Cotton water and nitrogen stress online monitoring method and system fusing multi-temporal remote sensing data

Abstract

This application relates to a method and system for online monitoring of water and nitrogen stress in cotton by fusing multi-temporal remote sensing data, belonging to the field of industrial vision technology. The method includes: acquiring temporal fluctuation characteristics of water evapotranspiration and nitrogen absorption in a target field area within a target monitoring time period based on a target remote sensing image, and obtaining joint trend data; identifying abnormal fluctuation points of water evapotranspiration and nitrogen absorption in the joint trend data, and obtaining stress type classification results for each fluctuation point; identifying water-dominant stress points and obtaining stress intensity quantification scores; obtaining a stress identification parameter set based on the stress intensity quantification scores; and performing water and nitrogen stress identification on real-time remote sensing images based on the stress identification parameter set to obtain a water and nitrogen stress distribution map. The technical solution of this application embodiment can realize online monitoring of water and nitrogen stress in cotton by fusing multi-source dynamic features, improving the spatiotemporal accuracy of stress identification and the pertinence of intervention measures.

Description

Technical Field

[0001] This application belongs to the field of industrial vision technology, specifically a method and system for online monitoring of water and nitrogen stress in cotton that integrates multi-temporal remote sensing data. Background Technology

[0002] As an important economic crop in agriculture, the supply of water and nitrogen during cotton's growth directly determines its yield and quality, which is crucial for ensuring food security and stabilizing farmers' income. Accurately monitoring cotton growth and responding promptly to water and nitrogen stress using modern technology is a key issue in improving agricultural production efficiency, and related research also has profound significance for sustainable agricultural development, resource conservation, and environmental protection.

[0003] Current cotton growth monitoring has significant shortcomings: most methods can only analyze a single time point or environmental condition, failing to comprehensively capture the dynamic changes at different growth stages of cotton, resulting in delayed judgment of water and nitrogen stress and missed opportunities for optimal intervention; moreover, when integrating observation information from multiple time periods, existing technologies do not delve into the intrinsic relationships between various growth-influencing factors, making it difficult to accurately pinpoint the root causes of stress. Summary of the Invention

[0004] To address the above issues, this application provides an online monitoring method and system for water and nitrogen stress in cotton that integrates multi-temporal remote sensing data. This system aims to solve the problem that existing technologies lack the ability to accurately identify the causes and effects of water and nitrogen stress in cotton by comprehensively analyzing the dynamic changes of factors such as water evapotranspiration and nitrogen uptake based on observations over multiple time periods.

[0005] To achieve the above objectives, the technical solution adopted in this application includes an online monitoring method for water and nitrogen stress in cotton that integrates multi-temporal remote sensing data, comprising: Based on the target remote sensing image of the target field area, obtain the temporal fluctuation characteristics of water evapotranspiration and nitrogen absorption of the target field area during the target monitoring time period. Obtain the joint trend data between the temporal fluctuation characteristics of water evapotranspiration and the temporal fluctuation characteristics of nitrogen absorption; Identify abnormal fluctuation points of water evapotranspiration and nitrogen absorption in the residual terms of the joint trend data, and obtain the stress type classification results for each of the abnormal fluctuation points of water evapotranspiration and nitrogen absorption. Based on the stress type classification results, water-dominant stress points are identified among the abnormal fluctuation points of water evapotranspiration and the abnormal fluctuation points of nitrogen absorption, and the stress degree quantification score of the water-dominant stress points is obtained. Based on the feature mapping relationship between the stress degree quantification score and the temporal fluctuation characteristics of water evapotranspiration and nitrogen absorption at the water-dominant stress point, a stress identification parameter set is obtained. Based on the stress identification parameter set, water and nitrogen stress are identified in each remote sensing spatial grid of the real-time remote sensing image to obtain a water and nitrogen stress distribution map.

[0006] Optionally, the step of acquiring the temporal fluctuation characteristics of water evapotranspiration and nitrogen uptake in the target field area during the target monitoring period based on the target remote sensing image of the target field area includes: Normalized difference vegetation index is obtained sequentially from the target remote sensing image in each acquisition time period, and it is determined whether each normalized difference vegetation index is lower than a preset water stress threshold. When there are consecutive target numbers of the normalized difference vegetation index below the preset water stress threshold, the collection time period corresponding to the normalized difference vegetation index of the consecutive target numbers is determined as the target monitoring time period. The temporal fluctuation characteristics of water evapotranspiration are obtained based on the normalized difference vegetation index within the target monitoring period. Acquire the chlorophyll fluorescence index from the remote sensing images of the target during the target monitoring period; The timing fluctuation characteristics of nitrogen absorption were obtained based on the chlorophyll fluorescence index.

[0007] Optionally, obtaining the joint trend data between the temporal fluctuation characteristics of water evapotranspiration and the temporal fluctuation characteristics of nitrogen absorption includes: Calculate the correlation strength parameter between the temporal fluctuation characteristics of water evapotranspiration and the temporal fluctuation characteristics of nitrogen absorption within each target time window; Based on the relevant intensity parameters, joint change time intervals with a joint change relationship between the temporal fluctuation characteristics of water evapotranspiration and the temporal fluctuation characteristics of nitrogen absorption are identified in each of the target time windows; Based on the temporal fluctuation characteristics of water evapotranspiration and the temporal fluctuation characteristics of nitrogen absorption, the rates of change of water evapotranspiration and nitrogen absorption in the joint change time interval are calculated respectively. The combined trend data is generated based on the rate of change in water evaporation and the rate of change in nitrogen absorption.

[0008] Optionally, the step of identifying anomalous fluctuation points in water evapotranspiration and nitrogen uptake in the residuals of the joint trend data, and obtaining stress type classification results for each of the anomalous fluctuation points in water evapotranspiration and nitrogen uptake, includes: Based on the time series decomposition algorithm, the trend term and the residual term of the joint change trend data are obtained; Based on the preset fluctuation threshold, the abnormal fluctuation points of moisture evaporation and nitrogen absorption are identified in the residual term. Calculate the phase difference data between the temporal fluctuation characteristics of water evapotranspiration and the temporal fluctuation characteristics of nitrogen absorption at each of the abnormal fluctuation points of water evapotranspiration and the abnormal fluctuation points of nitrogen absorption; Based on the phase difference data, cluster analysis is performed on the abnormal fluctuation points of water evapotranspiration and the abnormal fluctuation points of nitrogen absorption to obtain the stress type classification results.

[0009] Optionally, the step of identifying water-dominant stress points among the abnormal fluctuation points of water evapotranspiration and the abnormal fluctuation points of nitrogen absorption based on the stress type classification results, and obtaining a stress severity quantification score for the water-dominant stress points, includes: Obtain the dominant factor subset of the classification results for each of the aforementioned stress types; The abnormal fluctuation points of water evapotranspiration and / or the abnormal fluctuation points of nitrogen absorption that are higher than the preset factor threshold in the subset of dominant factors are identified as water-dominant stress points. Obtain the historical remote sensing baseline data corresponding to the water-dominant stress points; The stress severity quantification score is calculated based on the historical remote sensing baseline data and the temporal fluctuation characteristics of water evapotranspiration at the water-dominant stress point.

[0010] Optionally, the step of obtaining a stress identification parameter set based on the feature mapping relationship between the stress degree quantification score and the temporal fluctuation characteristics of water evapotranspiration and nitrogen absorption at the water-dominant stress point includes: Based on the random forest algorithm, a mapping relationship is modeled between the stress degree quantification score and the rate of change of water evapotranspiration, the rate of change of nitrogen absorption, and the phase difference data to obtain the weight coefficient of water evapotranspiration and the weight coefficient of nitrogen absorption. The water stress threshold parameter and nitrogen stress threshold parameter are obtained based on the water evapotranspiration weight coefficient and the nitrogen absorption weight coefficient, and the water stress threshold parameter, the nitrogen absorption weight coefficient, the water stress threshold parameter and the nitrogen stress threshold parameter are determined as the stress identification parameter set.

[0011] Optionally, the step of identifying water and nitrogen stress in each remote sensing spatial grid of the real-time remote sensing image based on the stress identification parameter set to obtain a water and nitrogen stress distribution map includes: Obtain the index feature deviation between the remote sensing image inversion index of each of the remote sensing spatial grids and the stress identification parameter set; The stress identification result of each remote sensing spatial grid is determined based on the deviation of the index characteristics; The water and nitrogen stress distribution map is generated based on the stress identification results.

[0012] Optionally, determining the stress identification result of each remote sensing spatial grid based on the index feature deviation includes: Based on the characteristic deviation of the index, if the preliminary estimate of water evapotranspiration of the remote sensing image inversion index is lower than the water stress threshold parameter, and the numerical relationship between the chlorophyll fluorescence index and the nitrogen stress threshold parameter meets the preset stability conditions, the stress identification result is determined to be nitrogen-dominated stress.

[0013] Optionally, after generating the water nitrogen stress distribution map based on each of the stress identification results, the method further includes: Identify the coordinates of the target risk area in the water and nitrogen stress distribution map; Based on the remote sensing image inversion index corresponding to the coordinates of the risk area, a temporal stress intervention signal for the coordinates of the target risk area is generated.

[0014] On the other hand, the technical solution adopted in this application also includes an online monitoring system for water and nitrogen stress in cotton that integrates multi-temporal remote sensing data, comprising: The feature acquisition module is used to acquire the temporal fluctuation characteristics of water evapotranspiration and nitrogen absorption of the target field area within the target monitoring time period based on the target remote sensing image of the target field area. The joint trend module is used to obtain joint trend data between the temporal fluctuation characteristics of water evapotranspiration and the temporal fluctuation characteristics of nitrogen absorption. The stress classification module is used to identify abnormal fluctuation points of water evapotranspiration and abnormal fluctuation points of nitrogen absorption in the residual terms of the joint trend data, and to obtain the stress type classification results of each abnormal fluctuation point of water evapotranspiration and abnormal fluctuation point of nitrogen absorption. The stress severity quantification module is used to identify water-dominant stress points among the abnormal fluctuation points of water evapotranspiration and the abnormal fluctuation points of nitrogen absorption based on the stress type classification results, and to obtain the stress severity quantification score of the water-dominant stress points. The parameter acquisition module is used to acquire a set of stress identification parameters based on the feature mapping relationship between the stress degree quantification score and the temporal fluctuation characteristics of water evapotranspiration and nitrogen absorption at the water-dominant stress point. The stress identification module is used to identify water and nitrogen stress in each remote sensing spatial grid of the real-time remote sensing image based on the stress identification parameter set, and obtain a water and nitrogen stress distribution map.

[0015] This application provides an online monitoring method and system for water and nitrogen stress in cotton, which integrates multi-temporal remote sensing data. Based on a target remote sensing image of a target field area, it acquires the temporal fluctuation characteristics of water evapotranspiration and nitrogen uptake within a target monitoring time period. It also acquires joint trend data between these two characteristics. Abnormal fluctuation points in water evapotranspiration and nitrogen uptake are identified in the residuals of this joint trend data. The stress type classification results for each abnormal fluctuation point are obtained. Based on these classification results, water-dominant stress points are identified among the abnormal fluctuation points in water evapotranspiration and nitrogen uptake. Finally, the system acquires water... The stress intensity quantification score of the dominant stress point is used to obtain a stress identification parameter set based on the feature mapping relationship between the stress intensity quantification score and the temporal fluctuation characteristics of water evapotranspiration and nitrogen absorption at the water-dominant stress point. Then, based on the stress identification parameter set, water and nitrogen stress is identified in each remote sensing spatial grid of the real-time remote sensing image, resulting in a water and nitrogen stress distribution map. This solves the problem of existing technologies lacking the ability to accurately identify the causes and effects of cotton water and nitrogen stress by comprehensively analyzing the dynamic changes of factors such as water evapotranspiration and nitrogen absorption based on observations over multiple time periods. It achieves online monitoring of cotton water and nitrogen stress by integrating multi-source dynamic features, significantly improving the spatiotemporal accuracy of stress identification and the targeting of intervention measures, providing a scientific basis for precision agricultural management of cotton. Attached Figure Description

[0016] Figure 1 A flowchart illustrating an online monitoring method for water and nitrogen stress in cotton that integrates multi-temporal remote sensing data, provided in this application embodiment; Figure 2 This is a schematic diagram of the structure of an online monitoring system for water and nitrogen stress in cotton that integrates multi-temporal remote sensing data, provided in an embodiment of this application. Detailed Implementation

[0017] To enable those skilled in the art to better understand the technical solution, the present application will be described in detail below with reference to the embodiments. The description in this section is only exemplary and explanatory, and should not be used to limit the scope of protection of the present application in any way.

[0018] Figure 1 The flowchart illustrates an online monitoring method for water and nitrogen stress in cotton that integrates multi-temporal remote sensing data, provided in this application embodiment. This method can be executed by an online monitoring system for water and nitrogen stress in cotton that integrates multi-temporal remote sensing data. This system can be implemented in hardware and / or software and is generally integrated into an electronic device, such as a computer. The method specifically includes: Step 110: Based on the target remote sensing image of the target field area, obtain the temporal fluctuation characteristics of water evapotranspiration and nitrogen absorption in the target field area during the target monitoring period.

[0019] The target field area typically refers to a cotton field area whose images have been acquired using remote sensing equipment. This area can be the entire cotton field, or preferably any one of several sub-areas obtained by dividing the entire field according to preset rules. The target remote sensing image can be spectral image data obtained by acquiring images of the target field area using remote sensing equipment. The target monitoring time period can be a time period set as needed for monitoring water and nitrogen stress in the cotton field. The temporal fluctuation characteristics of water evapotranspiration can be the characteristics of water evapotranspiration of the cotton in the target field area as reflected in the target remote sensing image changing over time within the target monitoring time period. The temporal fluctuation characteristics of nitrogen absorption can be the characteristics of nitrogen absorption of the cotton in the target field area as reflected in the target remote sensing image changing over time within the target monitoring time period.

[0020] Specifically, cotton fields are typically equipped with remote sensing acquisition equipment to continuously collect surface reflectance spectral images of the cotton field area at fixed time intervals. The target remote sensing images can be directly acquired by the remote sensing equipment or obtained by specific preprocessing of the raw image data acquired by the remote sensing equipment; there is no limitation here. However, it should be noted that the target remote sensing images in this embodiment usually need to include images acquired at multiple time points within the target monitoring period. Therefore, the target remote sensing images can reflect the changes in the surface reflectance spectrum of cotton in the target field area over time within the target monitoring period. Surface reflectance spectra can effectively characterize the water evaporation and nitrogen absorption of cotton, thus allowing for the analysis and extraction of temporal fluctuation characteristics of water evaporation and nitrogen absorption from the changes in surface reflectance spectra.

[0021] Optionally, the temporal fluctuation characteristics of cotton water evapotranspiration can be obtained based on the normalized difference vegetation index (NDVI) reflected in the target remote sensing image. This is because the NDVI is a key spectral index characterizing crop canopy coverage and growth vigor, reflecting the proportion of photosynthetically active radiation absorbed by the crop. Therefore, this index is related to the intensity of cotton water evapotranspiration. By analyzing the changes in the NDVI reflected in the target remote sensing image during the target monitoring period, the temporal fluctuation characteristics of cotton water evapotranspiration can be reflected.

[0022] Optionally, the temporal fluctuation characteristics of nitrogen absorption in cotton can be obtained based on the chlorophyll fluorescence index reflected in the remote sensing image of the target. This is because the chlorophyll fluorescence index is a spectral indicator reflecting the chlorophyll content and photosynthetic system activity of crop leaves, directly characterizing the nitrogen nutrition level of crops. Since this index is highly correlated with the nitrogen absorption of cotton, the temporal fluctuation characteristics of nitrogen absorption in cotton can be reflected by analyzing the changes in the chlorophyll fluorescence index reflected in the remote sensing image of the target during the target monitoring period.

[0023] In an optional implementation, based on the target remote sensing image of the target field area, the temporal fluctuation characteristics of water evapotranspiration and nitrogen absorption in the target field area during the target monitoring time period are obtained, including: sequentially acquiring the normalized difference vegetation index (NDV) in the target remote sensing image within each acquisition time period and determining whether each NPV is lower than a preset water stress threshold; when there is a consecutive number of targets with NPVs lower than the preset water stress threshold, the acquisition time period corresponding to the consecutive number of targets with NPVs is determined as the target monitoring time period; acquiring the temporal fluctuation characteristics of water evapotranspiration based on the NPV within the target monitoring time period; acquiring the chlorophyll fluorescence index in the target remote sensing image within the target monitoring time period; and acquiring the temporal fluctuation characteristics of nitrogen absorption based on the chlorophyll fluorescence index.

[0024] The data acquisition period can be a preset time period, with at least one remote sensing image of the target acquired in each acquisition period. Normalized Difference Vegetation Index (NDV) is obtained from the target remote sensing images within any acquisition period. Specifically, this can be achieved by extracting near-infrared and red reflectance from multi-temporal remote sensing images and calculating the NVC using a standard formula. Each NVC value is compared with a preset water stress threshold, which is a pre-defined critical value for the NVC used to determine the risk of water stress in cotton. Therefore, when the NVC value for a consecutive number of targets is lower than the preset water stress threshold, it indicates that the cotton evapotranspiration in the target field has remained consistently low over a sufficient number of consecutive acquisition periods, indicating abnormal water use behavior and a clear trend towards water stress. This suggests a potential for water and nitrogen stress, necessitating water and nitrogen stress monitoring of the target field area during this period, thus defining the corresponding acquisition period as the target monitoring time period. Optionally, the number of consecutive targets can be set to three.

[0025] Furthermore, after determining the target monitoring time period, the chlorophyll fluorescence index can be obtained by using the spectral reflectance characteristics of the red edge band and fluorescence sensitive band of the remote sensing image. Then, the temporal fluctuation characteristics of water evapotranspiration and nitrogen absorption can be obtained based on the normalized difference vegetation index and the chlorophyll fluorescence index, respectively.

[0026] Optionally, the temporal fluctuation characteristics of water evapotranspiration and nitrogen uptake can refer to the amplitude and frequency distributions of water evapotranspiration and nitrogen uptake values ​​in the time domain, respectively. Specifically, the preliminary estimate of water evapotranspiration obtained from the normalized difference vegetation index can be segmented for each time window to obtain the amplitude and frequency information of water evapotranspiration within each time window, serving as the temporal fluctuation characteristics of water evapotranspiration. Similarly, the temporal data of nitrogen uptake obtained from the chlorophyll fluorescence index can be segmented within the same time window to obtain its amplitude and frequency information, serving as the temporal fluctuation characteristics of nitrogen uptake.

[0027] For example, Landsat-8 imagery data of the target field area over the past three months can be obtained through a satellite remote sensing data platform. This data has a temporal resolution of 16 days and can contain six periods of target remote sensing images with a spatial resolution of 30 meters. Each period of the target remote sensing image can be preprocessed, for example, including radiometric and atmospheric corrections to eliminate the effects of cloud and atmospheric interference, obtaining accurate spectral reflectance data. Reflectance values ​​in the near-infrared band (0.85-0.88 micrometers) and the red band (0.64-0.67 micrometers) can then be extracted. The corresponding Normalized Difference Vegetation Index (NDVI) can be calculated using the formula NDVI = (NIR - Red) / (NIR + Red), where NDVI represents the Normalized Difference Vegetation Index, NIR represents near-infrared reflectance, and Red represents red reflectance. For example, if the near-infrared reflectance in any target remote sensing image is 0.45 and the red reflectance is 0.15, the corresponding Normalized Difference Vegetation Index is 0.5. This method analyzes all six target remote sensing images one by one, generating a normalized difference vegetation index (NDVI) time series, such as 0.5, 0.48, 0.42, 0.38, 0.35, and 0.32. Based on this series, preliminary evapotranspiration (ET) values ​​are further estimated. For example, a simplified linear model ET = NDVI * 100 can be used to calculate preliminary evapotranspiration estimates of 50, 48, 42, 38, 35, and 32, in mm / day. The preset water stress threshold corresponds to a preliminary evapotranspiration estimate of 40. If three consecutive preliminary evapotranspiration estimates are below 40, for example, the last three values ​​in the above series (38, 35, and 32) are all below the threshold of 40, it indicates that the normalized difference vegetation index is below the preset water stress threshold for three consecutive time periods. These three time periods can be defined as the target monitoring period, totaling 48 days. The chlorophyll fluorescence index for this period is then calculated, for example, 0.8, 0.75, and 0.7, in mW / m². 2Based on the relationship between chlorophyll fluorescence index and nitrogen absorption, / sr / nm can be directly used as nitrogen absorption time-series data. Furthermore, by segmenting the aforementioned preliminary estimates of water evapotranspiration and nitrogen absorption time-series data within each preset time window, such as daily, the fluctuation amplitude and frequency of each data point are obtained. Thus, based on the data corresponding to each time window within the target monitoring period, the time-series fluctuation characteristics of water evapotranspiration and nitrogen absorption are obtained respectively.

[0028] Step 120: Obtain the joint trend data between the temporal fluctuation characteristics of water evapotranspiration and the temporal fluctuation characteristics of nitrogen absorption.

[0029] Among them, the joint trend data can be data used to describe the dynamic relationship between water evapotranspiration and nitrogen uptake in cotton in the target field area, as reflected by the temporal fluctuation characteristics of water evapotranspiration and nitrogen uptake.

[0030] Specifically, by jointly analyzing the temporal fluctuation characteristics of water evapotranspiration and nitrogen absorption, we can obtain the changing characteristics of water evapotranspiration and nitrogen absorption at the same time, thereby extracting the dynamic correlation between the two and forming joint trend data.

[0031] In an optional implementation, acquiring joint trend data between temporal fluctuation characteristics of water evapotranspiration and temporal fluctuation characteristics of nitrogen absorption includes: calculating the correlation intensity parameter between temporal fluctuation characteristics of water evapotranspiration and temporal fluctuation characteristics of nitrogen absorption within each target time window; identifying joint change time intervals with a joint change relationship between temporal fluctuation characteristics of water evapotranspiration and temporal fluctuation characteristics of nitrogen absorption within each target time window based on the correlation intensity parameter; calculating the rate of change of water evapotranspiration and the rate of change of nitrogen absorption within the joint change time interval based on the temporal fluctuation characteristics of water evapotranspiration and the temporal fluctuation characteristics of nitrogen absorption, respectively; and generating joint trend data based on the rate of change of water evapotranspiration and the rate of change of nitrogen absorption.

[0032] The target time series window can be a pre-defined time period of a certain length, specifically set according to the length of the target monitoring period, so that the target monitoring period contains a certain number of consecutive target time series windows. The correlation intensity parameter can be a parameter describing the degree of correlation between the temporal fluctuation characteristics of moisture evapotranspiration and nitrogen absorption within the target time series window. Therefore, calculating the correlation intensity parameter for the temporal fluctuation characteristics of moisture evapotranspiration and nitrogen absorption in each target time series window can reflect the correlation between moisture evapotranspiration and nitrogen absorption within each target time series window.

[0033] Optionally, the target time window can be each time window corresponding to the fluctuation amplitude and frequency in the time fluctuation characteristics of water evapotranspiration and nitrogen absorption. The relevant intensity parameters can include the correlation coefficient between the fluctuation amplitudes of water evapotranspiration and nitrogen absorption and the correlation coefficient between the frequencies within the target time window.

[0034] Furthermore, the joint variation relationship can refer to a highly correlated relationship between the temporal fluctuation characteristics of water evapotranspiration and nitrogen absorption, which can be determined based on the description of the correlation strength parameters. The joint variation time series interval is the target time series window in which the temporal fluctuation characteristics of water evapotranspiration and nitrogen absorption exhibit a joint variation relationship. Specifically, if the correlation strength parameter exceeds a preset correlation strength threshold, a joint variation relationship between the temporal fluctuation characteristics of water evapotranspiration and nitrogen absorption can be determined, and the corresponding target time series window can be defined as the joint variation time series interval. The preset correlation strength threshold can be a pre-set critical correlation coefficient value used to determine whether there is a significant joint variation relationship between the temporal fluctuations of water evapotranspiration and nitrogen absorption.

[0035] Within the combined change time series, the rates of change in water evapotranspiration and nitrogen absorption can be calculated based on their temporal fluctuation characteristics. The rate of change in water evapotranspiration can refer to the rate of change in water evapotranspiration in the target field area, and the rate of change in nitrogen absorption can refer to the rate of change in nitrogen absorption in the target field area. Based on the trend direction, comparison results, and matching degree of the rates of change in water evapotranspiration and nitrogen absorption, the coordinated change trend of water evapotranspiration and nitrogen absorption in each target time series window can be determined, thereby establishing the dynamic mapping relationship between water evapotranspiration and nitrogen absorption in the time domain and generating corresponding combined change trend data.

[0036] For example, within a 48-day target monitoring period, eight target time-series windows can be set. Based on the daily water evapotranspiration and nitrogen absorption time-series fluctuation characteristics over six days in each target time-series window, the amplitude correlation coefficient and frequency correlation coefficient of the two can be calculated using the Pearson correlation coefficient formula, and then fused into a correlation strength parameter with preset weights. When the correlation strength parameter is sufficiently high, for example, when the Pearson correlation coefficient reaches 0.85, it indicates that there is a joint change relationship between the water evapotranspiration and nitrogen absorption time-series fluctuation characteristics, and the corresponding target time-series window is determined as the joint change time-series interval. Subsequently, the rate of change of water evapotranspiration and the rate of change of nitrogen absorption can be obtained by calculating the amplitude difference between each two adjacent days and averaging the results, and a joint change trend data can be constructed to describe the daily mapping relationship between the two.

[0037] Step 130: Identify abnormal fluctuation points of water evapotranspiration and nitrogen absorption in the residual terms of the joint trend data, and obtain the stress type classification results for each abnormal fluctuation point of water evapotranspiration and nitrogen absorption.

[0038] The residual term can be any data item in the joint trend data that describes the rate of change of water evapotranspiration and the rate of change of nitrogen uptake but does not exhibit a coordinated trend. Anomaly points in water evapotranspiration fluctuations can be the time points corresponding to the residual term where the rate of change of water evapotranspiration changes abnormally, and anomaly points in nitrogen uptake fluctuations can be the time points corresponding to the residual term where the rate of change of nitrogen uptake changes abnormally.

[0039] Specifically, by processing the joint trend data using a time series decomposition algorithm, the trend term and residual term can be separated. Based on the specific values ​​and changes of the rate of change of water evapotranspiration and the rate of change of nitrogen absorption at the corresponding time points of the residual term, it can be determined whether the time point corresponding to each residual term is an abnormal fluctuation point of water evapotranspiration or an abnormal fluctuation point of nitrogen absorption, thereby identifying all abnormal fluctuation points of water evapotranspiration and nitrogen absorption.

[0040] Furthermore, based on the abnormal fluctuations in water evapotranspiration and nitrogen absorption at each point, the corresponding stress type classification can be determined by analyzing the relationship between the two, according to pre-defined stress type classification rules. The stress type classification can be based on the stress situation caused by the main factors leading to the stress.

[0041] In an optional implementation, abnormal fluctuation points of water evapotranspiration and nitrogen absorption are identified in the residual terms of the joint trend data, and the stress type classification results of each abnormal fluctuation point of water evapotranspiration and nitrogen absorption are obtained. This includes: obtaining the trend term and residual term of the joint trend data based on a time series decomposition algorithm; identifying abnormal fluctuation points of water evapotranspiration and nitrogen absorption in the residual terms according to a preset fluctuation threshold; calculating the phase difference data between the time-series fluctuation characteristics of water evapotranspiration and nitrogen absorption at each abnormal fluctuation point of water evapotranspiration and nitrogen absorption; and performing cluster analysis on the abnormal fluctuation points of water evapotranspiration and nitrogen absorption based on the phase difference data to obtain the stress type classification results.

[0042] In the joint trend data, the trend term describes the coordinated trend of changes in the rates of water evapotranspiration and nitrogen absorption. Therefore, the trend term and the residual term can be distinguished in the joint trend data. The preset fluctuation threshold can be a pre-set critical value used to determine if the rate of change in water evapotranspiration or nitrogen absorption shows a significant abnormality. If the rate of change in water evapotranspiration or nitrogen absorption corresponding to any residual term exceeds the corresponding preset fluctuation threshold, it can be said that the rate of change in water evapotranspiration or nitrogen absorption has experienced abnormal fluctuations. The time point corresponding to the residual term can be determined as the abnormal fluctuation point of water evapotranspiration or nitrogen absorption.

[0043] Furthermore, by calculating the phase difference data between the temporal fluctuation characteristics of water evapotranspiration and nitrogen absorption at each abnormal fluctuation point of water evapotranspiration and nitrogen absorption, the specific value and direction of the phase difference can be obtained, thus determining the difference in the synchronicity of water and nitrogen changes among the fluctuation points. Based on the value and direction of the phase difference, the fluctuation points can be clustered to obtain a preliminary set of fluctuation point classifications. Since the phase difference can characterize the fluctuation characteristics and rate of change of water evapotranspiration and nitrogen absorption, it can be used as a basis for judging the stress type. Specifically, the stress type can be classified according to the sign of the phase difference and the range of its value.

[0044] For example, a classic seasonal-trend decomposition algorithm can be used to decompose the joint trend data into a trend term, a seasonal term, and a residual term. Abnormal fluctuation points are extracted from the residual term. Using the Z-score method with a threshold of 2.5, at least one abnormal point in water evapotranspiration residual and at least one abnormal fluctuation point in nitrogen absorption are identified in the residual term. Subsequently, the instantaneous phase is extracted using Hilbert transform, and the phase difference between the temporal fluctuation characteristics of water evapotranspiration and nitrogen absorption is calculated. For example, if the phase difference on day n is positive 0.3 rad, it indicates that water change slightly leads nitrogen change. Further, clustering is performed based on the sign and value of the phase difference. Using the K-means algorithm, the phase difference can be divided into three categories: a large positive value category greater than 0.5 rad, a small positive value category between 0 and 0.5 rad, and a negative value category. Day n can then be classified into the small positive value category, representing a stress type dominated by mild water evapotranspiration.

[0045] Step 140: Based on the stress type classification results, identify the water-dominant stress points among the abnormal fluctuation points of water evapotranspiration and nitrogen absorption, and obtain the stress degree quantification score of the water-dominant stress points.

[0046] Among them, the water-dominant stress point can be either the abnormal fluctuation point of water evapotranspiration or the abnormal fluctuation point of nitrogen absorption, where the dominant stress factor of the stress type is water evapotranspiration. The stress degree quantification score can be data that quantitatively describes the stress degree caused by water evapotranspiration as the main factor at the water-dominant stress point.

[0047] Specifically, based on the stress type classification results, the main factors causing stress at each fluctuation point can be identified. If the main factor is water evaporation, it can be determined as a water-dominant stress point. Furthermore, based on the relevant data on water evaporation and nitrogen absorption at the water-dominant stress point, the degree of water and nitrogen stress can be quantitatively assessed, resulting in a stress degree quantitative score.

[0048] In one optional implementation, based on the stress type classification results, water-dominant stress points are identified among the abnormal fluctuation points of water evapotranspiration and nitrogen absorption, and a stress severity quantification score is obtained for the water-dominant stress points. This includes: obtaining a subset of dominant factors for each stress type classification result; identifying water evapotranspiration and / or nitrogen absorption abnormal fluctuation points whose proportion of water evapotranspiration-related factors in the subset of dominant factors is higher than a preset factor threshold as water-dominant stress points; obtaining historical remote sensing baseline data corresponding to the water-dominant stress points; and calculating a stress severity quantification score based on the historical remote sensing baseline data and the temporal fluctuation characteristics of water evapotranspiration at the water-dominant stress points.

[0049] The dominant factor subset can be a subset containing data describing various factors leading to water and nitrogen stress, including data on factors related to the corresponding stress type classification results. Water evapotranspiration-related factors can be data describing factors related to water evapotranspiration. The preset factor threshold can be a pre-defined critical value for determining the dominant contribution of water evapotranspiration-related factors in the stress causes. If the proportion of water evapotranspiration-related factors in the dominant factor subset is higher than the preset factor threshold, it indicates that the influence of factors related to water evapotranspiration among the stress-causing factors in the stress type classification results is sufficiently large, and therefore the corresponding fluctuation point can be determined as the water-dominant stress point. The proportion of water evapotranspiration-related factors can, for example, refer to the proportion of the number of water evapotranspiration-related factors to the total number, preferably the proportion of the influence of water evapotranspiration-related factors to the total influence of all factors within the subset.

[0050] Furthermore, the historical remote sensing baseline data corresponding to the water-dominant stress point can refer to the data describing water evapotranspiration extracted from remote sensing images acquired during the same historical period at the water-dominant stress point. Therefore, by comparing the historical remote sensing baseline data with the temporal fluctuation characteristics of water evapotranspiration at the water-dominant stress point, the degree of anomaly in water evapotranspiration at the water-dominant stress point can be quantitatively assessed, thereby obtaining a stress intensity quantification score.

[0051] For example, the dominant factor subset corresponding to any stress type classification result may include factors related to water evapotranspiration such as soil salinity, temperature anomalies, sunshine duration, and rainfall, as well as factors unrelated to water evapotranspiration such as nitrogen fertilizer application amount and proportion. The contribution of each factor to the stress type classification result can be obtained according to preset rules or algorithm models. For example, if the contribution of soil salinity, temperature anomalies, sunshine duration, rainfall, nitrogen fertilizer application amount and proportion are 0.03, 0.07, 0.20, 0.15, 0.20, and 0.18 respectively, then the contribution of factors related to water evapotranspiration is 0.45, and the total contribution of all factors is 0.83. If the proportion of factors related to water evapotranspiration exceeds the preset factor threshold of 50%, the fluctuation point corresponding to the stress type classification result can be determined as the water-dominant stress point. Subsequently, historical remote sensing images from the same period are acquired to obtain historical remote sensing benchmark data, such as daily evapotranspiration data provided by the MOD16 data product. The average evapotranspiration sequence of each remote sensing image pixel from the same period in history is calculated as the benchmark. The deviation of the preliminary estimate of water evapotranspiration sequence, characterized by the amplitude of the temporal fluctuation characteristics of water evapotranspiration at the water-dominant stress point, from the benchmark is calculated. The regional spatial environmental differences between the two time points can be combined with the spatial heterogeneity parameter to weightedly calculate the stress degree quantification score. For example, if the average deviation is -35% and the standard deviation of the spatial heterogeneity parameter in the region is 12.5%, it can be normalized to 70 points and 25 points respectively. Based on the preset deviation weight of 0.7 and spatial heterogeneity weight of 0.3, the final stress degree quantification score of 56.5 points is obtained.

[0052] Step 150: Obtain the stress identification parameter set based on the feature mapping relationship between the stress degree quantification score and the temporal fluctuation characteristics of water evapotranspiration and nitrogen absorption at the water-dominant stress point.

[0053] The stress identification parameter set can be the parameters used in calculating the stress intensity quantification score based on the temporal fluctuation characteristics of water evapotranspiration and nitrogen absorption. Based on the feature mapping relationship between the stress intensity quantification score of the currently identified water-dominant stress point and its temporal fluctuation characteristics of water evapotranspiration and nitrogen absorption at that time point, a corresponding stress identification parameter set can be obtained. This ensures that after calculating the stress intensity quantification score based on the temporal fluctuation characteristics of water evapotranspiration and nitrogen absorption at any time point using this stress identification parameter set, the numerical relationship between the obtained stress intensity quantification score and the corresponding temporal fluctuation characteristics of water evapotranspiration and nitrogen absorption at that time point conforms to the feature mapping relationship. Therefore, this stress identification parameter set can be used to calculate the stress intensity quantification score based on the temporal fluctuation characteristics of water evapotranspiration and nitrogen absorption at any time point.

[0054] In an optional implementation, a stress identification parameter set is obtained based on the feature mapping relationship between the stress intensity quantification score and the temporal fluctuation characteristics of water evapotranspiration and nitrogen absorption at the water-dominant stress point, including: Based on the random forest algorithm, a mapping relationship model is built between the stress degree quantification score and the rate of change of water evapotranspiration, the rate of change of nitrogen absorption, and the phase difference data to obtain the water evapotranspiration weight coefficient and the nitrogen absorption weight coefficient. The water stress threshold parameters and nitrogen stress threshold parameters are obtained based on the water evapotranspiration weight coefficient and nitrogen absorption weight coefficient. The water evapotranspiration weight coefficient, nitrogen absorption weight coefficient, water stress threshold parameters and nitrogen stress threshold parameters are determined as the stress identification parameter set.

[0055] Phase difference data refers to the data describing the phase difference in the time-domain distribution between the water evapotranspiration rate sequence and the nitrogen uptake rate sequence. Water evapotranspiration weighting coefficients and nitrogen uptake weighting coefficients can be data describing the influence of water evapotranspiration and nitrogen stress factors on the final calculated stress intensity quantification score, respectively. Water stress threshold parameters and nitrogen stress threshold parameters can be threshold data used to determine whether water and nitrogen stress has occurred based on the normalized difference vegetation index and chlorophyll fluorescence index, respectively.

[0056] Specifically, by using the rates of change in water evapotranspiration, nitrogen uptake, and phase difference as input features, and the stress intensity quantification score as the target variable, a random forest algorithm is employed for training to model the mapping relationship. Cross-validation is used during training to iteratively calibrate the weight coefficients of water evapotranspiration and nitrogen uptake. This allows the extraction of these weight coefficients from the mapping relationship modeling results. Based on these modeling results and the water and nitrogen stress threshold parameters, water stress threshold parameters and nitrogen stress threshold parameters can be further obtained, forming a stress identification parameter set.

[0057] Step 160: Based on the stress identification parameter set, perform water and nitrogen stress identification on each remote sensing spatial grid of the real-time remote sensing image to obtain a water and nitrogen stress distribution map.

[0058] The real-time remote sensing image can be a reflectance spectrum image of a cotton field acquired in real time, and the remote sensing spatial grid can be a rasterized region division of the real-time remote sensing image according to preset rules. Water and nitrogen stress identification can be the operation of identifying whether the cotton in the field area reflected by each remote sensing spatial grid of the real-time remote sensing image is under water and nitrogen stress. The water and nitrogen stress distribution map can be a regional image composed of the water and nitrogen stress identification results of each remote sensing spatial grid of the real-time remote sensing image, reflecting the spatial distribution of the water and nitrogen stress status of cotton in the field area of ​​the real-time remote sensing image.

[0059] Specifically, based on the stress identification parameter set obtained from the aforementioned steps, water and nitrogen stress can be identified in cotton fields in any region. By dividing the real-time remote sensing image into multiple remote sensing spatial grids and identifying water and nitrogen stress in each grid, the water and nitrogen stress status of the corresponding field area can be obtained, thus forming a water and nitrogen stress distribution map.

[0060] In one optional implementation, water and nitrogen stress is identified for each remote sensing spatial grid in a real-time remote sensing image based on a stress identification parameter set to obtain a water and nitrogen stress distribution map. This includes: obtaining the index feature deviation between the remote sensing image inversion index of each remote sensing spatial grid and the stress identification parameter set; determining the stress identification result of each remote sensing spatial grid based on the index feature deviation; and generating a water and nitrogen stress distribution map based on each stress identification result.

[0061] Among them, the remote sensing image inversion index can be a factor index described by data related to the identification of water and nitrogen stress extracted from remote sensing spatial raster images. The index feature bias is data describing the difference between the remote sensing image inversion index and the corresponding parameter type in the stress identification parameter set.

[0062] Specifically, for each remote sensing spatial grid in the real-time remote sensing image, sequence inversion technology is used to process and extract the corresponding remote sensing image inversion indices. These indices are then compared with a stress identification parameter set to obtain the indicator feature deviations used to determine the stress identification results. Optionally, the remote sensing image inversion indices may include a preliminary estimate of water evapotranspiration and chlorophyll fluorescence index. The indicator feature deviations are obtained between the preliminary estimate of water evapotranspiration and the water stress threshold parameter in the stress identification parameter set, and between the chlorophyll fluorescence index and the chlorophyll fluorescence index corresponding to the nitrogen stress threshold parameter in the stress identification parameter set. These deviations are then weighted and evaluated based on the water evapotranspiration weight coefficient and nitrogen absorption weight coefficient in the stress identification parameter set to ultimately obtain the stress identification result.

[0063] In an optional implementation, the stress identification result of each remote sensing spatial grid is determined based on the index feature deviation, including: based on the index feature deviation, if the preliminary estimate of water evapotranspiration of the remote sensing image inversion index is lower than the water stress threshold parameter, and the numerical relationship between the chlorophyll fluorescence index and the nitrogen stress threshold parameter meets the preset stability conditions, the stress identification result is determined to be nitrogen-dominated stress.

[0064] The preset stability condition can refer to the chlorophyll fluorescence index being within the nitrogen stress threshold range and the temporal change tending to be stable. If the numerical relationship between the chlorophyll fluorescence index and the nitrogen stress threshold parameter meets the preset stability condition, it can be said that a nitrogen absorption imbalance has occurred due to continuous water shortage. In this case, the stress identification result can be determined to be nitrogen-dominated stress, thereby distinguishing between stress dominated by simple water evaporation and stress dominated by nitrogen absorption caused by water evaporation.

[0065] In an optional implementation, after generating a water and nitrogen stress distribution map based on the stress identification results, the method further includes: identifying the coordinates of the target risk area in the water and nitrogen stress distribution map; and generating a temporal stress intervention signal for the target risk area coordinates based on the remote sensing image inversion index corresponding to the risk area coordinates.

[0066] Among them, the target risk area coordinates can be used to locate the field area with more severe water and nitrogen stress, and the time-domain stress intervention signal can be used to dynamically intervene in the relevant factors of the field area located by the target risk area coordinates over time, so as to reduce or eliminate the water and nitrogen stress in the area.

[0067] Specifically, based on the stress identification results in the water and nitrogen stress distribution map, high-risk image areas of water and nitrogen stress are identified, and the coordinates of the target risk areas corresponding to these image areas are obtained. Furthermore, the remote sensing image inversion indicators corresponding to the risk area coordinates can specifically characterize the degree of water and nitrogen stress, the dominant factors, and their temporal variation patterns in that area. Based on this, corresponding temporal stress intervention signals can be generated, such as time-series intervention signals for irrigation and topdressing amounts.

[0068] Optionally, intervention priorities can be ranked according to the degree of water and nitrogen stress risk of the target risk area coordinates, so as to prioritize intervention in field areas with severe water and nitrogen stress.

[0069] The technical solution of this embodiment obtains the temporal fluctuation characteristics of water evapotranspiration and nitrogen uptake in the target field area during the target monitoring period based on the target remote sensing image of the target field area. It also obtains joint trend data between these two characteristics, identifies anomalous fluctuation points in water evapotranspiration and nitrogen uptake from the residuals of the joint trend data, and obtains the stress type classification results for each anomalous fluctuation point. Based on these stress type classification results, it identifies water-dominant stress points among the anomalous fluctuation points in water evapotranspiration and nitrogen uptake, and obtains a quantitative score for the stress degree of these water-dominant stress points. Based on the feature mapping relationship between the stress degree quantification score and the temporal fluctuation characteristics of water evapotranspiration and nitrogen absorption at the water-dominant stress point, a stress identification parameter set is obtained. Then, based on this parameter set, water and nitrogen stress is identified in each remote sensing spatial grid of the real-time remote sensing image, resulting in a water and nitrogen stress distribution map. This solves the problem of existing technologies lacking the ability to accurately identify the causes and effects of cotton water and nitrogen stress by comprehensively analyzing the dynamic changes of factors such as water evapotranspiration and nitrogen absorption based on observations over multiple time periods. It achieves online monitoring of cotton water and nitrogen stress by integrating multi-source dynamic features, significantly improving the spatiotemporal accuracy of stress identification and the targeting of intervention measures, providing a scientific basis for precision agricultural management of cotton.

[0070] Figure 2 This is a schematic diagram of the structure of an online monitoring system for water and nitrogen stress in cotton that integrates multi-temporal remote sensing data, provided as an embodiment of this application. Figure 2 As shown, the system includes: a feature acquisition module 210, a joint trend module 220, a stress classification module 230, a degree quantification module 240, a parameter acquisition module 250, and a stress identification module 260, wherein... The feature acquisition module 210 is used to acquire the temporal fluctuation characteristics of water evapotranspiration and nitrogen absorption of the target field area within the target monitoring time period based on the target remote sensing image of the target field area. The joint trend module 220 is used to acquire joint trend data between the temporal fluctuation characteristics of water evapotranspiration and the temporal fluctuation characteristics of nitrogen absorption. The stress classification module 230 is used to identify abnormal fluctuation points of water evapotranspiration and abnormal fluctuation points of nitrogen absorption in the residual terms of the joint trend data, and to obtain the stress type classification results of each abnormal fluctuation point of water evapotranspiration and abnormal fluctuation point of nitrogen absorption. The stress quantification module 240 is used to identify water-dominant stress points among the abnormal fluctuation points of water evapotranspiration and the abnormal fluctuation points of nitrogen absorption based on the stress type classification results, and to obtain the stress quantification score of the water-dominant stress points. The parameter acquisition module 250 is used to acquire a set of stress identification parameters based on the feature mapping relationship between the stress degree quantification score and the water evapotranspiration time-series fluctuation characteristics and the nitrogen absorption time-series fluctuation characteristics of the water-dominant stress point. The stress identification module 260 is used to identify water and nitrogen stress in each remote sensing spatial grid of the real-time remote sensing image based on the stress identification parameter set, and obtain a water and nitrogen stress distribution map.

[0071] The online monitoring system for water and nitrogen stress in cotton that integrates multi-temporal remote sensing data provided in this application can execute the online monitoring method for water and nitrogen stress in cotton that integrates multi-temporal remote sensing data provided in any embodiment of this application, and has the corresponding functional modules and beneficial effects of the method.

[0072] The technical solution of this embodiment obtains the temporal fluctuation characteristics of water evapotranspiration and nitrogen uptake in the target field area during the target monitoring period based on the target remote sensing image of the target field area. It also obtains joint trend data between these two characteristics, identifies anomalous fluctuation points in water evapotranspiration and nitrogen uptake from the residuals of the joint trend data, and obtains the stress type classification results for each anomalous fluctuation point. Based on these stress type classification results, it identifies water-dominant stress points among the anomalous fluctuation points in water evapotranspiration and nitrogen uptake, and obtains a quantitative score for the stress degree of these water-dominant stress points. Based on the feature mapping relationship between the stress degree quantification score and the temporal fluctuation characteristics of water evapotranspiration and nitrogen absorption at the water-dominant stress point, a stress identification parameter set is obtained. Then, based on this parameter set, water and nitrogen stress is identified in each remote sensing spatial grid of the real-time remote sensing image, resulting in a water and nitrogen stress distribution map. This solves the problem of existing technologies lacking the ability to accurately identify the causes and effects of cotton water and nitrogen stress by comprehensively analyzing the dynamic changes of factors such as water evapotranspiration and nitrogen absorption based on observations over multiple time periods. It achieves online monitoring of cotton water and nitrogen stress by integrating multi-source dynamic features, significantly improving the spatiotemporal accuracy of stress identification and the targeting of intervention measures, providing a scientific basis for precision agricultural management of cotton.

[0073] It should be noted that, in this document, the terms "comprising," "including," and any other variations are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Specific examples have been used in this document to illustrate the principles and implementation methods of the technical solutions of this application. The above examples are only for the purpose of helping to understand the methods and core ideas of this application. The above descriptions are merely preferred embodiments of this application. It should be pointed out that, due to the limitations of written expression and the objective existence of infinite specific structures, those skilled in the art can make several improvements, modifications, or changes without departing from the principles of this application, and can also combine the above technical features in an appropriate manner; these improvements, modifications, changes, or combinations, or the direct application of the concept and technical solutions of this application to other situations without modification, should all be considered within the scope of protection of this application.

Claims

1. A method for online monitoring of water and nitrogen stress in cotton by integrating multi-temporal remote sensing data, characterized in that, include: Based on the target remote sensing image of the target field area, obtain the temporal fluctuation characteristics of water evapotranspiration and nitrogen absorption of the target field area during the target monitoring time period. Obtain the joint trend data between the temporal fluctuation characteristics of water evapotranspiration and the temporal fluctuation characteristics of nitrogen absorption; Identify abnormal fluctuation points of water evapotranspiration and nitrogen absorption in the residual terms of the joint trend data, and obtain the stress type classification results for each of the abnormal fluctuation points of water evapotranspiration and nitrogen absorption. Based on the stress type classification results, water-dominant stress points are identified among the abnormal fluctuation points of water evapotranspiration and the abnormal fluctuation points of nitrogen absorption, and the stress degree quantification score of the water-dominant stress points is obtained. Based on the feature mapping relationship between the stress degree quantification score and the temporal fluctuation characteristics of water evapotranspiration and nitrogen absorption at the water-dominant stress point, a stress identification parameter set is obtained. Based on the stress identification parameter set, water and nitrogen stress are identified in each remote sensing spatial grid of the real-time remote sensing image to obtain a water and nitrogen stress distribution map.

2. The method according to claim 1, characterized in that, The step of acquiring the temporal fluctuation characteristics of water evapotranspiration and nitrogen uptake in the target field area within the target monitoring time period based on the target remote sensing image of the target field area includes: Normalized difference vegetation index is obtained sequentially from the target remote sensing image in each acquisition time period, and it is determined whether each normalized difference vegetation index is lower than a preset water stress threshold. When there are consecutive target numbers of the normalized difference vegetation index below the preset water stress threshold, the collection time period corresponding to the normalized difference vegetation index of the consecutive target numbers is determined as the target monitoring time period. The temporal fluctuation characteristics of water evapotranspiration are obtained based on the normalized difference vegetation index within the target monitoring period. Acquire the chlorophyll fluorescence index from the remote sensing images of the target during the target monitoring period; The timing fluctuation characteristics of nitrogen absorption were obtained based on the chlorophyll fluorescence index.

3. The method according to claim 2, characterized in that, The acquisition of joint trend data between the temporal fluctuation characteristics of water evapotranspiration and the temporal fluctuation characteristics of nitrogen absorption includes: Calculate the correlation strength parameter between the temporal fluctuation characteristics of water evapotranspiration and the temporal fluctuation characteristics of nitrogen absorption within each target time window; Based on the relevant intensity parameters, joint change time intervals with a joint change relationship between the temporal fluctuation characteristics of water evapotranspiration and the temporal fluctuation characteristics of nitrogen absorption are identified in each of the target time windows; Based on the temporal fluctuation characteristics of water evapotranspiration and the temporal fluctuation characteristics of nitrogen absorption, the rates of change of water evapotranspiration and nitrogen absorption in the joint change time interval are calculated respectively. The combined trend data is generated based on the rate of change in water evaporation and the rate of change in nitrogen absorption.

4. The method according to claim 3, characterized in that, The process involves identifying anomalous fluctuations in water evapotranspiration and nitrogen uptake in the residuals of the joint trend data, and obtaining stress type classification results for each of these anomalous fluctuations, including: Based on the time series decomposition algorithm, the trend term and the residual term of the joint change trend data are obtained; Based on the preset fluctuation threshold, the abnormal fluctuation points of moisture evaporation and nitrogen absorption are identified in the residual term. Calculate the phase difference data between the temporal fluctuation characteristics of water evapotranspiration and the temporal fluctuation characteristics of nitrogen absorption at each of the abnormal fluctuation points of water evapotranspiration and the abnormal fluctuation points of nitrogen absorption; Based on the phase difference data, cluster analysis is performed on the abnormal fluctuation points of water evapotranspiration and the abnormal fluctuation points of nitrogen absorption to obtain the stress type classification results.

5. The method according to claim 4, characterized in that, Based on the stress type classification results, the process involves identifying water-dominant stress points among the abnormal fluctuation points of water evapotranspiration and nitrogen absorption, and obtaining a stress severity quantification score for each water-dominant stress point, including: Obtain the dominant factor subset of the classification results for each of the aforementioned stress types; The abnormal fluctuation points of water evapotranspiration and / or the abnormal fluctuation points of nitrogen absorption that are higher than the preset factor threshold in the subset of dominant factors are identified as water-dominant stress points. Obtain the historical remote sensing baseline data corresponding to the water-dominant stress points; The stress severity quantification score is calculated based on the historical remote sensing baseline data and the temporal fluctuation characteristics of water evapotranspiration at the water-dominant stress point.

6. The method according to claim 5, characterized in that, The stress identification parameter set is obtained based on the feature mapping relationship between the stress degree quantification score and the temporal fluctuation characteristics of water evapotranspiration and nitrogen absorption at the water-dominant stress point, including: Based on the random forest algorithm, a mapping relationship is modeled between the stress degree quantification score and the rate of change of water evapotranspiration, the rate of change of nitrogen absorption, and the phase difference data to obtain the weight coefficient of water evapotranspiration and the weight coefficient of nitrogen absorption. The water stress threshold parameter and nitrogen stress threshold parameter are obtained based on the water evapotranspiration weight coefficient and the nitrogen absorption weight coefficient, and the water stress threshold parameter, the nitrogen absorption weight coefficient, the water stress threshold parameter and the nitrogen stress threshold parameter are determined as the stress identification parameter set.

7. The method according to claim 6, characterized in that, The step of identifying water and nitrogen stress in each remote sensing spatial grid of the real-time remote sensing image based on the stress identification parameter set to obtain a water and nitrogen stress distribution map includes: Obtain the index feature deviation between the remote sensing image inversion index of each of the remote sensing spatial grids and the stress identification parameter set; The stress identification result of each remote sensing spatial grid is determined based on the deviation of the index characteristics; The water and nitrogen stress distribution map is generated based on the stress identification results.

8. The method according to claim 7, characterized in that, The step of determining the stress identification result of each remote sensing spatial grid based on the deviation of the index features includes: Based on the characteristic deviation of the index, if the preliminary estimate of water evapotranspiration of the remote sensing image inversion index is lower than the water stress threshold parameter, and the numerical relationship between the chlorophyll fluorescence index and the nitrogen stress threshold parameter meets the preset stability conditions, the stress identification result is determined to be nitrogen-dominated stress.

9. The method according to claim 7, characterized in that, After generating the water nitrogen stress distribution map based on each of the stress identification results, the method further includes: Identify the coordinates of the target risk area in the water and nitrogen stress distribution map; Based on the remote sensing image inversion index corresponding to the coordinates of the risk area, a temporal stress intervention signal for the coordinates of the target risk area is generated.

10. An online monitoring system for water and nitrogen stress in cotton that integrates multi-temporal remote sensing data, characterized in that, include: The feature acquisition module is used to acquire the temporal fluctuation characteristics of water evapotranspiration and nitrogen absorption of the target field area within the target monitoring time period based on the target remote sensing image of the target field area. The joint trend module is used to obtain joint trend data between the temporal fluctuation characteristics of water evapotranspiration and the temporal fluctuation characteristics of nitrogen absorption. The stress classification module is used to identify abnormal fluctuation points of water evapotranspiration and abnormal fluctuation points of nitrogen absorption in the residual terms of the joint trend data, and to obtain the stress type classification results of each abnormal fluctuation point of water evapotranspiration and abnormal fluctuation point of nitrogen absorption. The stress severity quantification module is used to identify water-dominant stress points among the abnormal fluctuation points of water evapotranspiration and the abnormal fluctuation points of nitrogen absorption based on the stress type classification results, and to obtain the stress severity quantification score of the water-dominant stress points. The parameter acquisition module is used to acquire a set of stress identification parameters based on the feature mapping relationship between the stress degree quantification score and the temporal fluctuation characteristics of water evapotranspiration and nitrogen absorption at the water-dominant stress point. The stress identification module is used to identify water and nitrogen stress in each remote sensing spatial grid of the real-time remote sensing image based on the stress identification parameter set, and obtain a water and nitrogen stress distribution map.

Images