaves and localized curling of plants.

[0024] Preferably, the model management module includes:

[0025] The correlation analysis submodule performs correlation analysis based on the extracted plant phenotypic characteristics and SPAC factor data, using Pearson correlation coefficient or Spearman correlation coefficient to calculate the correlation coefficient r between each SPAC factor and the plant phenotypic characteristics. ij , where i represents the i-th SPAC factor and j represents the j-th plant phenotypic trait; construct a correlation matrix, the elements of which are the correlation coefficients between each SPAC factor and the plant phenotypic trait. By analyzing the correlation matrix, the association between each SPAC factor and the plant phenotypic trait is preliminarily determined.

[0026] The weight determination submodule determines the influence weights w of each SPAC factor on plant phenotypic traits based on the results of correlation analysis. i The weights are determined using an information entropy-based method. First, the information entropy H of each SPAC factor is calculated. i :

[0027]

[0028] Where, p k The probability of the i-th SPAC factor taking the k-th value; then, the weights are determined based on information entropy:

[0029]

[0030] The model building submodule establishes a correlation model between SPAC factors and plant phenotypes based on determined weights. This model employs a multiple linear regression model, with n SPAC factors x1, x2, ..., xn. n and m plant phenotypic characteristics y1, y2, ..., y m The model is represented as:

[0031]

[0032] Where, β j0 For the intercept term, β ji For regression coefficients, ∈ j The error term is used; the model parameters are estimated using the least squares method to obtain the specific model equations.

[0033] Preferably, the deduction module includes:

[0034] The trend prediction submodule predicts the trend of plant phenotypic changes over a future period based on the established correlation model and real-time collected SPAC factor data. It substitutes the real-time SPAC factor data into the model equation to calculate the predicted values ​​of plant phenotypic characteristics at future times. Where t is the current time and k is the prediction time step; confidence intervals are introduced to evaluate the reliability of the prediction results; and the standard error of the prediction value is calculated. Determine the confidence interval for the predicted value:

[0035]

[0036] Among them, z a / 2 denoted as the quantile of the standard normal distribution, and α is the significance level;

[0037] The scenario analysis submodule simulates the changes in plant phenotypes under different SPAC factor change scenarios set by the user, based on the correlation model.

[0038] Preferably, the data evaluation module includes:

[0039] The accuracy assessment submodule compares the prediction results from the inference module with the actual observed plant phenotypic data to calculate the prediction accuracy (ACC). The formula for calculating the prediction accuracy is as follows:

[0040]

[0041] Where m is the number of plant phenotypic traits, and N cj The number of samples that correctly predict the j-th plant phenotypic trait, N tj The total number of samples for the j-th plant phenotypic trait;

[0042] The feedback adjustment submodule feeds back the accuracy evaluation index to the model management module based on the data evaluation results. The model management module then dynamically adjusts the correlation weights of each factor in the correlation model based on the feedback information. The adjustment method uses either gradient descent or genetic algorithm to continuously optimize the model parameters and weights, thereby improving the accuracy and reliability of the model.

[0043] Preferably, the accuracy evaluation submodule further uses root mean square error (RMSE) and mean absolute error (MAE) to evaluate the accuracy of the prediction results; wherein, the formula for calculating the RMSE is:

[0044]

[0045] The formula for calculating the mean absolute error is:

[0046]

[0047] in, Let y be the predicted value of the j-th plant phenotypic trait on the i-th sample. ji These are actual observed values.

[0048] A high-throughput plant phenotypic dynamic monitoring method based on SPAC factors, implemented using the aforementioned system, specifically includes the following steps:

[0049] S1: The data acquisition module continuously acquires plant image data and SPAC factor data according to the set time interval and acquisition method, and transmits the data to the data processing module;

[0050] S2: The data processing module performs preprocessing operations on the acquired image data and SPAC factor data to obtain the processed information;

[0051] S3: The feature extraction module extracts the plant's conventional and abnormal phenotypic features based on the processed data and transmits the feature data to the model management module.

[0052] S4: The model management module performs correlation analysis, weight determination, and model building based on the extracted features and SPAC factor data, generating a correlation model between SPAC factors and plant phenotypes.

[0053] S5: The inference module predicts and analyzes future plant phenotypic trends based on the established correlation model and real-time collected SPAC factor data, and outputs the results.

[0054] S6: The data evaluation module compares the prediction results of the inference module with the actual observation data, evaluates the accuracy of the prediction, and feeds back the evaluation results to the model management module.

[0055] S7: The model management module dynamically adjusts model parameters and weights based on feedback information to continuously optimize model performance.

[0056] The beneficial effects of this invention are as follows:

[0057] This invention employs correlation analysis and a weight determination method based on information entropy to establish a correlation model between SPAC factors and plant phenotypes. This model accurately reflects the degree of influence of various environmental factors on plant phenotypes, providing a basis for plant phenotype prediction and scenario analysis.

[0058] Based on the established correlation model, this invention enables the system to accurately predict future trends in plant phenotypes by combining real-time collected SPAC factor data, and introduces confidence intervals to assess the reliability of the prediction results; it helps researchers and agricultural producers understand the impact of different environmental factors on plant growth.

[0059] 1. This invention can evaluate the accuracy of the inference module in real time and feed the evaluation results back to the model management module. The model management module dynamically adjusts the correlation weights of each factor in the correlation model based on the feedback information, continuously optimizing the model parameters and weights, thereby improving the accuracy and reliability of the model. Attached Figure Description

[0060] Figure 1 This is a framework diagram of the high-throughput plant phenotypic dynamic monitoring system based on SPAC factors proposed in this invention;

[0061] Figure 2 This is a flowchart of the high-throughput plant phenotypic dynamic monitoring method based on SPAC factors proposed in this invention. Detailed Implementation

[0062] The technical solution of the present invention will be further described in detail below with reference to specific embodiments.

[0063] Example 1:

[0064] A high-throughput plant phenotypic dynamic monitoring system based on SPAC factors, the system comprising:

[0065] The data acquisition module is used to collect image data of plants and SPAC factor data of plant growth environment;

[0066] The data processing module processes the collected data;

[0067] The feature extraction module extracts features from the processed data, such as common features like plant size and leaf density, as well as abnormal features like localized yellowing of plant leaves and localized curling of plant parts.

[0068] The model management module initially establishes a correlation model between SPAC factors and plant phenotypes based on the extracted features. This model should reflect the influence of various environmental factors on various plant characteristics. Each factor should be assigned a correlation weight, and the factors with larger correlation weights have more significant effects.

[0069] The extrapolation module extrapolates future trends in plant phenotypic changes based on the established correlation model and real-time collected SPAC factor data.

[0070] The data evaluation module assesses the accuracy of the simulation module and provides feedback to the model management module. The model management module then dynamically adjusts the correlation weights of each factor in the correlation model based on the accuracy.

[0071] The data acquisition module includes:

[0072] The image data acquisition submodule uses a multispectral imaging device to acquire data at a preset acquisition frequency, capturing phenotypic changes in plants at different growth stages, such as a camera array.

[0073] The SPAC factor data acquisition submodule uses soil sensors, plant physiological monitoring instruments, and meteorological sensors to measure SPAC factor data in real time during plant growth. The soil sensors are placed at the plant roots to monitor soil environmental parameters in real time, including soil moisture, soil temperature, and soil pH. The plant physiological monitoring instruments, such as chlorophyll fluorometers and photosynthetic rate meters, acquire photosynthetic characteristic parameters of plants in real time. The meteorological sensors monitor atmospheric environmental parameters, including air temperature, relative humidity, air pressure, and light intensity.

[0074] The data processing module includes:

[0075] The image data processing submodule preprocesses the acquired raw images, including denoising and contrast enhancement, to improve image quality and clarity. It also employs an image segmentation algorithm to separate the plants from the background for subsequent extraction of plant phenotypic features. Furthermore, it performs morphological processing on the segmented plant images, such as dilation and erosion, to fill small holes in the image, remove noise such as burrs, and obtain clearer plant image outlines.

[0076] The SPAC factor data processing submodule removes outliers and erroneous data from the collected SPAC factor data according to a set range; for example, the collected data is arranged along a time axis and labeled C1, C2, ..., C... n The entire data exhibits a linear variation; if at a certain moment C... i The value compared to C i-1 C i+1 There is a significant change, namely C i ≥[(C i-1 +C i+1 ) / 2]×(1+X)%, or C i <[(C i-1 +C i+1If ) / 2]×(1-X)%, where X is a set value, then it is considered an abnormal value, and the C i Delete, or remove C i The value is replaced with (C) i-1 +C i+1 ) / 2.

[0077] The SPAC factor data processing submodule also normalizes the collected SPAC factor data, converting it into a uniform scale for subsequent data analysis and model building.

[0078] The feature extraction module includes:

[0079] The conventional phenotypic feature extraction submodule is used to extract conventional phenotypic features such as plant size and leaf density. Specifically, for plant size features, conventional size features such as canopy area, plant height, and stem diameter are calculated by analyzing the processed plant image; canopy area is obtained through pixel statistics and geometric calculations, while plant height and stem diameter are measured through calibration scale and feature point recognition in the image; for leaf density features, image analysis technology is used to count the number of plant leaves and leaf area, and leaf density is calculated.

[0080] The abnormal phenotypic feature extraction submodule is used to extract abnormal phenotypic features such as localized yellowing of leaves and localized curling of plants. Specifically, for localized yellowing of leaves, a threshold range for yellowing is set in the RGB color space of the image. The yellowed areas are identified by judging the RGB values ​​of each pixel. The proportion of the yellowed area to the total leaf area is calculated as an indicator describing the abnormal feature of localized yellowing of leaves. For localized curling of plants, the curvature changes in the plant area are detected by analyzing the contour curve of the plant image. When the curvature exceeds a set threshold, curling is determined to exist in that area. The area, perimeter, maximum curling curvature, and average curling curvature of the curled area are calculated to quantify the abnormal feature of localized curling of plants.

[0081] The model management module includes:

[0082] The correlation analysis submodule performs correlation analysis based on the extracted plant phenotypic characteristics and SPAC factor data, using Pearson correlation coefficient or Spearman correlation coefficient to calculate the correlation coefficient r between each SPAC factor and the plant phenotypic characteristics. ij , where i represents the i-th SPAC factor and j represents the j-th plant phenotypic trait; construct a correlation matrix, the elements of which are the correlation coefficients between each SPAC factor and the plant phenotypic trait. By analyzing the correlation matrix, the association between each SPAC factor and the plant phenotypic trait is preliminarily determined.

[0083] The weight determination submodule determines the influence weights w of each SPAC factor on plant phenotypic traits based on the results of correlation analysis. i The weights are determined using an information entropy-based method. First, the information entropy H of each SPAC factor is calculated. i :

[0084]

[0085] Where, p k The probability of the i-th SPAC factor taking the k-th value; then, the weights are determined based on information entropy:

[0086]

[0087] By adopting an information entropy-based method, the information content of each SPAC factor and its influence on plant phenotypic characteristics can be fully considered, making the weight allocation more reasonable.

[0088] The model building submodule establishes a correlation model between SPAC factors and plant phenotypes based on determined weights. This model employs a multiple linear regression model, with n SPAC factors x1, x2, ..., xn. n and m plant phenotypic characteristics y1, y2, ..., y m The model is represented as:

[0089]

[0090] Where, β j0 For the intercept term, β ji For regression coefficients, ∈ j The error term is used; the model parameters are estimated using the least squares method to obtain the specific model equations.

[0091] The inference module includes:

[0092] The trend prediction submodule predicts the trend of plant phenotypic changes over a future period based on the established correlation model and real-time collected SPAC factor data. It substitutes the real-time SPAC factor data into the model equation to calculate the predicted values ​​of plant phenotypic characteristics at future times. Where t is the current time and k is the prediction time step; confidence intervals are introduced to evaluate the reliability of the prediction results; and the standard error of the prediction value is calculated. Determine the confidence interval for the predicted value:

[0093]

[0094] Among them, z a / 2 denoted as the quantile of the standard normal distribution, and α is the significance level;

[0095] The scenario analysis submodule simulates the changes in plant phenotypes under different SPAC factor change scenarios set by the user, such as changes in soil moisture and temperature, based on the correlation model.

[0096] By setting up a scenario analysis submodule, researchers and agricultural producers can understand the impact of different environmental factors on plant growth, so as to better formulate reasonable cultivation and management measures.

[0097] The data evaluation module includes:

[0098] The accuracy assessment submodule compares the prediction results from the inference module with the actual observed plant phenotypic data to calculate the prediction accuracy (ACC). The formula for calculating the prediction accuracy is as follows:

[0099]

[0100] Where m is the number of plant phenotypic traits, and N cj The number of samples that correctly predict the j-th plant phenotypic trait, N tj The total number of samples for the j-th plant phenotypic trait;

[0101] The feedback adjustment submodule feeds back the accuracy evaluation index to the model management module based on the data evaluation results. The model management module then dynamically adjusts the correlation weights of each factor in the correlation model based on the feedback information. The adjustment method uses either gradient descent or genetic algorithm to continuously optimize the model parameters and weights, thereby improving the accuracy and reliability of the model.

[0102] The accuracy assessment submodule further uses root mean square error (RMSE) and mean absolute error (MAE) to evaluate the accuracy of the prediction results; the formula for calculating RMSE is:

[0103]

[0104] The formula for calculating the mean absolute error is:

[0105]

[0106] in, Let y be the predicted value of the j-th plant phenotypic trait on the i-th sample. ji These are actual observed values.

[0107] Example 2:

[0108] This embodiment of the high-throughput plant phenotypic dynamic monitoring method based on SPAC factor, based on the system implementation of Embodiment 1, specifically includes the following steps:

[0109] S1: The data acquisition module continuously acquires plant image data and SPAC factor data according to the set time interval and acquisition method, and transmits the data to the data processing module;

[0110] S2: The data processing module performs preprocessing operations on the acquired image data and SPAC factor data to obtain the processed information;

[0111] S3: The feature extraction module extracts the plant's conventional and abnormal phenotypic features based on the processed data and transmits the feature data to the model management module.

[0112] S4: The model management module performs correlation analysis, weight determination, and model building based on the extracted features and SPAC factor data, generating a correlation model between SPAC factors and plant phenotypes.

[0113] S5: The inference module predicts and analyzes future plant phenotypic trends based on the established correlation model and real-time collected SPAC factor data, and outputs the results.

[0114] S6: The data evaluation module compares the prediction results of the inference module with the actual observation data, evaluates the accuracy of the prediction, and feeds back the evaluation results to the model management module.

[0115] S7: The model management module dynamically adjusts model parameters and weights based on feedback information to continuously optimize model performance.

[0116] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A high-throughput plant phenotypic dynamic monitoring system based on SPAC factors, characterized in that, The system includes: The data acquisition module is used to collect image data of plants and SPAC factor data of plant growth environment; The data processing module processes the collected data; The feature extraction module extracts features from the processed data; The model management module initially establishes a correlation model between SPAC factors and plant phenotypes based on the extracted features. This model should reflect the influence of various environmental factors on various plant characteristics, and a correlation weight should be set for each factor. The extrapolation module extrapolates future trends in plant phenotypic changes based on the established correlation model and real-time collected SPAC factor data. The data evaluation module assesses the accuracy of the simulation module and provides feedback to the model management module. The model management module then dynamically adjusts the correlation weights of each factor in the correlation model based on the accuracy.

2. The high-throughput plant phenotypic dynamic monitoring system based on SPAC factors according to claim 1, characterized in that, The data acquisition module includes: The image data acquisition submodule uses a multispectral imaging device to acquire data at a preset acquisition frequency, capturing phenotypic changes in plants at different growth stages. The SPAC factor data acquisition submodule uses soil sensors, plant physiological monitoring instruments, and meteorological sensors to measure SPAC factor data in real time during plant growth. Specifically, the soil sensors are placed at the plant roots to monitor soil environmental parameters in real time; the plant physiological monitoring instruments acquire photosynthetic characteristic parameters of the plants in real time; and the meteorological sensors monitor atmospheric environmental parameters.

3. The high-throughput plant phenotypic dynamic monitoring system based on SPAC factors according to claim 1, characterized in that, The data processing module includes: The image data processing submodule preprocesses the acquired raw images to improve their quality and clarity; and uses an image segmentation algorithm to separate the plants from the background. The SPAC factor data processing submodule removes outliers and erroneous data from the collected SPAC factor data according to a set range.

4. The high-throughput plant phenotypic dynamic monitoring system based on SPAC factors according to claim 3, characterized in that, The SPAC factor data processing submodule arranges the collected data according to the time axis, and labels them as C1, C2, ..., C... n The entire data exhibits a linear variation; if at a certain moment C... i The value compared to C i-1 C i+1 There is a significant change, namely C i ≥[(C i-1 +C i+1 ) / 2]×(1+X)%, or C i <[(C i-1 +C i+1 If ) / 2]×(1-X)%, where X is a set value, then it is considered an abnormal value, and the C i Delete, or remove C i The value is replaced with (C) i-1 +C i+1 ) / 2.

5. The high-throughput plant phenotypic dynamic monitoring system based on SPAC factors according to claim 1, characterized in that, The feature extraction module includes: The conventional phenotypic feature extraction submodule is used to extract conventional phenotypic features, including plant size features and leaf density features; The abnormal phenotypic feature extraction submodule is used to extract abnormal phenotypic features, including localized yellowing of leaves and localized curling of plants.

6. The high-throughput plant phenotypic dynamic monitoring system based on SPAC factors according to claim 1, characterized in that, The model management module includes: The correlation analysis submodule performs correlation analysis based on the extracted plant phenotypic characteristics and SPAC factor data, using Pearson correlation coefficient or Spearman correlation coefficient to calculate the correlation coefficient r between each SPAC factor and the plant phenotypic characteristics. ij , where i represents the i-th SPAC factor and j represents the j-th plant phenotypic trait; construct a correlation matrix, the elements of which are the correlation coefficients between each SPAC factor and the plant phenotypic trait. By analyzing the correlation matrix, the association between each SPAC factor and the plant phenotypic trait is preliminarily determined. The weight determination submodule determines the influence weights w of each SPAC factor on plant phenotypic traits based on the results of correlation analysis. i The weights are determined using an information entropy-based method. First, the information entropy H of each SPAC factor is calculated. i : Where, p k The probability of the i-th SPAC factor taking the k-th value; then, the weights are determined based on information entropy: The model building submodule establishes a correlation model between SPAC factors and plant phenotypes based on determined weights. This model employs a multiple linear regression model, with n SPAC factors x1, x2, ..., xn. n and m plant phenotypic characteristics y1, y2, ..., y m The model is represented as: Where, β j0 For the intercept term, β ji For regression coefficients, ∈ j The error term is used; the model parameters are estimated using the least squares method to obtain the specific model equations.

7. The high-throughput plant phenotypic dynamic monitoring system based on SPAC factors according to claim 1, characterized in that, The deduction module includes: The trend prediction submodule predicts the trend of plant phenotypic changes over a future period based on the established correlation model and real-time collected SPAC factor data. It substitutes the real-time SPAC factor data into the model equation to calculate the predicted values ​​of plant phenotypic characteristics at future times. Where t is the current time and k is the prediction time step; confidence intervals are introduced to evaluate the reliability of the prediction results; and the standard error of the prediction value is calculated. Determine the confidence interval for the predicted value: Among them, Z a / 2 denoted as the quantile of the standard normal distribution, and α is the significance level; The scenario analysis submodule simulates the changes in plant phenotypes under different SPAC factor change scenarios set by the user, based on the correlation model.

8. The high-throughput plant phenotypic dynamic monitoring system based on SPAC factors according to claim 1, characterized in that, The data evaluation module includes: The accuracy assessment submodule compares the prediction results from the inference module with the actual observed plant phenotypic data to calculate the prediction accuracy (ACC). The formula for calculating the prediction accuracy is as follows: Where m is the number of plant phenotypic traits, and N cj The number of samples that correctly predict the j-th plant phenotypic trait, N tj The total number of samples for the j-th plant phenotypic trait; The feedback adjustment submodule feeds back the accuracy evaluation index to the model management module based on the data evaluation results. The model management module then dynamically adjusts the correlation weights of each factor in the correlation model based on the feedback information. The adjustment method is either gradient descent or genetic algorithm.

9. The high-throughput plant phenotypic dynamic monitoring system based on SPAC factors according to claim 8, characterized in that, The accuracy assessment submodule further uses root mean square error (RMSE) and mean absolute error (MAE) to evaluate the accuracy of the prediction results; the formula for calculating RMSE is as follows: The formula for calculating the mean absolute error is: in, Let y be the predicted value of the j-th plant phenotypic trait on the i-th sample. ji These are actual observed values.

10. A method for high-throughput monitoring of plant phenotypic dynamics based on SPAC factors, characterized in that, The system implementation based on any one of claims 1-9 specifically includes the following steps: S1: The data acquisition module continuously acquires plant image data and SPAC factor data according to the set time interval and acquisition method, and transmits the data to the data processing module; S2: The data processing module performs preprocessing operations on the acquired image data and SPAC factor data to obtain the processed information; S3: The feature extraction module extracts the plant's conventional and abnormal phenotypic features based on the processed data and transmits the feature data to the model management module. S4: The model management module performs correlation analysis, weight determination, and model building based on the extracted features and SPAC factor data, generating a correlation model between SPAC factors and plant phenotypes. S5: The inference module predicts and analyzes future plant phenotypic trends based on the established correlation model and real-time collected SPAC factor data, and outputs the results. S6: The data evaluation module compares the prediction results of the inference module with the actual observation data, evaluates the accuracy of the prediction, and feeds back the evaluation results to the model management module. S7: The model management module dynamically adjusts model parameters and weights based on feedback information to continuously optimize model performance.

Images