Soybean growth monitoring method based on satellite remote sensing

Abstract

The invention provides a soybean growth monitoring method based on satellite remote sensing, and relates to the technical field of crop growth monitoring, and the method specifically comprises the steps: setting two monitoring nodes according to a soybean growth period, and collecting multispectral and thermal infrared remote sensing images and normal growth leaf area density and stem diameter sample data; a soybean area and a non-soybean area are distinguished by using a vegetation index threshold, cloud / shadow pixels are eliminated by using a multiband combination-threshold method, effective pixels are screened, and two types of remote sensing data are matched. Extracting leaf area density, canopy temperature and collaborative inversion stem diameter from the effective data; calculating the relative growth rate of the two types of indexes based on samples, and constructing a collaborative growth reference equation and a residual standard deviation; and substituting the growth rate of the planting area into an equation to obtain a theoretical value, calculating a deviation value, and judging the soybean growth state by combining the comparison between the deviation and a normal fluctuation interval and the positive and negative characteristics of the growth rate.

Description

Technical Field

[0001] This invention relates to the field of crop growth monitoring technology, specifically a method for monitoring soybean growth based on satellite remote sensing. Background Technology

[0002] Soybeans, as a globally important dual-purpose crop for both grain and oil, are crucial to food security and the stability of the agricultural product market. Traditional soybean growth monitoring relies on manual field observation, which consumes significant manpower and resources. Its intelligent acquisition of location information cannot comprehensively reflect the growth differences across large areas of farmland, and is highly subjective and inefficient. Therefore, developing a technology for monitoring soybean growth based on satellite remote sensing is particularly important.

[0003] With the rise of remote sensing technology, various types of remote sensing satellites have been gradually improved, and the performance of remote sensing data has been continuously optimized, leading to its increasing application in the agricultural field. Simultaneously, advancements in data processing and modeling technologies have provided computational support and reliable model support for remote sensing satellite monitoring of crop growth. Satellite remote sensing has become an effective method for monitoring soybean growth.

[0004] In existing technologies, CN109187356 discloses a method for predicting soybean growth based on satellite remote sensing data. The specific method includes: Remote sensing data acquisition: In a designated soybean planting area, data (including CCD, HSI, and IRS data) is acquired using remote sensing satellites at different spatial locations, altitudes, and resolutions, combined with the characteristics of different satellite bands; simultaneously, remote sensing images are acquired according to the soybean growth stages, such as emergence and seedling stages, and the images are divided into several equal parts according to the number of soybeans. Data preprocessing: This includes four key operations: First, image mosaicking, stitching images together based on a reference image and matching brightness and contrast; second, using ENVI 4.7 software to perform geometric correction on CCD and IRS data; third, converting DN values ​​to radiance using a radiometric calibration formula to eliminate sensor noise; and fourth, using ENVI's FLAASH module for atmospheric correction, completing data unit and storage format conversion, and setting relevant parameters. Growth information extraction and analysis: First, the image is converted to grayscale, the pixel percentage of each leaf segment is calculated, a growth curve is plotted, and the result is compared with a preset threshold. Then, the RGB values ​​of each pixel in the image are obtained, the mean of each segment is calculated, a curve is plotted again, and the color threshold is checked. Finally, the image and growth information can be uploaded to the server for staff to view and judge through a client. However, this solution only considers the area percentage and color of the leaves, which cannot reflect the overall growth status of the plant, and it does not consider environmental data, thus its support for actual planting management is limited.

[0005] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0006] The purpose of this invention is to provide a method for monitoring soybean growth based on satellite remote sensing, so as to solve the problems mentioned in the background art.

[0007] To achieve the above objectives, the present invention provides the following technical solution:

[0008] A method for monitoring soybean growth based on satellite remote sensing, comprising the following steps:

[0009] The first and second monitoring nodes were determined based on the soybean growth period. Remote sensing image data and sample data were collected in the planting area at the first and second monitoring nodes, respectively. The remote sensing image data included multispectral image data and thermal infrared image data. The sample data included leaf area density sample data and stem diameter sample data of normal growth at the two monitoring nodes.

[0010] Furthermore, the initial monitoring time point was determined based on the soybean growth cycle, which includes the sowing period, branching period, flowering and pod-setting period, and grain-filling period. The branching period was selected as the initial monitoring time point, and the second monitoring time point was 30 days after the first. The spectral resolution of the multispectral data... Spatial resolution Temperature resolution of thermal infrared data Spatial resolution This ensures that the data covers soybean growing areas.

[0011] Based on multispectral image data, the vegetation index within each pixel is calculated. The soybean planting area is distinguished from the non-soybean planting area by the vegetation index threshold method. Cloud and shadow pixels are removed by the multi-band combination-threshold method. Effective pixels of soybean planting plots in the planting area are selected and mapped to thermal infrared image data. The effective pixel data is retained in the multispectral image data and thermal infrared image data to obtain effective multispectral image data and effective thermal infrared image data.

[0012] Furthermore, the effective pixel selection method is as follows: after selecting non-soybean planting areas based on the normalized vegetation index and the enhanced vegetation index, cloud and shadow pixels are removed to obtain effective pixels;

[0013] The method for screening non-soybean-growing areas is as follows: The normalized vegetation index is calculated based on the near-infrared and red band reflectance in multispectral image data.

[0014] ;

[0015] in, For near-infrared reflectivity, For red band reflectance, the normalized difference in vegetation index (NDPI) during the soybean growth period is typically between 0.3 and 0.8. Using this as a threshold, when... When a pixel is identified as a non-vegetation pixel, it is marked as invalid. At that time, it was initially identified as a vegetation pixel and proceeded to the next screening step;

[0016] Enhanced vegetation indices are calculated based on near-infrared, red, and blue reflectance from multispectral image data. These enhanced vegetation indices are then supplemented by areas with high vegetation index coverage for further screening.

[0017] ;

[0018] in, For near-infrared reflectivity, For red band reflectivity, For blue band reflectance, the enhanced vegetation index for soybeans during the growing season is between 0.2 and 0.6. If the normalized vegetation index (NDI) of a pixel is within the threshold range but the enhanced vegetation index exceeds the threshold, it is marked as invalid. If both the NDI and the enhanced vegetation index are within the threshold range, the pixel is considered valid. The method for removing cloud and shadow pixels is as follows: radial cloud and shadow removal is performed on pixels within the soybean planting area using a multi-band combined threshold.

[0019] Cloud pixel removal method: Cloud pixels are identified and removed by combining the blue band reflectance and the normalized vegetation index threshold in multispectral image data. If the blue band reflectance of the pixel is low, the cloud pixel is removed. ,and If so, the pixel is identified as a cloud pixel and removed;

[0020] Shadow pixel removal method: Shadow pixels are removed based on red band reflectance and near-infrared reflectance. If the red band reflectance of the pixel is low... Near-infrared reflectivity ,and If the pixel is identified as a shadow pixel, it will be removed.

[0021] Within the planting area, soybean leaf area density was extracted based on vegetation index, soybean canopy temperature was extracted from thermal infrared image data, and soybean stem diameter was calculated by combining multispectral image data.

[0022] Furthermore, the red-edge normalized vegetation index was calculated based on the red-edge reflectance and the near-infrared reflectance:

[0023] ;

[0024] in, The red-edged normalized vegetation index, For near-infrared reflectivity, Reflectivity in the red-edge band;

[0025] The normalized vegetation index, the red-edge normalized vegetation index, and the enhanced vegetation index were input into a multiple linear regression model to obtain leaf area density:

[0026] ;

[0027] in, Leaf area density, Normalized Difference Vegetation Index (NDVI) The red-edged normalized vegetation index, To enhance the vegetation index, , The fitting coefficients are calculated using the least squares method.

[0028] The method for calculating soybean stem diameter is as follows: Near-infrared surface reflectance is extracted from multispectral data, and soybean canopy temperature is extracted from thermal infrared data. A nonlinear coupling model is constructed based on the positive correlation between near-infrared surface reflectance and stem diameter, and the negative correlation between canopy temperature and stem diameter.

[0029] ;

[0030] in, The stem is thick. These are the model coefficients. For near-infrared reflectivity, Canopy temperature;

[0031] The model was trained by using sample soybean stem diameter and corresponding near-infrared reflectance and canopy temperature. The model coefficients were optimized, and the soybean stem diameter was obtained by inverting the input of near-infrared reflectance and canopy temperature into the trained model.

[0032] Based on normal growth sample data within the planting area, the relative growth rate of leaf area density was calculated. Relative growth rate of stem diameter A benchmark equation for leaf area density-stem diameter synergistic growth was constructed using the synergistic growth parameter. The standard deviation of the residuals for all samples was calculated based on the benchmark equation. ;

[0033] Furthermore, based on the leaf area density of the two collected samples... Sample stem diameter Calculate the relative growth rate of leaf area density Relative growth rate of stem diameter :

[0034] ;

[0035] in, This represents the relative growth rate of leaf area density. The leaf area density of the first monitoring node sample. The leaf area density is the second detection node.

[0036] ;

[0037] in, The relative growth rate of stem diameter. The stem diameter is for the first monitoring node sample. The stem diameter is for the second monitoring node sample.

[0038] by As the independent variable, Using linear fitting as the dependent variable, a baseline equation for coordinated growth under normal growth conditions is constructed:

[0039] ;

[0040] Based on the baseline equation, calculate the standard deviation of the residuals for all normal samples. Sample standard deviation The 95% confidence interval is used as the normal fluctuation range for soybean growth monitoring. The calculated relative growth rates of leaf area density and stem diameter within the planting area are input into the baseline equation to obtain the theoretical relative growth rate of stem diameter. :

[0041] ;

[0042] Calculate the synergistic growth deviation value .

[0043] Calculate the relative growth rate of leaf area density within the planting area. Relative growth rate of stem diameter Substituting these values ​​into the baseline equation yields the theoretical relative growth rate of stem diameter. The co-growth deviation value was calculated based on the relative growth rate of the sample stem diameter and the relative growth rate of the theoretical stem diameter. The growth status of soybeans in the planting area is determined by comparing the deviation value with the normal fluctuation range and combining the positive and negative values ​​of the relative growth rate of area density and the relative growth rate of stem diameter.

[0044] Furthermore, based on the deviation value By comparing the growth rate with the normal fluctuation range, and combining the positive and negative values ​​of the relative growth rates of leaf area density and stem diameter, the growth status of soybeans can be determined. ,at the same time and If the soybean growth is normal, then the soybean growth is considered to be normal; if ,at the same time If the soybean growth is determined to be abnormal, it is classified as excessive vegetative growth; if ,at the same time If the soybean growth is abnormal due to nutritional imbalance, then the soybean growth is determined to be abnormal. or If so, the soybean growth is determined to be of a declining abnormal type.

[0045] This invention also provides a soybean growth monitoring system based on satellite remote sensing, the system being used to implement the aforementioned soybean growth monitoring method based on satellite remote sensing, comprising:

[0046] Data acquisition module: used to determine the first and second monitoring nodes according to the soybean growth stage, and to collect remote sensing image data and sample data in the planting area at the first and second monitoring nodes respectively. The remote sensing image data includes multispectral image data and thermal infrared image data, and the sample data includes leaf area density sample data and stem diameter sample data of normal growth at the two monitoring nodes.

[0047] Effective pixel filtering module: Based on multispectral image data, it calculates the normalized vegetation index and enhanced vegetation index within each effective pixel, distinguishes soybean vegetation from non-soybean planting areas through the vegetation index threshold method, removes cloud and shadow pixels through multi-band combination-threshold method, filters effective pixels of soybean planting plots within the planting area, and maps them to thermal infrared image data. It retains the data of effective pixels in multispectral image data and thermal infrared image data to obtain effective multispectral image data and effective thermal infrared image data.

[0048] Soybean growth data calculation module: used to extract soybean leaf area density based on vegetation index within the planting area, extract soybean canopy temperature from thermal infrared image data and invert and calculate soybean stem diameter by combining multispectral image data;

[0049] Growth range calculation module: used to calculate the relative growth rate of leaf area density based on normal growth sample data within the planting area. Relative growth rate of stem diameter A benchmark equation for leaf area density-stem diameter synergistic growth was constructed using the synergistic growth parameter. The standard deviation of the residuals for all samples was calculated based on the benchmark equation. This is considered a normal fluctuation range;

[0050] Growth determination module: used to calculate the relative growth rate of leaf area density within the planting area. Relative growth rate of stem diameter Substituting these values ​​into the baseline equation yields the theoretical relative growth rate of stem diameter. The co-growth deviation value was calculated based on the relative growth rate of the sample stem diameter and the relative growth rate of the theoretical stem diameter. The growth status of soybeans in the planting area is determined by comparing the deviation value with the normal fluctuation range and combining the positive and negative values ​​of the relative growth rate of area density and the relative growth rate of stem diameter.

[0051] Compared with existing technologies, the beneficial effects of this invention are as follows: This scheme determines dual monitoring nodes based on the soybean growth stage, and selectively collects multispectral and thermal infrared remote sensing data for key growth stages, as well as leaf area density and stem diameter sample data for normal growth. Compared with traditional random time monitoring or single data type collection, this design can accurately capture the core dynamic characteristics of soybean growth, avoiding the problems of insufficient data timeliness or one-sided information, and providing high-quality basic data that fits the physiological laws of the plant for subsequent growth analysis. At the same time, this scheme breaks through the limitation of existing technologies that only focus on leaf-related indicators. It extracts leaf area density (reflecting photosynthetic capacity) through vegetation indices and inverts stem diameter (reflecting plant support and nutrient accumulation capacity) by combining thermal infrared canopy temperature and multispectral data, achieving the simultaneous acquisition of dual core growth parameters of "leaf-stem". The two types of parameters correspond to the key dimensions of soybean vegetative growth, and the synergistic inversion can more comprehensively and three-dimensionally depict soybean growth, avoiding the defect that a single parameter cannot reflect the overall growth status of the plant. Attached Figure Description

[0052] Figure 1 This is a schematic diagram of the overall method flow of the present invention;

[0053] Figure 2 This is a graph showing the relative growth rate of leaf area density, the relative growth rate of stem diameter, and the synergistic growth deviation value of the present invention.

[0054] Figure 3 This is a structural block diagram of the system of the present invention. Detailed Implementation

[0055] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0056] It should be noted that, unless otherwise defined, the technical or scientific terms used in this invention should have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0057] Example:

[0058] Please see Figures 1 to 2 The present invention provides a technical solution:

[0059] A method for monitoring soybean growth based on satellite remote sensing, specifically including:

[0060] Step 1: Determine the first and second monitoring nodes based on the soybean growth cycle. Collect remote sensing image data and sample data within the planting area at both nodes. The remote sensing image data includes multispectral image data and thermal infrared image data. The sample data includes leaf area density and stem diameter samples of normal growth at both monitoring nodes. Determine the first monitoring time node based on the soybean growth cycle, which includes sowing, branching, flowering and pod-setting, and grain-filling stages. Select the branching stage as the first monitoring time node. The second monitoring time node is 30 days after the first. The spectral resolution of the multispectral data... Spatial resolution Temperature resolution of thermal infrared data Spatial resolution This ensures that the data covers soybean growing areas.

[0061] Significant differences exist in the growth characteristics and physiological indicators of soybeans at different growth stages. Dividing soybean growth into first and second monitoring nodes based on growth stage allows for precise identification of key growth phases, avoiding the randomness of monitoring time and ensuring that the collected data reflects the true state of soybeans at their core growth stages. This provides targeted foundational data for subsequent growth assessment and stress diagnosis. The normal growth sample data from the two monitoring nodes form a time-series data sequence, directly reflecting the growth rate of soybeans from the first node to the second. This provides direct data evidence for determining whether soybean growth is normal and whether there are problems such as growth retardation or excessive vegetative growth, while also providing temporal characteristic support for subsequent growth trend prediction. The sample data are explicitly limited to leaf area density and stem diameter under normal growth conditions, enabling the construction of a benchmark database of normal soybean growth status. If subsequent diagnosis of the impact of drought, pests, and other stresses on soybean growth is required, remote sensing inversion data and ground monitoring data from the stressed area can be compared with this benchmark data to quickly quantify the degree of stress, improving the accuracy and efficiency of stress diagnosis.

[0062] Step 2: Based on multispectral image data, calculate the vegetation index within each pixel. Using a vegetation index thresholding method, distinguish between soybean planting areas and non-soybean planting areas. Remove cloud and shadow pixels using a multi-band combination-thresholding method to filter effective pixels from soybean planting areas within the planting region. Map these effective pixels to thermal infrared image data. Retain the effective pixel data from both the multispectral and thermal infrared image data to obtain effective multispectral and thermal infrared image data. The method for filtering effective pixels is as follows: after filtering out non-soybean planting areas based on the normalized vegetation index and enhanced vegetation index, remove cloud and shadow pixels to obtain effective pixels.

[0063] The method for screening non-soybean-growing areas is as follows: The normalized vegetation index is calculated based on the near-infrared and red band reflectance in multispectral image data.

[0064] ;

[0065] in, For near-infrared reflectivity, For red band reflectance, the normalized difference in vegetation index (NDPI) during the soybean growth period is typically between 0.3 and 0.8. Using this as a threshold, when... When a pixel is identified as a non-vegetation pixel, it is marked as invalid. At that time, it was initially identified as a vegetation pixel and proceeded to the next screening step;

[0066] Enhanced vegetation indices are calculated based on near-infrared, red, and blue reflectance from multispectral image data. These enhanced vegetation indices are then supplemented by areas with high vegetation index coverage for further screening.

[0067] ;

[0068] in, For near-infrared reflectivity, For red band reflectivity, For blue band reflectance, the enhanced vegetation index for soybeans during the growing season is between 0.2 and 0.6. If the normalized vegetation index (NDI) of a pixel is within the threshold range but the enhanced vegetation index exceeds the threshold, it is marked as invalid. If both the NDI and the enhanced vegetation index are within the threshold range, the pixel is considered valid. The method for removing cloud and shadow pixels is as follows: radial cloud and shadow removal is performed on pixels within the soybean planting area using a multi-band combined threshold.

[0069] Cloud pixel removal method: Cloud pixels are identified and removed by combining the blue band reflectance and the normalized vegetation index threshold in multispectral image data. If the blue band reflectance of the pixel is low, the cloud pixel is removed. ,and If so, the pixel is identified as a cloud pixel and removed;

[0070] Shadow pixel removal method: Shadow pixels are removed based on red band reflectance and near-infrared reflectance. If the red band reflectance of the pixel is low... Near-infrared reflectivity ,and If the pixel is identified as a shadow pixel, it will be removed.

[0071] The Normalized Difference Vegetation Index (NDI) is sensitive to vegetation cover and can quickly identify green vegetation areas. The Enhanced Vegetation Index (EDI) optimizes the NDI's tendency to saturate in high-vegetation-coverage areas, providing a more accurate reflection of soybean canopy growth vitality. Combining these two indicators creates a dual-criteria system, effectively eliminating interference from bare soil, weeds, and crop residues compared to a single vegetation index. This accurately delineates the boundaries of soybean planting areas, laying a precise spatial foundation for subsequent targeted monitoring. Clouds and shadows can obscure soybean canopy information, leading to invalid or erroneous remote sensing data. A multi-band combination-threshold method utilizes the reflectance differences of different bands for comprehensive discrimination, thoroughly eliminating cloud-covered pixels, shadow pixels, and sensor noise pixels. This reduces interference from invalid data in subsequent analysis, ensuring the authenticity and accuracy of the filtered valid pixel data. The process involves first distinguishing between soybean and non-soybean areas and then removing clouds and shadows, rather than directly denoising the entire area. This reduces the scope of invalid pixel removal, decreases computational resource consumption, and improves the efficiency of the screening process. At the same time, it avoids misjudging clouds and shadows in non-soybean growing areas as invalid data from soybean areas, ensuring the relevance and rationality of the screening results.

[0072] Step 3: Within the planting area, extract soybean leaf area density based on vegetation indices; extract soybean canopy temperature from thermal infrared image data and calculate soybean stem diameter by combining it with multispectral image data; calculate the red-edge normalized vegetation index based on red-edge reflectance and near-infrared reflectance.

[0073] ;

[0074] in, The red-edged normalized vegetation index, For near-infrared reflectivity, Reflectivity in the red-edge band;

[0075] The normalized vegetation index, the red-edge normalized vegetation index, and the enhanced vegetation index were input into a multiple linear regression model to obtain leaf area density:

[0076] ;

[0077] in, Leaf area density, Normalized Difference Vegetation Index (NDVI) The red-edged normalized vegetation index, To enhance the vegetation index, , The fitting coefficients are calculated using the least squares method.

[0078] The method for calculating soybean stem diameter is as follows: Near-infrared surface reflectance is extracted from multispectral data, and soybean canopy temperature is extracted from thermal infrared data. A nonlinear coupling model is constructed based on the positive correlation between near-infrared surface reflectance and stem diameter, and the negative correlation between canopy temperature and stem diameter.

[0079] ;

[0080] in, The stem is thick. These are the model coefficients. For near-infrared reflectivity, Canopy temperature;

[0081] The model was trained by using sample soybean stem diameter and corresponding near-infrared reflectance and canopy temperature. The model coefficients were optimized, and the soybean stem diameter was obtained by inverting the input of near-infrared reflectance and canopy temperature into the trained model.

[0082] Normalized Difference Vegetation Index (NDVI), Enhanced Vegetation Index (EDI), and soybean leaf area density (LAD) show a significant positive correlation. Based on the selected effective multispectral pixel data, the leaf area density parameter of the soybean canopy can be directly obtained through a correlation model between vegetation index and leaf area density. Compared with traditional manual sampling and measurement methods, this approach does not require crop damage and can achieve large-scale, full-coverage quantitative monitoring, significantly improving monitoring efficiency and spatial representativeness. Thermal infrared image data can capture the radiation temperature of the soybean canopy. The canopy temperature extracted based on effective thermal infrared pixels is a direct indicator of the soybean's water metabolism status—an abnormally high canopy temperature usually corresponds to weakened plant transpiration and intensified water stress. This step, through precise extraction of canopy temperature, provides crucial quantitative evidence for subsequent assessment of soybean water supply and demand balance and identification of drought-stressed areas.

[0083] Step 4: Calculate the relative growth rate of leaf area density based on normal growth sample data within the planting area. Relative growth rate of stem diameter A benchmark equation for leaf area density-stem diameter synergistic growth was constructed using the synergistic growth parameter. The standard deviation of the residuals for all samples was calculated based on the benchmark equation. Based on the leaf area density of the two collected samples Sample stem diameter Calculate the relative growth rate of leaf area density Relative growth rate of stem diameter :

[0084] ;

[0085] in, This represents the relative growth rate of leaf area density. The leaf area density of the first monitoring node sample. The leaf area density is the second detection node.

[0086] ;

[0087] in, The relative growth rate of stem diameter. The stem diameter is for the first monitoring node sample. The stem diameter is for the second monitoring node sample.

[0088] by As the independent variable, Using linear fitting as the dependent variable, a baseline equation for coordinated growth under normal growth conditions is constructed:

[0089] ;

[0090] Based on the baseline equation, calculate the standard deviation of the residuals for all normal samples. Sample standard deviation The 95% confidence interval is used as the normal fluctuation range for soybean growth monitoring. The calculated relative growth rates of leaf area density and stem diameter within the planting area are input into the baseline equation to obtain the theoretical relative growth rate of stem diameter. :

[0091] ;

[0092] Calculate the synergistic growth deviation value .

[0093] Leaf area density and stem diameter are core indicators of soybean vegetative growth, and they exhibit a synergistic physiological relationship. By calculating the relative growth rate of two nodes, a synergistic growth benchmark equation is constructed, transforming the qualitative growth relationship between "leaf-stem" under normal soybean growth into a quantitative mathematical model. This clarifies the optimal matching relationship between the two indicators under normal growth conditions, providing a referable "standard scale" for subsequent abnormal growth identification. Compared to directly using the absolute values ​​of leaf area density and stem diameter to construct the model, using the relative growth rate as a synergistic parameter effectively eliminates the influence of differences in initial soybean seedling conditions across different plots, enabling the benchmark equation to adapt to soybean plots with different initial growth states, thus improving the universality and applicability of the benchmark model. The sample residuals reflect the degree of deviation between the normal growth sample data and the benchmark equation, and the standard deviation of the residuals further quantifies the natural fluctuation range of the "leaf-stem" synergistic relationship under normal growth conditions. This standard deviation can serve as the core threshold for subsequent anomaly identification, transforming the vague concept of abnormal growth into a quantifiable mathematical discrimination standard, significantly improving the objectivity and accuracy of growth status assessment. The baseline equation and standard deviation threshold output by this step can serve as the core basis for soybean growth grading, providing precise decision support for differentiated field management and avoiding resource waste caused by blind management.

[0094] Step 5: Calculate the relative growth rate of leaf area density within the planting area. Relative growth rate of stem diameter Substituting these values ​​into the baseline equation yields the theoretical relative growth rate of stem diameter. The co-growth deviation value was calculated based on the relative growth rate of the sample stem diameter and the relative growth rate of the theoretical stem diameter. The growth status of soybeans in the planting area is determined by comparing the deviation value with the normal fluctuation range and combining the positive and negative values ​​of the relative growth rate of area density and the relative growth rate of stem diameter.

[0095] Based on deviation value By comparing the growth rate with the normal fluctuation range, and combining the positive and negative values ​​of the relative growth rates of leaf area density and stem diameter, the growth status of soybeans can be determined. ,at the same time and If the soybean growth is normal, then the soybean growth is considered to be normal; if ,at the same time If the soybean growth is determined to be abnormal, it is classified as excessive vegetative growth; if ,at the same time If the soybean growth is abnormal due to nutritional imbalance, then the soybean growth is determined to be abnormal. or If so, the soybean growth is determined to be of a declining abnormal type.

[0096] Based on the effective image data selected in the early stage, this step can perform batch calculations and classifications for all soybean planting plots within the planting area, eliminating the need for manual sampling and investigation of each plot. Compared with traditional field inspection methods, this significantly shortens the classification cycle and improves the timeliness of large-scale regional growth monitoring, making it particularly suitable for comprehensive growth assessment of large-scale soybean planting bases. The accurate growth classification results can directly guide differentiated management measures: targeted spraying of growth regulators for overly vigorous plots, topdressing with nitrogen fertilizer for plots with insufficient leaf area, and investigating pests, diseases, or water stress in plots with declining growth. This targeted management approach avoids the traditional one-size-fits-all fertilization and pesticide application model, reduces agricultural input waste, lowers production costs, and simultaneously improves soybean yield and quality.

[0097] As shown in Table 1, the data in Table 1 reflect the relative growth rate of leaf area density, the relative growth rate of stem diameter, the co-growth deviation value, and the soybean growth status in some data groups.

[0098] Table 1 shows the relative growth rate of leaf area density, relative growth rate of stem diameter, co-growth deviation value, and soybean growth status for some data sets.

[0099] Group number Relative growth rate of leaf area density Relative growth rate of stem diameter Cooperative growth deviation value growth status 1 28.5 15.2 0.82 normal 2 25.3 13.8 -0.65 normal 3 42.7 8.3 -2.91 Eccentric growth abnormality 4 18.6 25.9 3.02 Nutritional Imbalance Abnormal 5 22.1 12.5 -0.37 normal 6 -5.2 3.1 -1.89 Degenerative anomaly 7 31.4 16.7 0.59 normal 8 45.9 9.1 -3.24 Eccentric growth abnormality 9 15.8 22.4 2.78 Nutritional Imbalance Abnormal 10 -3.7 -2.8 -1.56 Degenerative anomaly 11 26.9 14.3 -0.71 normal 12 35.2 18.5 0.43 normal 13 51.3 7.8 -3.57 Eccentric growth abnormality 14 12.4 28.6 3.15 Nutritional Imbalance Abnormal 15 -7.9 1.2 -2.13 Degenerative anomaly 16 29.7 15.9 0.68 normal 17 38.5 10.2 -2.89 Eccentric growth abnormality 18 19.3 24.1 2.63 Nutritional Imbalance Abnormal 19 -2.4 -4.5 -1.72 Degenerative anomaly 20 24.8 13.1 -0.52 normal

[0100] Table 1 shows the accuracy of the judgment method based on the relative growth rates of leaf area density, stem diameter, synergistic growth deviation, and soybean growth status for each group. For example, in Group 1, the relative growth rates of leaf area density and stem diameter are positive, the synergistic growth deviation is within the range, and the soybean growth status is normal; in Group 3, the data shows... ,at the same time The soybeans showed abnormal, excessive growth; in the four sets of data... ,at the same time The soybean growth was abnormal due to nutritional imbalance; if in the 6 sets of data... The fact that the soybean growth was abnormally declining indicates that the soybean growth monitoring method was reasonable.

[0101] Please see Figure 3 The present invention further provides a soybean growth monitoring system based on satellite remote sensing, the system being used to implement the aforementioned soybean growth monitoring method based on satellite remote sensing, comprising:

[0102] Data acquisition module: used to determine the first and second monitoring nodes according to the soybean growth stage, and to collect remote sensing image data and sample data in the planting area at the first and second monitoring nodes respectively. The remote sensing image data includes multispectral image data and thermal infrared image data, and the sample data includes leaf area density sample data and stem diameter sample data of normal growth at the two monitoring nodes.

[0103] Effective pixel filtering module: Based on multispectral image data, it calculates the normalized vegetation index and enhanced vegetation index within each effective pixel, distinguishes soybean vegetation from non-soybean planting areas through the vegetation index threshold method, removes cloud and shadow pixels through multi-band combination-threshold method, filters effective pixels of soybean planting plots within the planting area, and maps them to thermal infrared image data. It retains the data of effective pixels in multispectral image data and thermal infrared image data to obtain effective multispectral image data and effective thermal infrared image data.

[0104] Soybean growth data calculation module: used to extract soybean leaf area density based on vegetation index within the planting area, extract soybean canopy temperature from thermal infrared image data and invert and calculate soybean stem diameter by combining multispectral image data;

[0105] Growth range calculation module: used to calculate the relative growth rate of leaf area density based on normal growth sample data within the planting area. Relative growth rate of stem diameter A benchmark equation for leaf area density-stem diameter synergistic growth was constructed using the synergistic growth parameter. The standard deviation of the residuals for all samples was calculated based on the benchmark equation. This is considered a normal fluctuation range;

[0106] Growth determination module: used to calculate the relative growth rate of leaf area density within the planting area. Relative growth rate of stem diameter Substituting these values ​​into the baseline equation yields the theoretical relative growth rate of stem diameter. The co-growth deviation value was calculated based on the relative growth rate of the sample stem diameter and the relative growth rate of the theoretical stem diameter. The growth status of soybeans in the planting area is determined by comparing the deviation value with the normal fluctuation range and combining the positive and negative values ​​of the relative growth rate of area density and the relative growth rate of stem diameter.

[0107] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0108] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented in software, the above embodiments can be implemented, in whole or in part, as a computer program product. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution.

[0109] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0110] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A method for monitoring soybean growth based on satellite remote sensing, characterized in that, include: The first and second monitoring nodes were determined based on the soybean growth period. Remote sensing image data and sample data were collected in the planting area at the first and second monitoring nodes, respectively. The remote sensing image data included multispectral image data and thermal infrared image data. The sample data included leaf area density sample data and stem diameter sample data of normal growth at the two monitoring nodes. Based on multispectral image data, the vegetation index within each pixel is calculated. The soybean planting area is distinguished from the non-soybean planting area by the vegetation index threshold method. Cloud and shadow pixels are removed by the multi-band combination-threshold method. Effective pixels of soybean planting plots in the planting area are selected and mapped to thermal infrared image data. The effective pixel data is retained in the multispectral image data and thermal infrared image data to obtain effective multispectral image data and effective thermal infrared image data. Within the planting area, soybean leaf area density was extracted based on vegetation index, soybean canopy temperature was extracted from thermal infrared image data, and soybean stem diameter was calculated by combining multispectral image data. Based on normal growth sample data in the planting area, the relative growth rate of leaf area density and the relative growth rate of stem diameter were calculated as synergistic growth parameters to construct the benchmark equation for synergistic growth of leaf area density and stem diameter. The standard deviation of the residuals of all samples was calculated based on the benchmark equation. The relative growth rate of leaf area density and the relative growth rate of stem diameter within the planting area are calculated and substituted into the baseline equation to obtain the theoretical relative growth rate of stem diameter. The co-growth deviation value is calculated based on the sample relative growth rate of stem diameter and the theoretical relative growth rate of stem diameter. The growth status of soybeans within the planting area is determined by comparing the deviation value with the normal fluctuation range and by combining the positive and negative values ​​of the relative growth rate of area density and the relative growth rate of stem diameter.

2. The method for monitoring soybean growth based on satellite remote sensing according to claim 1, characterized in that: The initial monitoring time point was determined based on the soybean growth cycle, which includes the sowing period, branching period, flowering and pod-setting period, and grain-filling period. The branching period was selected as the initial monitoring time point, and the second monitoring time point was 30 days after the first. The spectral resolution of the multispectral data... Spatial resolution Temperature resolution of thermal infrared data Spatial resolution This ensures that the data covers soybean growing areas.

3. The method for monitoring soybean growth based on satellite remote sensing according to claim 1, characterized in that: The effective pixel selection method is as follows: after screening out non-soybean planting areas based on the normalized vegetation index and the enhanced vegetation index, cloud and shadow pixels are removed to obtain effective pixels. The method for screening non-soybean-growing areas is as follows: The normalized vegetation index is calculated based on the near-infrared and red band reflectance in multispectral image data. ； in, For near-infrared reflectivity, For red band reflectance, the normalized difference in vegetation index (NDPI) during the soybean growth period is typically between 0.3 and 0.

8. Using this as a threshold, when... When a pixel is identified as a non-vegetation pixel, it is marked as invalid. At that time, it was initially identified as a vegetation pixel and proceeded to the next screening step; Enhanced vegetation indices are calculated based on near-infrared, red, and blue reflectance from multispectral image data. These enhanced vegetation indices are then supplemented by areas with high vegetation index coverage for further screening. ； in, For near-infrared reflectivity, For red band reflectivity, The range of enhanced vegetation index for soybean growth period is 0.2-0.6 for blue band reflectance. If the normalized vegetation index of a pixel is within the threshold range but the enhanced vegetation index exceeds the threshold, it is marked as invalid. If both the normalized vegetation index and the enhanced vegetation index of a pixel are within the threshold range, it is judged as a valid pixel.

4. The method for monitoring soybean growth based on satellite remote sensing according to claim 3, characterized in that: The method for removing cloud and shadow pixels is as follows: radial cloud and shadow removal of pixels in the soybean planting area is performed by combining multiple band thresholds. Cloud pixel removal method: Cloud pixels are identified and removed by combining the blue band reflectance and the normalized vegetation index threshold in multispectral image data. If the blue band reflectance of the pixel is low, the cloud pixel is removed. ,and If so, the pixel is identified as a cloud pixel and removed; Shadow pixel removal method: Shadow pixels are removed based on red band reflectance and near-infrared reflectance. If the red band reflectance of the pixel is low... Near-infrared reflectivity ,and If the pixel is identified as a shadow pixel, it will be removed.

5. The method for monitoring soybean growth based on satellite remote sensing according to claim 1, characterized in that: The red-edge normalized vegetation index is calculated based on the red-edge reflectance and near-infrared reflectance: ； in, The red-edged normalized vegetation index, For near-infrared reflectivity, Reflectivity in the red-edge band; The normalized vegetation index, the red-edge normalized vegetation index, and the enhanced vegetation index were input into a multiple linear regression model to obtain leaf area density: ； in, Leaf area density, Normalized Difference Vegetation Index (NDVI) The red-edged normalized vegetation index, To enhance the vegetation index, , The fitting coefficients are calculated using the least squares method.

6. The method for monitoring soybean growth based on satellite remote sensing according to claim 1, characterized in that: The method for calculating soybean stem diameter is as follows: Near-infrared surface reflectance is extracted from multispectral data, and soybean canopy temperature is extracted from thermal infrared data. A nonlinear coupling model is constructed based on the positive correlation between near-infrared surface reflectance and stem diameter, and the negative correlation between canopy temperature and stem diameter. ； in, The stem is thick. These are the model coefficients. For near-infrared reflectivity, Canopy temperature; The model was trained by using sample soybean stem diameter and corresponding near-infrared reflectance and canopy temperature. The model coefficients were optimized, and the soybean stem diameter was obtained by inverting the input of near-infrared reflectance and canopy temperature into the trained model.

7. The method for monitoring soybean growth based on satellite remote sensing according to claim 1, characterized in that: Based on the leaf area density of the two collected samples Sample stem diameter Calculate the relative growth rate of leaf area density Relative growth rate of stem diameter : ； in, This represents the relative growth rate of leaf area density. The leaf area density of the first monitoring node sample. The leaf area density is the second detection node. ； in, The relative growth rate of stem diameter. The stem diameter is for the first monitoring node sample. The stem diameter is for the second monitoring node sample. by As the independent variable, Using linear fitting as the dependent variable, a baseline equation for coordinated growth under normal growth conditions is constructed: ； Based on the baseline equation, calculate the standard deviation of the residuals for all normal samples. Sample standard deviation The 95% confidence interval is used as the normal fluctuation range for soybean growth monitoring. The calculated relative growth rates of leaf area density and stem diameter within the planting area are input into the baseline equation to obtain the theoretical relative growth rate of stem diameter. : ； Calculate the synergistic growth deviation value .

8. The method for monitoring soybean growth based on satellite remote sensing according to claim 1, characterized in that: Based on deviation value By comparing the growth rate with the normal fluctuation range, and combining the positive and negative values ​​of the relative growth rates of leaf area density and stem diameter, the growth status of soybeans can be determined. ,at the same time and If the soybean growth is normal, then the soybean growth is considered to be normal; if ,at the same time If the soybean growth is determined to be abnormal, it is classified as excessive vegetative growth; if ,at the same time If so, the soybean growth is determined to be abnormal due to nutritional imbalance. like or If so, the soybean growth is determined to be of a declining abnormal type.

Images